From thibaut.barrere at gmail.com Fri Dec 1 03:19:41 2006 From: thibaut.barrere at gmail.com (=?ISO-8859-1?Q?Thibaut_Barr=E8re?=) Date: Fri, 1 Dec 2006 09:19:41 +0100 Subject: [Rubygems-developers] Good practices to freeze gems ? (outside rails) In-Reply-To: <9D66D500-BB41-4D66-84A5-9E1CFBC5B78C@segment7.net> References: <4a68b8cf0611300023i14a3d1abj8b9a878e96d48aae@mail.gmail.com> <9D66D500-BB41-4D66-84A5-9E1CFBC5B78C@segment7.net> Message-ID: <4a68b8cf0612010019x21b73865m2d09a9d5e761eaaf@mail.gmail.com> Thanks for these answers. I'll try them out. Thibaut -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20061201/04fb7b85/attachment.html From anatol.pomozov at gmail.com Sat Dec 2 02:36:57 2006 From: anatol.pomozov at gmail.com (Cerberus) Date: Sat, 2 Dec 2006 10:36:57 +0300 Subject: [Rubygems-developers] [rubygems] Build broken by drbrain (#1109) Message-ID: <45712d19ea600_be419cdaf011a@ubuntu.tmail> ------------------------------------------------------------------------ r1109 | drbrain | 2006-12-01 03:44:41 +0300 | 1 line Changed paths: M /trunk/ChangeLog M /trunk/lib/rubygems/specification.rb M /trunk/test/test_specification.rb Allow gemspecs to be used intelligently with Array#-, Array#uniq ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/rake/rake_test_loader Started .........................................................................................................F...........................................EEE.E.E..................................................................................................... Finished in 14.891 seconds. 1) Failure: test_cache_data_none_readable(TestGemSourceInfoCache) [./test/test_gem_source_info_cache.rb:82]: <{}> expected but was <{"key"=>"sys"}>. 2) Error: test_explicit_proxy(TestRemoteFetcher): ArgumentError: unrecognized option: proxy_http_basic_authentication c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:106:in `check_options' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:103:in `check_options' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:127:in `open_uri' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:531:in `open' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:86:in `open' ./lib/rubygems/remote_fetcher.rb:124:in `open_uri_or_path' ./lib/rubygems/remote_fetcher.rb:49:in `fetch_path' ./test/test_remote_fetcher.rb:73:in `test_explicit_proxy' ./lib/rubygems/user_interaction.rb:45:in `use_ui' ./lib/rubygems/user_interaction.rb:27:in `use_ui' ./test/test_remote_fetcher.rb:70:in `test_explicit_proxy' 3) Error: test_explicit_proxy_with_user_auth(TestRemoteFetcher): ArgumentError: unrecognized option: proxy_http_basic_authentication c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:106:in `check_options' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:103:in `check_options' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:127:in `open_uri' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:531:in `open' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:86:in `open' ./lib/rubygems/remote_fetcher.rb:124:in `open_uri_or_path' ./lib/rubygems/remote_fetcher.rb:49:in `fetch_path' ./test/test_remote_fetcher.rb:85:in `test_explicit_proxy_with_user_auth' ./lib/rubygems/user_interaction.rb:45:in `use_ui' ./lib/rubygems/user_interaction.rb:27:in `use_ui' ./test/test_remote_fetcher.rb:78:in `test_explicit_proxy_with_user_auth' 4) Error: test_explicit_proxy_with_user_auth_in_env(TestRemoteFetcher): ArgumentError: unrecognized option: proxy_http_basic_authentication c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:106:in `check_options' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:103:in `check_options' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:127:in `open_uri' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:531:in `open' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:86:in `open' ./lib/rubygems/remote_fetcher.rb:124:in `open_uri_or_path' ./lib/rubygems/remote_fetcher.rb:49:in `fetch_path' ./test/test_remote_fetcher.rb:118:in `test_explicit_proxy_with_user_auth_in_env' ./lib/rubygems/user_interaction.rb:45:in `use_ui' ./lib/rubygems/user_interaction.rb:27:in `use_ui' ./test/test_remote_fetcher.rb:110:in `test_explicit_proxy_with_user_auth_in_env' 5) Error: test_implicit_proxy(TestRemoteFetcher): ArgumentError: unrecognized option: proxy_http_basic_authentication c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:106:in `check_options' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:103:in `check_options' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:127:in `open_uri' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:531:in `open' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:86:in `open' ./lib/rubygems/remote_fetcher.rb:124:in `open_uri_or_path' ./lib/rubygems/remote_fetcher.rb:49:in `fetch_path' ./test/test_remote_fetcher.rb:145:in `test_implicit_proxy' ./lib/rubygems/user_interaction.rb:45:in `use_ui' ./lib/rubygems/user_interaction.rb:27:in `use_ui' ./test/test_remote_fetcher.rb:142:in `test_implicit_proxy' 6) Error: test_implicit_upper_case_proxy(TestRemoteFetcher): ArgumentError: unrecognized option: proxy_http_basic_authentication c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:106:in `check_options' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:103:in `check_options' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:127:in `open_uri' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:531:in `open' c:/PROGRA~1/ruby/lib/ruby/1.8/open-uri.rb:86:in `open' ./lib/rubygems/remote_fetcher.rb:124:in `open_uri_or_path' ./lib/rubygems/remote_fetcher.rb:49:in `fetch_path' ./test/test_remote_fetcher.rb:153:in `test_implicit_upper_case_proxy' ./lib/rubygems/user_interaction.rb:45:in `use_ui' ./lib/rubygems/user_interaction.rb:27:in `use_ui' ./test/test_remote_fetcher.rb:150:in `test_implicit_upper_case_proxy' 257 tests, 933 assertions, 1 failures, 5 errors rake aborted! Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib "c:/PRO...] (See full trace by running task with --trace) -- Cerberus 0.3.2, http://cerberus.rubyforge.org/ From jim.weirich at gmail.com Sat Dec 2 11:54:16 2006 From: jim.weirich at gmail.com (Jim Weirich) Date: Sat, 2 Dec 2006 11:54:16 -0500 Subject: [Rubygems-developers] Enhancement Proposal: Would like be able to programatically specify platform In-Reply-To: <01C905AA-0D98-4346-B444-7A68EDE44C0C@segment7.net> References: <01C905AA-0D98-4346-B444-7A68EDE44C0C@segment7.net> Message-ID: On 11/29/06, Eric Hodel wrote:> Can you discuss details? I'd like the tinderbox to not waste time > and bandwidth attempting to run non-platform gems. (And be capable > of running platform gems.) One of the things we will do is make the Ruby platform be for pure ruby gems only. Things that have a C extension will have a platform name of "C" or "EXT" (or something like that). This allows us to not install gems requiring a compile environment on computers that can't compile (e.g. most non-developer windows machines). Other than that, we will try to use the most appropriate hardware platform we can find that matches. We don't at the moment know exactly how to determine the "best" platform match, thus the reason for tattle which will hopefully report one what's being used today. -- -- -- Jim Weirich jim at weirichhouse.org http://onestepback.org ----------------------------------------------------------------- "Beware of bugs in the above code; I have only proved it correct, not tried it." -- Donald Knuth (in a memo to Peter van Emde Boas) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20061202/f23ee353/attachment.html From drbrain at segment7.net Sat Dec 2 14:56:19 2006 From: drbrain at segment7.net (Eric Hodel) Date: Sat, 2 Dec 2006 11:56:19 -0800 Subject: [Rubygems-developers] [rubygems] Build broken by drbrain (#1109) In-Reply-To: <45712d19ea600_be419cdaf011a@ubuntu.tmail> References: <45712d19ea600_be419cdaf011a@ubuntu.tmail> Message-ID: <3E403BF4-3714-4466-977D-3B956A5A82DD@segment7.net> On Dec 1, 2006, at 23:36 , Cerberus wrote: > ---------------------------------------------------------------------- > -- > r1109 | drbrain | 2006-12-01 03:44:41 +0300 | 1 line > Changed paths: > M /trunk/ChangeLog > M /trunk/lib/rubygems/specification.rb > M /trunk/test/test_specification.rb > > Allow gemspecs to be used intelligently with Array#-, Array#uniq > ---------------------------------------------------------------------- > -- > > (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) > Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/ > rake/rake_test_loader > Started > ...................................................................... > ...................................F.................................. > .........EEE.E.E...................................................... > ............................................... > Finished in 14.891 seconds. > > 1) Failure: > test_cache_data_none_readable(TestGemSourceInfoCache) [./test/ > test_gem_source_info_cache.rb:82]: > <{}> expected but was > <{"key"=>"sys"}>. Possibly... > 2) Error: > test_explicit_proxy(TestRemoteFetcher): > ArgumentError: unrecognized option: proxy_http_basic_authentication > 3) Error: > test_explicit_proxy_with_user_auth(TestRemoteFetcher): > ArgumentError: unrecognized option: proxy_http_basic_authentication > > 4) Error: > test_explicit_proxy_with_user_auth_in_env(TestRemoteFetcher): > ArgumentError: unrecognized option: proxy_http_basic_authentication > > 5) Error: > test_implicit_proxy(TestRemoteFetcher): > ArgumentError: unrecognized option: proxy_http_basic_authentication > > 6) Error: > test_implicit_upper_case_proxy(TestRemoteFetcher): > ArgumentError: unrecognized option: proxy_http_basic_authentication Lies! These were caused by the removal of rubygems/open-uri... -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From jim.weirich at gmail.com Sat Dec 2 15:03:28 2006 From: jim.weirich at gmail.com (Jim Weirich) Date: Sat, 2 Dec 2006 15:03:28 -0500 Subject: [Rubygems-developers] [rubygems] Build broken by drbrain (#1109) In-Reply-To: <3E403BF4-3714-4466-977D-3B956A5A82DD@segment7.net> References: <45712d19ea600_be419cdaf011a@ubuntu.tmail> <3E403BF4-3714-4466-977D-3B956A5A82DD@segment7.net> Message-ID: On 12/2/06, Eric Hodel wrote: > > On Dec 1, 2006, at 23:36 , Cerberus wrote: > > > ---------------------------------------------------------------------- > > -- > > r1109 | drbrain | 2006-12-01 03:44:41 +0300 | 1 line > > Changed paths: > > M /trunk/ChangeLog > > M /trunk/lib/rubygems/specification.rb > > M /trunk/test/test_specification.rb > > > > Allow gemspecs to be used intelligently with Array#-, Array#uniq > > ---------------------------------------------------------------------- > > -- > > > > (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) > > Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/ > > rake/rake_test_loader > > Started > > ...................................................................... > > ...................................F.................................. > > .........EEE.E.E...................................................... > > ............................................... > > Finished in 14.891 seconds. > > > > 1) Failure: > > test_cache_data_none_readable(TestGemSourceInfoCache) [./test/ > > test_gem_source_info_cache.rb:82]: > > <{}> expected but was > > <{"key"=>"sys"}>. > > Possibly... Don't know about this. I'm not seeing it. > 2) Error: > > test_explicit_proxy(TestRemoteFetcher): > > ArgumentError: unrecognized option: proxy_http_basic_authentication [... other errors elided ...] Lies! These were caused by the removal of rubygems/open-uri... I've got these fixed locally ... will commit soon. -- -- -- Jim Weirich jim at weirichhouse.org http://onestepback.org ----------------------------------------------------------------- "Beware of bugs in the above code; I have only proved it correct, not tried it." -- Donald Knuth (in a memo to Peter van Emde Boas) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20061202/79ff4369/attachment-0001.html From jim.weirich at gmail.com Sat Dec 2 15:57:36 2006 From: jim.weirich at gmail.com (Jim Weirich) Date: Sat, 2 Dec 2006 15:57:36 -0500 Subject: [Rubygems-developers] [rubygems] Build broken by drbrain (#1109) In-Reply-To: <45712d19ea600_be419cdaf011a@ubuntu.tmail> References: <45712d19ea600_be419cdaf011a@ubuntu.tmail> Message-ID: Ok, all the tests are passing for me at the moment. The standard open-uri handles proxy specification in a different way than the patched version we were using. Fortunately, the new way is a bit easier. So, everybody, check out svn head and give it a try. Let me know if there are any tests that are still broken for you. Also, try it in your real environment and make user proxies (especially authenticated proxies) are still working. Thanks. -- -- -- Jim Weirich jim at weirichhouse.org http://onestepback.org ----------------------------------------------------------------- "Beware of bugs in the above code; I have only proved it correct, not tried it." -- Donald Knuth (in a memo to Peter van Emde Boas) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20061202/5fd40cc9/attachment.html From anatol.pomozov at gmail.com Sat Dec 2 16:02:05 2006 From: anatol.pomozov at gmail.com (Cerberus) Date: Sun, 3 Dec 2006 00:02:05 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1111) Message-ID: <4571e9cde2ce8_d3419b780e18d@ubuntu.tmail> ------------------------------------------------------------------------ r1111 | jimweirich | 2006-12-02 23:53:47 +0300 | 3 lines Changed paths: M /trunk/ChangeLog M /trunk/lib/rubygems/loadpath_manager.rb M /trunk/lib/rubygems.rb Removed duplicate copies of the SUFFIX_PATTERN by providing a global method Gem.suffixes and Gem.suffix_pattern. ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/rake/rake_test_loader Started .........................................................................................................F....................................................................................................................................................... Finished in 21.5 seconds. 1) Failure: test_cache_data_none_readable(TestGemSourceInfoCache) [./test/test_gem_source_info_cache.rb:82]: <{}> expected but was <{"key"=>"sys"}>. 257 tests, 947 assertions, 1 failures, 0 errors rake aborted! Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib "c:/PRO...] (See full trace by running task with --trace) -- Cerberus 0.3.2, http://cerberus.rubyforge.org/ From anatol.pomozov at gmail.com Sat Dec 2 16:36:44 2006 From: anatol.pomozov at gmail.com (Anatol Pomozov) Date: Sun, 3 Dec 2006 00:36:44 +0300 Subject: [Rubygems-developers] [rubygems] Build broken by drbrain (#1109) In-Reply-To: <3E403BF4-3714-4466-977D-3B956A5A82DD@segment7.net> References: <45712d19ea600_be419cdaf011a@ubuntu.tmail> <3E403BF4-3714-4466-977D-3B956A5A82DD@segment7.net> Message-ID: <3665a1a00612021336s12819bd3i359b5be88439b49e@mail.gmail.com> On 12/2/06, Eric Hodel wrote: > > On Dec 1, 2006, at 23:36 , Cerberus wrote: > > > ---------------------------------------------------------------------- > > -- > > r1109 | drbrain | 2006-12-01 03:44:41 +0300 | 1 line > > Changed paths: > > M /trunk/ChangeLog > > M /trunk/lib/rubygems/specification.rb > > M /trunk/test/test_specification.rb > > > > Allow gemspecs to be used intelligently with Array#-, Array#uniq > > ---------------------------------------------------------------------- > > -- > > > > (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) > > Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/ > > rake/rake_test_loader > > Started > > ...................................................................... > > ...................................F.................................. > > .........EEE.E.E...................................................... > > ............................................... > > Finished in 14.891 seconds. > > > > 1) Failure: > > test_cache_data_none_readable(TestGemSourceInfoCache) [./test/ > > test_gem_source_info_cache.rb:82]: > > <{}> expected but was > > <{"key"=>"sys"}>. > > Possibly... Looks like code FileUtils.chmod 0222, @sic.system_cache_file does not work for Windows. (Current version of Cerberus runs on my WinXP home computer) > 2) Error: > > test_explicit_proxy(TestRemoteFetcher): > > ArgumentError: unrecognized option: proxy_http_basic_authentication > > 3) Error: > > test_explicit_proxy_with_user_auth(TestRemoteFetcher): > > ArgumentError: unrecognized option: proxy_http_basic_authentication > > > > 4) Error: > > test_explicit_proxy_with_user_auth_in_env(TestRemoteFetcher): > > ArgumentError: unrecognized option: proxy_http_basic_authentication > > > > 5) Error: > > test_implicit_proxy(TestRemoteFetcher): > > ArgumentError: unrecognized option: proxy_http_basic_authentication > > > > 6) Error: > > test_implicit_upper_case_proxy(TestRemoteFetcher): > > ArgumentError: unrecognized option: proxy_http_basic_authentication > > Lies! These were caused by the removal of rubygems/open-uri... Well, yes. Errors were caused by open-uri. But Cerberus shows only last commit message. If Cerberus (WinXP OS) was running whole the time then we have gotten message right after open-uri removal. But last week I worked under Kubuntu OS so there was no possibility for Cerberus to check RubyGems tests. -- anatol (http://feeds.feedburner.com/apomozov-eng) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20061203/bb7a40f5/attachment.html From jim.weirich at gmail.com Sat Dec 2 17:05:44 2006 From: jim.weirich at gmail.com (Jim Weirich) Date: Sat, 2 Dec 2006 17:05:44 -0500 Subject: [Rubygems-developers] [rubygems] Build still broken (#1111) In-Reply-To: <4571e9cde2ce8_d3419b780e18d@ubuntu.tmail> References: <4571e9cde2ce8_d3419b780e18d@ubuntu.tmail> Message-ID: On 12/2/06, Cerberus wrote: 1) Failure: > test_cache_data_none_readable(TestGemSourceInfoCache) > [./test/test_gem_source_info_cache.rb:82]: > <{}> expected but was > <{"key"=>"sys"}>. > > 257 tests, 947 assertions, 1 failures, 0 errors > rake aborted! > Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib > "c:/PRO...] It is working for me. Perhaps we have a test with external dependencies. What system/ruby version are you using? -- -- -- Jim Weirich jim at weirichhouse.org http://onestepback.org ----------------------------------------------------------------- "Beware of bugs in the above code; I have only proved it correct, not tried it." -- Donald Knuth (in a memo to Peter van Emde Boas) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20061202/2b9446d0/attachment.html From anatol.pomozov at gmail.com Sun Dec 3 02:51:30 2006 From: anatol.pomozov at gmail.com (Anatol Pomozov) Date: Sun, 3 Dec 2006 10:51:30 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1111) In-Reply-To: References: <4571e9cde2ce8_d3419b780e18d@ubuntu.tmail> Message-ID: <3665a1a00612022351m20a3bcaw50014933965df686@mail.gmail.com> Hi, Jim. I believe that it is Windows specific problem. Look at test code at test_gem_source_info_cache.rb:80 . Chmod and stat functions are very limited on Windows platform. I have added code that shows us what permission file has after we try to set 0222 mode FileUtils.chmod 0222, @sic.system_cache_file puts sprintf("%o", File.stat(@sic.system_cache_file).mode) It returns me 100644 It means that "real" file mode is 644. So RubyGems engine could read cache data. Some more explanation you could find here http://blade.nagaokaut.ac.jp/cgi-bin/scat.rb/ruby/ruby-core/1149 Looks like this test is *nix specific and could be run only on platforms that fully support FIle.mode. On 12/3/06, Jim Weirich wrote: > > On 12/2/06, Cerberus wrote: 1) Failure: > > > test_cache_data_none_readable(TestGemSourceInfoCache) > > [./test/test_gem_source_info_cache.rb:82]: > > <{}> expected but was > > <{"key"=>"sys"}>. > > > > 257 tests, 947 assertions, 1 failures, 0 errors > > rake aborted! > > Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib > > "c:/PRO...] > > > It is working for me. Perhaps we have a test with external dependencies. > What system/ruby version are you using? > > -- > -- > -- Jim Weirich jim at weirichhouse.org http://onestepback.org > -- anatol (http://feeds.feedburner.com/apomozov-eng) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20061203/a9b70a83/attachment.html From anatol.pomozov at gmail.com Mon Dec 4 00:41:23 2006 From: anatol.pomozov at gmail.com (Cerberus) Date: Mon, 4 Dec 2006 08:41:23 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1112) Message-ID: <4573b50340740_68019b787c170@ubuntu.tmail> ------------------------------------------------------------------------ r1112 | drbrain | 2006-12-04 08:26:09 +0300 | 3 lines Changed paths: M /trunk/ChangeLog M /trunk/lib/rubygems/remote_installer.rb Don't download cached files. Allow cache dir to be overridden. ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/rake/rake_test_loader Started .........................................................................................................F....................................................................................................................................................... Finished in 13.328 seconds. 1) Failure: test_cache_data_none_readable(TestGemSourceInfoCache) [./test/test_gem_source_info_cache.rb:82]: <{}> expected but was <{"key"=>"sys"}>. 257 tests, 947 assertions, 1 failures, 0 errors rake aborted! Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib "c:/PRO...] (See full trace by running task with --trace) -- Cerberus 0.3.2, http://cerberus.rubyforge.org/ From drbrain at segment7.net Mon Dec 4 13:36:46 2006 From: drbrain at segment7.net (Eric Hodel) Date: Mon, 4 Dec 2006 10:36:46 -0800 Subject: [Rubygems-developers] [rubygems] Build broken by drbrain (#1109) In-Reply-To: References: <45712d19ea600_be419cdaf011a@ubuntu.tmail> Message-ID: <545238ED-9C0E-43E8-B43C-EBE88048AC56@segment7.net> On Dec 2, 2006, at 12:57 , Jim Weirich wrote: > Ok, all the tests are passing for me at the moment. The standard > open-uri handles proxy specification in a different way than the > patched version we were using. Fortunately, the new way is a bit > easier. > > So, everybody, check out svn head and give it a try. Let me know > if there are any tests that are still broken for you. Also, try it > in your real environment and make user proxies (especially > authenticated proxies) are still working. I've installed more gems than anybody else, ever, using trunk, and it is working fine. -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From drbrain at segment7.net Mon Dec 4 13:41:56 2006 From: drbrain at segment7.net (Eric Hodel) Date: Mon, 4 Dec 2006 10:41:56 -0800 Subject: [Rubygems-developers] [rubygems] Build still broken (#1111) In-Reply-To: <3665a1a00612022351m20a3bcaw50014933965df686@mail.gmail.com> References: <4571e9cde2ce8_d3419b780e18d@ubuntu.tmail> <3665a1a00612022351m20a3bcaw50014933965df686@mail.gmail.com> Message-ID: <7BCED878-ED5D-40A4-B8B7-98F59050DDC1@segment7.net> On Dec 2, 2006, at 23:51 , Anatol Pomozov wrote: > Look at test code at test_gem_source_info_cache.rb:80 . Chmod and > stat functions are very limited on Windows platform. > > I have added code that shows us what permission file has after we > try to set 0222 mode > FileUtils.chmod 0222, @sic.system_cache_file > puts sprintf("%o", File.stat(@sic.system_cache_file).mode) > > It returns me > 100644 > > It means that "real" file mode is 644. So RubyGems engine could > read cache data. How about this patch: Index: test/test_gem_source_info_cache.rb =================================================================== --- test/test_gem_source_info_cache.rb (revision 1111) +++ test/test_gem_source_info_cache.rb (working copy) @@ -79,6 +79,8 @@ def test_cache_data_none_readable FileUtils.chmod 0222, @sic.system_cache_file FileUtils.chmod 0222, @sic.user_cache_file + return if (File.stat(@sic.system_cache_file).mode & 0222) != 0222 + return if (File.stat(@sic.user_cache_file).mode & 0222) != 0222 assert_equal({}, @sic.cache_data) end -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From anatol.pomozov at gmail.com Mon Dec 4 16:05:24 2006 From: anatol.pomozov at gmail.com (Anatol Pomozov) Date: Tue, 5 Dec 2006 00:05:24 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1111) In-Reply-To: <7BCED878-ED5D-40A4-B8B7-98F59050DDC1@segment7.net> References: <4571e9cde2ce8_d3419b780e18d@ubuntu.tmail> <3665a1a00612022351m20a3bcaw50014933965df686@mail.gmail.com> <7BCED878-ED5D-40A4-B8B7-98F59050DDC1@segment7.net> Message-ID: <3665a1a00612041305q8518a6dr9d1b2bfb9070fb64@mail.gmail.com> Hi, Eric. Your solution works fine on Windows. Just add comments to these lines what those returns are mean. On 12/4/06, Eric Hodel wrote: > > On Dec 2, 2006, at 23:51 , Anatol Pomozov wrote: > > > Look at test code at test_gem_source_info_cache.rb:80 . Chmod and > > stat functions are very limited on Windows platform. > > > > I have added code that shows us what permission file has after we > > try to set 0222 mode > > FileUtils.chmod 0222, @sic.system_cache_file > > puts sprintf("%o", File.stat(@sic.system_cache_file).mode) > > > > It returns me > > 100644 > > > > It means that "real" file mode is 644. So RubyGems engine could > > read cache data. > > How about this patch: > > Index: test/test_gem_source_info_cache.rb > =================================================================== > --- test/test_gem_source_info_cache.rb (revision 1111) > +++ test/test_gem_source_info_cache.rb (working copy) > @@ -79,6 +79,8 @@ > def test_cache_data_none_readable > FileUtils.chmod 0222, @sic.system_cache_file > FileUtils.chmod 0222, @sic.user_cache_file > + return if (File.stat(@sic.system_cache_file).mode & 0222) != 0222 > + return if (File.stat(@sic.user_cache_file).mode & 0222) != 0222 > assert_equal({}, @sic.cache_data) > end > > -- > Eric Hodel - drbrain at segment7.net - http://blog.segment7.net > -- anatol (http://feeds.feedburner.com/apomozov-eng) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20061205/50462587/attachment.html From drbrain at segment7.net Tue Dec 5 03:42:40 2006 From: drbrain at segment7.net (Eric Hodel) Date: Tue, 5 Dec 2006 00:42:40 -0800 Subject: [Rubygems-developers] [rubygems] Build still broken (#1111) In-Reply-To: <3665a1a00612041305q8518a6dr9d1b2bfb9070fb64@mail.gmail.com> References: <4571e9cde2ce8_d3419b780e18d@ubuntu.tmail> <3665a1a00612022351m20a3bcaw50014933965df686@mail.gmail.com> <7BCED878-ED5D-40A4-B8B7-98F59050DDC1@segment7.net> <3665a1a00612041305q8518a6dr9d1b2bfb9070fb64@mail.gmail.com> Message-ID: <14C0E91D-5444-44DC-86EF-BB766640DA1F@segment7.net> On Dec 4, 2006, at 13:05 , Anatol Pomozov wrote: > Hi, Eric. > > Your solution works fine on Windows. Committed. -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From anatol.pomozov at gmail.com Tue Dec 5 18:19:38 2006 From: anatol.pomozov at gmail.com (Anatol Pomozov) Date: Wed, 6 Dec 2006 02:19:38 +0300 Subject: [Rubygems-developers] [rubygems] Build fixed by drbrain (#1113) Message-ID: <3665a1a00612051519n7931448g25445977b73d411e@mail.gmail.com> ------------------------------------------------------------------------ r1113 | drbrain | 2006-12-05 11:38:57 +0300 | 1 line Hack around systems that don't support chmod ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/rake/rake_test_loader Started ................................................................................................................................................................................................................................................................. Finished in 22.469 seconds. 257 tests, 946 assertions, 0 failures, 0 errors -- Cerberus 0.3.2, http://cerberus.rubyforge.org/ From drbrain at segment7.net Tue Dec 5 20:18:36 2006 From: drbrain at segment7.net (Eric Hodel) Date: Tue, 5 Dec 2006 17:18:36 -0800 Subject: [Rubygems-developers] [rubygems] Build fixed by drbrain (#1113) In-Reply-To: <3665a1a00612051519n7931448g25445977b73d411e@mail.gmail.com> References: <3665a1a00612051519n7931448g25445977b73d411e@mail.gmail.com> Message-ID: <26D38B05-E0B0-4AF1-A1C8-6E17020A0F56@segment7.net> On Dec 5, 2006, at 15:19 , Anatol Pomozov wrote: > ---------------------------------------------------------------------- > -- > r1113 | drbrain | 2006-12-05 11:38:57 +0300 | 1 line > > Hack around systems that don't support chmod > ---------------------------------------------------------------------- > -- > > (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) > Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/ > rake/rake_test_loader > Started > ...................................................................... > ...................................................................... > ...................................................................... > ............................................... > Finished in 22.469 seconds. > > 257 tests, 946 assertions, 0 failures, 0 errors w00t! -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From drbrain at segment7.net Fri Dec 8 03:04:54 2006 From: drbrain at segment7.net (Eric Hodel) Date: Fri, 8 Dec 2006 00:04:54 -0800 Subject: [Rubygems-developers] Gem::Platform Message-ID: Currently we have: module Platform RUBY = 'ruby' WIN32 = 'mswin32' LINUX_586 = 'i586-linux' DARWIN = 'powerpc-darwin' CURRENT = 'current' end And here's what today's source index has: require 'rubygems/source_info_cache' gems = Gem::SourceInfoCache.cache_data['http:// gems.rubyforge.org'].source_index.latest_specs counts = Hash.new { |h,k| h[k] = 0 } gems.each { |_,g| counts[g.platform] += 1 } puts counts.sort_by { |pl,c| c }.map { |pl,c| "\t#{c}\t#{pl}" }.join ("\n") 1 powerpc-darwin 1 i686-darwin8.4.1 1 windows 1 win32-1.8.2-VC7 3 i686-linux 3 i586-linux 5 i386-mswin32 22 mswin32 29 1108 ruby mswin32 has been outrageously successful, but linux and darwin less-so. I think we'll need a DARWIN_INTEL and DARWIN_UNIVERSAL, a more- generic linux, and ...? -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From tomp at earthlink.net Fri Dec 8 10:33:41 2006 From: tomp at earthlink.net (Tom Pollard) Date: Fri, 8 Dec 2006 10:33:41 -0500 Subject: [Rubygems-developers] New patch for 'gem contents' bug Message-ID: <22C65D5F-406C-4007-A192-08AB67716417@earthlink.net> Hi, In rubygems 0.9.0, the 'gem contents' command is broken, in that it can no longer find the installed gems. It looks like the code that used to fall back on the installed gem path when no directories were specified on the command line was replaced. This has been broken for months now, and is still broken in the current code on svn. I notice that there's a simple patch for this in the rubyforge bugtracker. I have a slightly more complete patch I'd like to offer. Unfortunately, I don't see any way to upload a new patch file to an existing case in rubyforge. What's the right thing to do here? Thanks, TomP From chad at chadfowler.com Fri Dec 8 11:10:29 2006 From: chad at chadfowler.com (Chad Fowler) Date: Fri, 8 Dec 2006 09:10:29 -0700 Subject: [Rubygems-developers] Gem::Platform In-Reply-To: References: Message-ID: On 12/8/06, Eric Hodel wrote: > Currently we have: > > module Platform > RUBY = 'ruby' > WIN32 = 'mswin32' > LINUX_586 = 'i586-linux' > DARWIN = 'powerpc-darwin' > CURRENT = 'current' > end > > And here's what today's source index has: > > require 'rubygems/source_info_cache' > gems = Gem::SourceInfoCache.cache_data['http:// > gems.rubyforge.org'].source_index.latest_specs > > counts = Hash.new { |h,k| h[k] = 0 } > gems.each { |_,g| counts[g.platform] += 1 } > puts counts.sort_by { |pl,c| c }.map { |pl,c| "\t#{c}\t#{pl}" }.join > ("\n") > > 1 powerpc-darwin > 1 i686-darwin8.4.1 > 1 windows > 1 win32-1.8.2-VC7 > 3 i686-linux > 3 i586-linux > 5 i386-mswin32 > 22 mswin32 > 29 > 1108 ruby > > mswin32 has been outrageously successful, but linux and darwin less-so. > > I think we'll need a DARWIN_INTEL and DARWIN_UNIVERSAL, a more- > generic linux, and ...? > Hi. This is exactly why we created tattle (http://rubyforge.org/projects/tattle). I just released a first version of the gem so people could try it out. Bruce is going to make some UI changes to the report (http://tattle.rubygarden.org) hopefully today. We also forgot to support HTTP proxies, so we'll need to fix that. Give it a go by doing: sudo gem install tattle tattle Chad From grant at antiflux.org Fri Dec 8 11:10:53 2006 From: grant at antiflux.org (Grant Hollingworth) Date: Fri, 8 Dec 2006 09:10:53 -0700 Subject: [Rubygems-developers] Gem::Platform In-Reply-To: References: Message-ID: <20061208161053.GB19630@antiflux.org> * Eric Hodel [2006-12-08 01:05]: >mswin32 has been outrageously successful, but linux and darwin less-so. Linux and Darwin don't have so many special cases. Or at least they're far more likely to have a compiler. >I think we'll need a DARWIN_INTEL and DARWIN_UNIVERSAL, a more- >generic linux, and ...? Separating architecture and OS might be a good idea. And perhaps a distinction between source and binary packages? What does a platform of i586-linux mean, exactly? That it's compiled already? That it contains assembly that only works on i586? Special compiler flags? From grant at antiflux.org Fri Dec 8 11:01:33 2006 From: grant at antiflux.org (Grant Hollingworth) Date: Fri, 8 Dec 2006 09:01:33 -0700 Subject: [Rubygems-developers] New patch for 'gem contents' bug In-Reply-To: <22C65D5F-406C-4007-A192-08AB67716417@earthlink.net> References: <22C65D5F-406C-4007-A192-08AB67716417@earthlink.net> Message-ID: <20061208160133.GA19630@antiflux.org> * Tom Pollard [2006-12-08 08:45]: >I notice that there's a simple patch for this in the rubyforge >bugtracker. I have a slightly more complete patch I'd like to >offer. Unfortunately, I don't see any way to upload a new patch file >to an existing case in rubyforge. What's the right thing to do here? I'd love to see this fixed, too. I supplied a patch back in June that's essentially the same as the one on Rubyforge. I think maybe it didn't work on gems with multiple versions, though. A more complete patch sounds excellent. From drbrain at segment7.net Fri Dec 8 14:04:10 2006 From: drbrain at segment7.net (Eric Hodel) Date: Fri, 8 Dec 2006 11:04:10 -0800 Subject: [Rubygems-developers] Gem::Platform In-Reply-To: <20061208161053.GB19630@antiflux.org> References: <20061208161053.GB19630@antiflux.org> Message-ID: <17A61DDF-8F61-4FB0-B377-BB11FE85B931@segment7.net> On Dec 8, 2006, at 08:10 , Grant Hollingworth wrote: > * Eric Hodel [2006-12-08 01:05]: >> mswin32 has been outrageously successful, but linux and darwin >> less-so. > > Linux and Darwin don't have so many special cases. Or at least > they're far more likely to have a compiler. But not 100% likely. Tattle will tell us more. >> I think we'll need a DARWIN_INTEL and DARWIN_UNIVERSAL, a more- >> generic linux, and ...? > > Separating architecture and OS might be a good idea. And perhaps a > distinction between source and binary packages? If you're going to make this split you'll need arch, platform and compiler (for win32). And it isn't backwards-compatible with the existing gems. > What does a platform of i586-linux mean, exactly? That it's > compiled already? That it contains assembly that only works on > i586? Special compiler flags? Just "compiled for linux x86". -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From drbrain at segment7.net Fri Dec 8 14:04:58 2006 From: drbrain at segment7.net (Eric Hodel) Date: Fri, 8 Dec 2006 11:04:58 -0800 Subject: [Rubygems-developers] New patch for 'gem contents' bug In-Reply-To: <22C65D5F-406C-4007-A192-08AB67716417@earthlink.net> References: <22C65D5F-406C-4007-A192-08AB67716417@earthlink.net> Message-ID: On Dec 8, 2006, at 07:33 , Tom Pollard wrote: > In rubygems 0.9.0, the 'gem contents' command is broken, in that it > can no longer find the installed gems. It looks like the code that > used to fall back on the installed gem path when no directories were > specified on the command line was replaced. This has been broken for > months now, and is still broken in the current code on svn. > > I notice that there's a simple patch for this in the rubyforge > bugtracker. I have a slightly more complete patch I'd like to > offer. Unfortunately, I don't see any way to upload a new patch file > to an existing case in rubyforge. What's the right thing to do here? If you can't submit it to rubyforge, post it here. -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From tomp at earthlink.net Fri Dec 8 23:18:14 2006 From: tomp at earthlink.net (Tom Pollard) Date: Fri, 8 Dec 2006 23:18:14 -0500 Subject: [Rubygems-developers] New patch for 'gem contents' bug In-Reply-To: <20061208160133.GA19630@antiflux.org> References: <22C65D5F-406C-4007-A192-08AB67716417@earthlink.net> <20061208160133.GA19630@antiflux.org> Message-ID: <9A43CA64-8F35-468F-BC20-BF08B133595E@earthlink.net> On Dec 8, 2006, at 11:01 AM, Grant Hollingworth wrote: > I supplied a patch back in June that's essentially the same as the > one on Rubyforge. I think maybe it didn't work on gems with > multiple versions, though. The --version option not working for 'gem contents' is a separate bug, which I hadn't noticed. It turns out it's just an typo in the configuration of that option in ContentsCommand. ( "--version" should be "--version VERSION" in the add_option() call.) Another thing that looks to me like a bug is that, when there are multiple versions, the earliest version is selected by default, rather than the most current. TomP From tomp at earthlink.net Sat Dec 9 00:01:02 2006 From: tomp at earthlink.net (Tom Pollard) Date: Sat, 9 Dec 2006 00:01:02 -0500 Subject: [Rubygems-developers] New patch for 'gem contents' bug In-Reply-To: References: <22C65D5F-406C-4007-A192-08AB67716417@earthlink.net> Message-ID: Hi, I've attached a patch file (a context diff) for gem_commands.rb, made against the current source file in svn. The changes are all in the ContentsCommand class. This fixes three bugs: 1) 'gem contents' wouldn't work unless you explicitly specified a specification directory, 2) the '--version' option didn't work at all, because the version number you specified was treated as a gem name, and 3) when multiple versions of a gem were installed, the earliest was selected by default, rather than the most current. (Now the most recent is selected.) The fixes for (2) and (3) were each a one-word change on a single line. The code for the ContentsCommand#execute function could also be simplified a little, because there's always a list of specification directories now. My guess is that the '--version' bug arose because someone didn't want to use the VersionOption mixin in this class. I think that's probably because the descriptive text used in the option definition in VersionOption ("Specify version of gem to #{taskname}") assumes the command name is a transitive verb, like 'install', or 'list'. If the string were changed to something like "Specify version of gem for #{taskname}"), it might be appropriate for a command like 'contents', too, even if it doesn't sound as good for 'install'. Maybe you could simplify it further, to just "Specify version of gem"? In an any case, it would be nice if the VersionOption mixin could be used here. Thanks, TomP On Dec 8, 2006, at 2:04 PM, Eric Hodel wrote: >> I don't see any way to upload a new patch file to an existing >> case in rubyforge. What's the right thing to do here? > > If you can't submit it to rubyforge, post it here. -------------- next part -------------- A non-text attachment was scrubbed... Name: gem_commands.rb.patch Type: application/octet-stream Size: 1915 bytes Desc: not available Url : http://rubyforge.org/pipermail/rubygems-developers/attachments/20061209/7cc31f5f/attachment.obj -------------- next part -------------- From tomp at earthlink.net Sat Dec 9 17:33:55 2006 From: tomp at earthlink.net (Tom Pollard) Date: Sat, 9 Dec 2006 17:33:55 -0500 Subject: [Rubygems-developers] New patch for 'gem contents' bug In-Reply-To: References: <22C65D5F-406C-4007-A192-08AB67716417@earthlink.net> Message-ID: On Dec 9, 2006, at 12:01 AM, Tom Pollard wrote: > 2) the '--version' option didn't work at all, because the version > number you specified was treated as a gem name, I notice now that this was already reported by Scott Sayles (with a patch) more than a year ago as bug #2723. https://rubyforge.org/tracker/index.php? func=detail&aid=2723&group_id=126&atid=575 TomP From ciaranm at ciaranm.org Mon Dec 11 13:57:58 2006 From: ciaranm at ciaranm.org (Ciaran McCreesh) Date: Mon, 11 Dec 2006 18:57:58 +0000 Subject: [Rubygems-developers] yaml apparently invalid Message-ID: <20061211185758.36eca471@snowdrop> The master yaml file at [1] appears to be invalid. Specifically, the entry for extra-1.0 has a description: key that starts with a backtick (currently this is line 234467). According to the yaml specification [2], around Example 4.10, "The ?@? and ?`? are reserved for future use." Although some yaml parsers currently don't have a problem with the backtick, libyaml treats it as a fatal error. Who would be the appropriate party to contact to get this fixed? Is it a problem with the gem itself, or is it caused by insufficient escaping being carried out by the library? Thanks, [1]: http://gems.rubyforge.org/yaml [2]: http://yaml.org/spec/current.html -- Ciaran McCreesh Mail : ciaranm at ciaranm.org Web : http://ciaranm.org/ -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: not available Url : http://rubyforge.org/pipermail/rubygems-developers/attachments/20061211/0fb45bb2/attachment.bin From drbrain at segment7.net Wed Dec 13 18:35:26 2006 From: drbrain at segment7.net (Eric Hodel) Date: Wed, 13 Dec 2006 18:35:26 -0500 Subject: [Rubygems-developers] Where does the quick index go? Message-ID: <9BD5514E-40AF-4269-85B0-223E7B5DDAE0@segment7.net> Sometimes rubygems likes to do a bulk update even when there's no possible way 50 gems could have been updated. I added some code and found that occasionally the quick index just isn't there: $ sudo gem install image_science Fetching quick Gem source index for: http://gems.rubyforge.org Quick index not found: 404 Not Found Bulk updating Gem source index for: http://gems.rubyforge.org [...] -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From jim.weirich at gmail.com Thu Dec 14 01:14:05 2006 From: jim.weirich at gmail.com (Jim Weirich) Date: Thu, 14 Dec 2006 01:14:05 -0500 Subject: [Rubygems-developers] Where does the quick index go? In-Reply-To: <9BD5514E-40AF-4269-85B0-223E7B5DDAE0@segment7.net> References: <9BD5514E-40AF-4269-85B0-223E7B5DDAE0@segment7.net> Message-ID: On 12/13/06, Eric Hodel wrote: > Sometimes rubygems likes to do a bulk update even when there's no > possible way 50 gems could have been updated. I added some code and > found that occasionally the quick index just isn't there: > > $ sudo gem install image_science > Fetching quick Gem source index for: http://gems.rubyforge.org > Quick index not found: 404 Not Found > Bulk updating Gem source index for: http://gems.rubyforge.org > [...] Is this a mirror issue? -- -- -- Jim Weirich jim at weirichhouse.org http://onestepback.org ----------------------------------------------------------------- "Beware of bugs in the above code; I have only proved it correct, not tried it." -- Donald Knuth (in a memo to Peter van Emde Boas) From drbrain at segment7.net Thu Dec 14 10:42:39 2006 From: drbrain at segment7.net (Eric Hodel) Date: Thu, 14 Dec 2006 10:42:39 -0500 Subject: [Rubygems-developers] Where does the quick index go? In-Reply-To: References: <9BD5514E-40AF-4269-85B0-223E7B5DDAE0@segment7.net> Message-ID: <82DBD792-241E-4C0A-920D-4153791F36AB@segment7.net> On Dec 14, 2006, at 01:14, Jim Weirich wrote: > On 12/13/06, Eric Hodel wrote: >> Sometimes rubygems likes to do a bulk update even when there's no >> possible way 50 gems could have been updated. I added some code and >> found that occasionally the quick index just isn't there: >> >> $ sudo gem install image_science >> Fetching quick Gem source index for: http://gems.rubyforge.org >> Quick index not found: 404 Not Found >> Bulk updating Gem source index for: http://gems.rubyforge.org >> [...] > > Is this a mirror issue? I'm not sure how I can easily tell. OpenURI/Net::HTTP are hard to dig into without getting tons of crap you don't want back. -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From hgs at dmu.ac.uk Thu Dec 14 11:11:27 2006 From: hgs at dmu.ac.uk (Hugh Sasse) Date: Thu, 14 Dec 2006 16:11:27 +0000 (WET) Subject: [Rubygems-developers] Betagems not available? Message-ID: I've been trying to update rubygems and I'm getting bash-2.05# gem update --system --source http://onestepback.org/betagems Updating RubyGems... ERROR: While executing gem ... (Gem::RemoteSourceException) HTTP Response 503 bash-2.05# I can't remember what that code means now. Oh, seems to be service unavailable... Hugh From drbrain at segment7.net Thu Dec 14 14:02:52 2006 From: drbrain at segment7.net (Eric Hodel) Date: Thu, 14 Dec 2006 14:02:52 -0500 Subject: [Rubygems-developers] Betagems not available? In-Reply-To: References: Message-ID: <97A961EF-4D7A-4E0D-97EB-2E0AB31AD8B5@segment7.net> On Dec 14, 2006, at 11:11, Hugh Sasse wrote: > I've been trying to update rubygems and I'm getting > > bash-2.05# gem update --system --source http://onestepback.org/ > betagems > Updating RubyGems... > ERROR: While executing gem ... (Gem::RemoteSourceException) > HTTP Response 503 > bash-2.05# > > I can't remember what that code means now. Oh, seems to be service > unavailable... Yes, rubyforge is having issue. -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From bspolarich at picometrix.com Thu Dec 14 15:14:29 2006 From: bspolarich at picometrix.com (Brian W. Spolarich ) Date: Thu, 14 Dec 2006 15:14:29 -0500 Subject: [Rubygems-developers] Apologies In Advance - How to Unsubscribe? Message-ID: Apologies in advance for sending this to the entire list, but I've tried the other usual routes to no avail (rubygems-developers-request at rubyforge.org, -owner, mailman commands, etc.) I'd like to unsubscribe to rubygems-developers. Mailman seems unwilling to do so, although I've tried a couple of times and have responded to the eventual auto-reply. Can someone help? Thanks, -bws From drbrain at segment7.net Fri Dec 15 02:21:08 2006 From: drbrain at segment7.net (Eric Hodel) Date: Thu, 14 Dec 2006 23:21:08 -0800 Subject: [Rubygems-developers] Trackker Message-ID: <13961314-7D6C-4844-9614-F14C5888541B@segment7.net> I've got the tracker down to 37 open items: http://rubyforge.org/tracker/?group_id=126 The rest of the items are patches or bugs that require actual work. A bit of help would be appreciated. There are three gem submissions that are so stale they should probably be closed. -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From jim.weirich at gmail.com Fri Dec 15 07:25:42 2006 From: jim.weirich at gmail.com (Jim Weirich) Date: Fri, 15 Dec 2006 07:25:42 -0500 Subject: [Rubygems-developers] Betagems not available? In-Reply-To: References: Message-ID: On 12/14/06, Hugh Sasse wrote: > I've been trying to update rubygems and I'm getting > > bash-2.05# gem update --system --source http://onestepback.org/betagems > Updating RubyGems... > ERROR: While executing gem ... (Gem::RemoteSourceException) > HTTP Response 503 > bash-2.05# > > I can't remember what that code means now. Oh, seems to be service > unavailable... My UML instance is down. So that means my blog, betagems, rake doc site and primary email are kaput. Blah. Hopefully it will be back soon. -- -- -- Jim Weirich jim at weirichhouse.org http://onestepback.org ----------------------------------------------------------------- "Beware of bugs in the above code; I have only proved it correct, not tried it." -- Donald Knuth (in a memo to Peter van Emde Boas) From chad at chadfowler.com Sat Dec 16 11:54:08 2006 From: chad at chadfowler.com (Chad Fowler) Date: Sat, 16 Dec 2006 09:54:08 -0700 Subject: [Rubygems-developers] Trackker In-Reply-To: <13961314-7D6C-4844-9614-F14C5888541B@segment7.net> References: <13961314-7D6C-4844-9614-F14C5888541B@segment7.net> Message-ID: On 12/15/06, Eric Hodel wrote: > I've got the tracker down to 37 open items: > > http://rubyforge.org/tracker/?group_id=126 > > The rest of the items are patches or bugs that require actual work. > A bit of help would be appreciated. > > There are three gem submissions that are so stale they should > probably be closed. > Eric, are you aware of fixing http://rubyforge.org/tracker/index.php?func=detail&aid=982&group_id=126&atid=575. It looks like you did. I added a test i expected to fail and it passed. Looks like svn has the desired behavior now. Chad From drbrain at segment7.net Mon Dec 18 17:18:05 2006 From: drbrain at segment7.net (Eric Hodel) Date: Mon, 18 Dec 2006 14:18:05 -0800 Subject: [Rubygems-developers] Trackker In-Reply-To: References: <13961314-7D6C-4844-9614-F14C5888541B@segment7.net> Message-ID: <64493C79-7AF2-4CF9-B393-CD61CABFC4FB@segment7.net> On Dec 16, 2006, at 08:54, Chad Fowler wrote: > On 12/15/06, Eric Hodel wrote: >> I've got the tracker down to 37 open items: >> >> http://rubyforge.org/tracker/?group_id=126 >> >> The rest of the items are patches or bugs that require actual work. >> A bit of help would be appreciated. >> >> There are three gem submissions that are so stale they should >> probably be closed. >> > > > Eric, are you aware of fixing > http://rubyforge.org/tracker/index.php? > func=detail&aid=982&group_id=126&atid=575. > It looks like you did. I added a test i expected to fail and it > passed. Looks like svn has the desired behavior now. I don't remember fixing it. If it is fixed go ahead and close the bug. -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From drbrain at segment7.net Tue Dec 19 02:13:17 2006 From: drbrain at segment7.net (Eric Hodel) Date: Mon, 18 Dec 2006 23:13:17 -0800 Subject: [Rubygems-developers] Bad extconf.rb (Was: Trackker) In-Reply-To: <64493C79-7AF2-4CF9-B393-CD61CABFC4FB@segment7.net> References: <13961314-7D6C-4844-9614-F14C5888541B@segment7.net> <64493C79-7AF2-4CF9-B393-CD61CABFC4FB@segment7.net> Message-ID: <64798902-3E2A-4C7B-B962-3471B1449469@segment7.net> On Dec 18, 2006, at 14:18, Eric Hodel wrote: > On Dec 16, 2006, at 08:54, Chad Fowler wrote: >> Eric, are you aware of fixing >> http://rubyforge.org/tracker/index.php? >> func=detail&aid=982&group_id=126&atid=575. >> It looks like you did. I added a test i expected to fail and it >> passed. Looks like svn has the desired behavior now. > > I don't remember fixing it. If it is fixed go ahead and close the > bug. Actually, this isn't our bug. Gem authors like to write broken extconf.rbs where they don't check the output of #have_library or #have_header (or whatever the header one is) so they'll build a Makefile when they shouldn't. There's really nothing we can do about this. If the C code is bad or the extconf.rb doesn't write a Makefile then things are good. -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From anatol.pomozov at gmail.com Tue Dec 19 12:23:28 2006 From: anatol.pomozov at gmail.com (Cerberus) Date: Tue, 19 Dec 2006 20:23:28 +0300 Subject: [Rubygems-developers] [rubygems] Build broken by drbrain (#1144) Message-ID: <45882010f0b90_37019b63aa151@ubuntu.tmail> ------------------------------------------------------------------------ r1144 | drbrain | 2006-12-19 13:11:39 +0300 | 1 line Changed paths: M /trunk/ChangeLog M /trunk/lib/rubygems/custom_require.rb M /trunk/lib/rubygems/gem_commands.rb M /trunk/lib/rubygems/source_index.rb M /trunk/lib/rubygems.rb Use File.join on glob strings. Fixes bug #1096 by Chad Fowler. ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/rake/rake_test_loader Started ......................................................................................'sh' is not recognized as an internal or external command, operable program or batch file. E'sh' is not recognized as an internal or external command, operable program or batch file. F Microsoft (R) Program Maintenance Utility Version 7.10.3077 Copyright (C) Microsoft Corporation. All rights reserved. ok Microsoft (R) Program Maintenance Utility Version 7.10.3077 Copyright (C) Microsoft Corporation. All rights reserved. ok Fcreating Makefile Microsoft (R) Program Maintenance Utility Version 7.10.3077 Copyright (C) Microsoft Corporation. All rights reserved. Microsoft (R) Program Maintenance Utility Version 7.10.3077 Copyright (C) Microsoft Corporation. All rights reserved. Fchecking for main() in nonexistent.lib... no creating Makefile Microsoft (R) Program Maintenance Utility Version 7.10.3077 Copyright (C) Microsoft Corporation. All rights reserved. c:\progra~1\ruby\bin\ruby -e "puts 'EXPORTS', 'Init_foo'" > foo-i386-mswin32.def cl -nologo -MD -Zi -O2b2xg- -G6 -I. -Ic:/progra~1/ruby/lib/ruby/1.8/i386-mswin32 -Ic:/progra~1/ruby/lib/ruby/1.8/i386-mswin32 -I. -c -Tcfoo.c foo.c cl -nologo -LD -Fefoo.so foo.obj msvcrt-ruby18.lib oldnames.lib user32.lib advapi32.lib wsock32.lib -link -incremental:no -debug -opt:ref -opt:icf -dll -libpath:"c:/progra~1/ruby/lib" -def:foo-i386-mswin32.def -implib:foo-i386-mswin32.lib -pdb:foo-i386-mswin32.pdb Creating library foo-i386-mswin32.lib and object foo-i386-mswin32.exp foo.obj : error LNK2019: unresolved external symbol _nonexistent referenced in function _Init_foo foo.so : fatal error LNK1120: 1 unresolved externals NMAKE : fatal error U1077: 'cl' : return code '0x2' Stop. Echecking for main() in nonexistent.lib... no need libnonexistent *** extconf.rb failed *** Could not create Makefile due to some reason, probably lack of necessary libraries and/or headers. Check the mkmf.log file for more details. You may need configuration options. Provided configuration options: --with-opt-dir --without-opt-dir --with-opt-include --without-opt-include=${opt-dir}/include --with-opt-lib --without-opt-lib=${opt-dir}/lib --with-make-prog --srcdir=. --curdir --ruby=c:/progra~1/ruby/bin/ruby --with-nonexistentlib --without-nonexistentlib F Microsoft (R) Program Maintenance Utility Version 7.10.3077 Copyright (C) Microsoft Corporation. All rights reserved. Microsoft (R) Program Maintenance Utility Version 7.10.3077 Copyright (C) Microsoft Corporation. All rights reserved. F.(in C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3428/ext) F(in C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3428/ext) rake aborted! fail C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3428/ext/rakefile:1:in `abort' (See full trace by running task with --trace) fail F...............................................................................F.............................................................................................. Finished in 80.578 seconds. 1) Error: test_self_build(TestGemExtConfigureBuilder): Gem::InstallError: configure failed: sh ./configure --prefix=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3428/prefix ./lib/rubygems/installer.rb:600:in `run' ./lib/rubygems/installer.rb:611:in `build' ./test/test_gem_ext_configure_builder.rb:28:in `test_self_build' ./test/test_gem_ext_configure_builder.rb:27:in `test_self_build' 2) Failure: test_self_build_fail(TestGemExtConfigureBuilder) [./test/test_gem_ext_configure_builder.rb:54]: <"configure failed:\n\nsh ./configure --prefix=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3428/prefix\n./configure: ./configure: No such file or directory\n"> expected but was <"configure failed:\n\nsh ./configure --prefix=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3428/prefix\n">. 3) Failure: test_self_build_has_makefile(TestGemExtConfigureBuilder) [./test/test_gem_ext_configure_builder.rb:76]: <["make", "ok\n", "make install", "ok\n"]> expected but was <["nmake", "", "nmake install", ""]>. 4) Failure: test_class_build(TestGemExtExtConfBuilder) [./test/test_gem_ext_ext_conf_builder.rb:37]: <["ruby extconf.rb", "creating Makefile\n", "make", "make: Nothing to be done for `all'.\n", "make install", "make: Nothing to be done for `install'.\n"]> expected but was <["ruby extconf.rb", "", "nmake", "", "nmake install", ""]>. 5) Error: test_class_build_extconf_bad(TestGemExtExtConfBuilder): Gem::InstallError: make failed: ruby extconf.rbnmake ./lib/rubygems/installer.rb:586:in `make' ./lib/rubygems/installer.rb:581:in `make' ./lib/rubygems/installer.rb:627:in `build' ./test/test_gem_ext_ext_conf_builder.rb:54:in `test_class_build_extconf_bad' ./test/test_gem_ext_ext_conf_builder.rb:53:in `test_class_build_extconf_bad' 6) Failure: test_class_build_extconf_fail(TestGemExtExtConfBuilder) [./test/test_gem_ext_ext_conf_builder.rb:113]: <"extconf failed:\n\nruby extconf.rb\nchecking for main() in -lnonexistent... no\nneed libnonexistent\n*** extconf.rb failed ***\nCould not create Makefile due to some reason, probably lack of\nnecessary libraries and/or headers. Check the mkmf.log file for more\ndetails. You may need configuration options.\n\nProvided configuration options:\n\t--with-opt-dir\n\t--without-opt-dir\n\t--with-opt-include\n\t--without-opt-include=${opt-dir}/include\n\t--with-opt-lib\n\t--without-opt-lib=${opt-dir}/lib\n\t--with-make-prog\n\t--without-make-prog\n\t--srcdir=.\n\t--curdir\n\t--ruby=/usr/local/bin/ruby\n\t--with-nonexistentlib\n\t--without-nonexistentlib\n"> expected but was <"extconf failed:\n\nruby extconf.rb\n">. 7) Failure: test_class_make(TestGemExtExtConfBuilder) [./test/test_gem_ext_ext_conf_builder.rb:141]: <["make", "make: Nothing to be done for `all'.\n", "make install", "make: Nothing to be done for `install'.\n"]> expected but was <["nmake", "", "nmake install", ""]>. 8) Failure: test_class_build(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:35]: <["rake RUBYARCHDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3428/prefix RUBYLIBDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3428/prefix extension", "(in C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3428/ext)\n"]> expected but was <["rake RUBYARCHDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3428/prefix RUBYLIBDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3428/prefix extension", ""]>. 9) Failure: test_class_build_fail(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:46]: exception expected but none was thrown. 10) Failure: test_source_index_hash(TestRemoteInstaller) [./test/test_remote_installer.rb:87]: <1> expected but was <5>. 270 tests, 960 assertions, 8 failures, 2 errors rake aborted! Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib "c:/PRO...] (See full trace by running task with --trace) -- Cerberus 0.3.2, http://cerberus.rubyforge.org/ From drbrain at segment7.net Tue Dec 19 13:19:12 2006 From: drbrain at segment7.net (Eric Hodel) Date: Tue, 19 Dec 2006 10:19:12 -0800 Subject: [Rubygems-developers] [rubygems] Build broken by drbrain (#1144) In-Reply-To: <45882010f0b90_37019b63aa151@ubuntu.tmail> References: <45882010f0b90_37019b63aa151@ubuntu.tmail> Message-ID: <1D9002DB-4E8F-4DEB-AF4D-D691B54EA99F@segment7.net> On Dec 19, 2006, at 09:23, Cerberus wrote: > ---------------------------------------------------------------------- > -- > r1144 | drbrain | 2006-12-19 13:11:39 +0300 | 1 line > Changed paths: > M /trunk/ChangeLog > M /trunk/lib/rubygems/custom_require.rb > M /trunk/lib/rubygems/gem_commands.rb > M /trunk/lib/rubygems/source_index.rb > M /trunk/lib/rubygems.rb > > Use File.join on glob strings. Fixes bug #1096 by Chad Fowler. > ---------------------------------------------------------------------- > -- > > (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) > Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/ > rake/rake_test_loader > Started > ...................................................................... > ................'sh' is not recognized as an internal or external > command, > operable program or batch file. What do we do about this one? How do you invoke (do you invoke?) configure for cmd.exe? Also, this should all get dumped into the results due to the redirector. Is "1<&2" not correct for cmd.exe? > E'sh' is not recognized as an internal or external command, > operable program or batch file. > F > Microsoft (R) Program Maintenance Utility Version 7.10.3077 > Copyright (C) Microsoft Corporation. All rights reserved. > > ok > > Microsoft (R) Program Maintenance Utility Version 7.10.3077 > Copyright (C) Microsoft Corporation. All rights reserved. > > ok > Fcreating Makefile > > Microsoft (R) Program Maintenance Utility Version 7.10.3077 > Copyright (C) Microsoft Corporation. All rights reserved. > > > Microsoft (R) Program Maintenance Utility Version 7.10.3077 > Copyright (C) Microsoft Corporation. All rights reserved. > > Fchecking for main() in nonexistent.lib... no > creating Makefile > > Microsoft (R) Program Maintenance Utility Version 7.10.3077 > Copyright (C) Microsoft Corporation. All rights reserved. > > c:\progra~1\ruby\bin\ruby -e "puts 'EXPORTS', 'Init_foo'" > foo- > i386-mswin32.def > cl -nologo -MD -Zi -O2b2xg- -G6 -I. -Ic:/progra~1/ruby/lib/ruby/ > 1.8/i386-mswin32 -Ic:/progra~1/ruby/lib/ruby/1.8/i386-mswin32 -I. - > c -Tcfoo.c > foo.c > cl -nologo -LD -Fefoo.so foo.obj msvcrt-ruby18.lib oldnames.lib > user32.lib advapi32.lib wsock32.lib -link -incremental:no -debug - > opt:ref -opt:icf -dll -libpath:"c:/progra~1/ruby/lib" -def:foo-i386- > mswin32.def -implib:foo-i386-mswin32.lib -pdb:foo-i386-mswin32.pdb > Creating library foo-i386-mswin32.lib and object foo-i386- > mswin32.exp > foo.obj : error LNK2019: unresolved external symbol _nonexistent > referenced in function _Init_foo > foo.so : fatal error LNK1120: 1 unresolved externals > NMAKE : fatal error U1077: 'cl' : return code '0x2' > Stop. > Echecking for main() in nonexistent.lib... no > need libnonexistent > *** extconf.rb failed *** > Could not create Makefile due to some reason, probably lack of > necessary libraries and/or headers. Check the mkmf.log file for more > details. You may need configuration options. > > Provided configuration options: > --with-opt-dir > --without-opt-dir > --with-opt-include > --without-opt-include=${opt-dir}/include > --with-opt-lib > --without-opt-lib=${opt-dir}/lib > --with-make-prog > --srcdir=. > --curdir > --ruby=c:/progra~1/ruby/bin/ruby > --with-nonexistentlib > --without-nonexistentlib > F > Microsoft (R) Program Maintenance Utility Version 7.10.3077 > Copyright (C) Microsoft Corporation. All rights reserved. > > > Microsoft (R) Program Maintenance Utility Version 7.10.3077 > Copyright (C) Microsoft Corporation. All rights reserved. > > F.(in C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3428/ext) > F(in C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3428/ext) > rake aborted! > fail > C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3428/ext/rakefile: > 1:in `abort' > (See full trace by running task with --trace) > fail > F..................................................................... > ..........F........................................................... > ................................... > Finished in 80.578 seconds. > > 1) Error: > test_self_build(TestGemExtConfigureBuilder): > Gem::InstallError: configure failed: > > sh ./configure --prefix=C:/DOCUME~1/anatol/LOCALS~1/Temp/ > test_rubygems_3428/prefix > > ./lib/rubygems/installer.rb:600:in `run' > ./lib/rubygems/installer.rb:611:in `build' > ./test/test_gem_ext_configure_builder.rb:28:in `test_self_build' > ./test/test_gem_ext_configure_builder.rb:27:in `test_self_build' > > 2) Failure: > test_self_build_fail(TestGemExtConfigureBuilder) [./test/ > test_gem_ext_configure_builder.rb:54]: > <"configure failed:\n\nsh ./configure --prefix=C:/DOCUME~1/anatol/ > LOCALS~1/Temp/test_rubygems_3428/prefix\n./configure: ./configure: > No such file or directory\n"> expected but was > <"configure failed:\n\nsh ./configure --prefix=C:/DOCUME~1/anatol/ > LOCALS~1/Temp/test_rubygems_3428/prefix\n">. > > 3) Failure: > test_self_build_has_makefile(TestGemExtConfigureBuilder) [./test/ > test_gem_ext_configure_builder.rb:76]: > <["make", "ok\n", "make install", "ok\n"]> expected but was > <["nmake", "", "nmake install", ""]>. I can probably fix these. > 4) Failure: > test_class_build(TestGemExtExtConfBuilder) [./test/ > test_gem_ext_ext_conf_builder.rb:37]: > <["ruby extconf.rb", > "creating Makefile\n", > "make", > "make: Nothing to be done for `all'.\n", > "make install", > "make: Nothing to be done for `install'.\n"]> expected but was > <["ruby extconf.rb", "", "nmake", "", "nmake install", ""]>. > > 5) Error: > test_class_build_extconf_bad(TestGemExtExtConfBuilder): > Gem::InstallError: make failed: > > ruby extconf.rbnmake > ./lib/rubygems/installer.rb:586:in `make' > ./lib/rubygems/installer.rb:581:in `make' > ./lib/rubygems/installer.rb:627:in `build' > ./test/test_gem_ext_ext_conf_builder.rb:54:in > `test_class_build_extconf_bad' > ./test/test_gem_ext_ext_conf_builder.rb:53:in > `test_class_build_extconf_bad' > > 6) Failure: > test_class_build_extconf_fail(TestGemExtExtConfBuilder) [./test/ > test_gem_ext_ext_conf_builder.rb:113]: > <"extconf failed:\n\nruby extconf.rb\nchecking for main() in - > lnonexistent... no\nneed libnonexistent\n*** extconf.rb failed *** > \nCould not create Makefile due to some reason, probably lack of > \nnecessary libraries and/or headers. Check the mkmf.log file for > more\ndetails. You may need configuration options.\n\nProvided > configuration options:\n\t--with-opt-dir\n\t--without-opt-dir\n\t-- > with-opt-include\n\t--without-opt-include=${opt-dir}/include\n\t-- > with-opt-lib\n\t--without-opt-lib=${opt-dir}/lib\n\t--with-make-prog > \n\t--without-make-prog\n\t--srcdir=.\n\t--curdir\n\t--ruby=/usr/ > local/bin/ruby\n\t--with-nonexistentlib\n\t--without-nonexistentlib > \n"> expected but was > <"extconf failed:\n\nruby extconf.rb\n">. > > 7) Failure: > test_class_make(TestGemExtExtConfBuilder) [./test/ > test_gem_ext_ext_conf_builder.rb:141]: > <["make", > "make: Nothing to be done for `all'.\n", > "make install", > "make: Nothing to be done for `install'.\n"]> expected but was > <["nmake", "", "nmake install", ""]>. > > 8) Failure: > test_class_build(TestGemExtRakeBuilder) [./test/ > test_gem_ext_rake_builder.rb:35]: > <["rake RUBYARCHDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/ > test_rubygems_3428/prefix RUBYLIBDIR=C:/DOCUME~1/anatol/LOCALS~1/ > Temp/test_rubygems_3428/prefix extension", > "(in C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3428/ext)\n"]> > expected but was > <["rake RUBYARCHDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/ > test_rubygems_3428/prefix RUBYLIBDIR=C:/DOCUME~1/anatol/LOCALS~1/ > Temp/test_rubygems_3428/prefix extension", > ""]>. What version of rake do you have installed? > 9) Failure: > test_class_build_fail(TestGemExtRakeBuilder) [./test/ > test_gem_ext_rake_builder.rb:46]: > exception expected but none was thrown. > > 10) Failure: > test_source_index_hash(TestRemoteInstaller) [./test/ > test_remote_installer.rb:87]: > <1> expected but was > <5>. > > 270 tests, 960 assertions, 8 failures, 2 errors > rake aborted! > Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib > "c:/PRO...] > > (See full trace by running task with --trace) > > -- > Cerberus 0.3.2, http://cerberus.rubyforge.org/ > _______________________________________________ > Rubygems-developers mailing list > Rubygems-developers at rubyforge.org > http://rubyforge.org/mailman/listinfo/rubygems-developers -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From anthonyeden at gmail.com Tue Dec 19 17:50:02 2006 From: anthonyeden at gmail.com (Anthony Eden) Date: Tue, 19 Dec 2006 17:50:02 -0500 Subject: [Rubygems-developers] Adding Tasks to Rake from a Gem? Message-ID: Is it possible to add tasks to Rake at runtime from a Gem? Right now I can include the following in Ruby code to get make the library available: require_gem 'activewarehouse' require 'active_warehouse' However this will not expose my custom tasks to Rake (they are currently defined in tasks/*.rake). Any thoughts on how I can cleanly get my Rake tasks added on the fly? I am currently using Rake 0.7.1. Thanks for any help on this. V/r Anthony -- Cell: 808 782-5046 Current Location: Melbourne, FL From drnicwilliams at gmail.com Tue Dec 19 18:27:23 2006 From: drnicwilliams at gmail.com (Nic Williams) Date: Wed, 20 Dec 2006 10:27:23 +1100 Subject: [Rubygems-developers] Adding Tasks to Rake from a Gem? In-Reply-To: References: Message-ID: <44b555bb0612191527y3700fe22y625e763967b25d64@mail.gmail.com> Two options: i) In your Rakefile, you need to load the target rake files. ii) In a Rails app, you could use gemsonrails project to link/freeze the gem into your app, and it will automatically find your Rakefiles. ( https://rubyforge.org/projects/gemsonrails/) Nic On 12/20/06, Anthony Eden wrote: > > Is it possible to add tasks to Rake at runtime from a Gem? Right now I > can include the following in Ruby code to get make the library > available: > > require_gem 'activewarehouse' > require 'active_warehouse' > > However this will not expose my custom tasks to Rake (they are > currently defined in tasks/*.rake). Any thoughts on how I can cleanly > get my Rake tasks added on the fly? I am currently using Rake 0.7.1. > Thanks for any help on this. > > V/r > Anthony > > -- > Cell: 808 782-5046 > Current Location: Melbourne, FL > _______________________________________________ > Rubygems-developers mailing list > Rubygems-developers at rubyforge.org > http://rubyforge.org/mailman/listinfo/rubygems-developers > -- Dr Nic Williams http://www.drnicwilliams.com - Ruby/Rails blog skype: nicwilliams (m) +31 62 494 8552 (p) +61 7 3102 3237 (finds me anywhere in the world) (f) +61 7 3305 7572 (sends fax to my email) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20061220/064347fa/attachment.html From drnicwilliams at gmail.com Tue Dec 19 18:35:53 2006 From: drnicwilliams at gmail.com (Nic Williams) Date: Wed, 20 Dec 2006 10:35:53 +1100 Subject: [Rubygems-developers] Adding Tasks to Rake from a Gem? In-Reply-To: <44b555bb0612191527y3700fe22y625e763967b25d64@mail.gmail.com> References: <44b555bb0612191527y3700fe22y625e763967b25d64@mail.gmail.com> Message-ID: <44b555bb0612191535v4aafa848ud729019c5dea60d6@mail.gmail.com> Sorry, it will load the *.rake files, not "Rakefiles" from the gem. On 12/20/06, Nic Williams wrote: > > Two options: > > i) In your Rakefile, you need to load the target rake files. > ii) In a Rails app, you could use gemsonrails project to link/freeze the > gem into your app, and it will automatically find your Rakefiles. ( > https://rubyforge.org/projects/gemsonrails/) > > Nic > > On 12/20/06, Anthony Eden < anthonyeden at gmail.com> wrote: > > > > Is it possible to add tasks to Rake at runtime from a Gem? Right now I > > can include the following in Ruby code to get make the library > > available: > > > > require_gem 'activewarehouse' > > require 'active_warehouse' > > > > However this will not expose my custom tasks to Rake (they are > > currently defined in tasks/*.rake). Any thoughts on how I can cleanly > > get my Rake tasks added on the fly? I am currently using Rake 0.7.1. > > Thanks for any help on this. > > > > V/r > > Anthony > > > > -- > > Cell: 808 782-5046 > > Current Location: Melbourne, FL > > _______________________________________________ > > Rubygems-developers mailing list > > Rubygems-developers at rubyforge.org > > http://rubyforge.org/mailman/listinfo/rubygems-developers > > > > > > -- > Dr Nic Williams > http://www.drnicwilliams.com - Ruby/Rails blog > skype: nicwilliams > (m) +31 62 494 8552 > (p) +61 7 3102 3237 (finds me anywhere in the world) > (f) +61 7 3305 7572 (sends fax to my email) -- Dr Nic Williams http://www.drnicwilliams.com - Ruby/Rails blog skype: nicwilliams (m) +31 62 494 8552 (p) +61 7 3102 3237 (finds me anywhere in the world) (f) +61 7 3305 7572 (sends fax to my email) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20061220/d9839828/attachment.html From anthonyeden at gmail.com Tue Dec 19 20:40:25 2006 From: anthonyeden at gmail.com (Anthony Eden) Date: Tue, 19 Dec 2006 20:40:25 -0500 Subject: [Rubygems-developers] Adding Tasks to Rake from a Gem? In-Reply-To: <44b555bb0612191527y3700fe22y625e763967b25d64@mail.gmail.com> References: <44b555bb0612191527y3700fe22y625e763967b25d64@mail.gmail.com> Message-ID: Thanks, I think gemsonrails is the way to go. V/r Anthony On 12/19/06, Nic Williams wrote: > Two options: > > i) In your Rakefile, you need to load the target rake files. > ii) In a Rails app, you could use gemsonrails project to link/freeze the gem > into your app, and it will automatically find your Rakefiles. ( > https://rubyforge.org/projects/gemsonrails/) > > Nic > > > On 12/20/06, Anthony Eden < anthonyeden at gmail.com> wrote: > > > > Is it possible to add tasks to Rake at runtime from a Gem? Right now I > > can include the following in Ruby code to get make the library > > available: > > > > require_gem 'activewarehouse' > > require 'active_warehouse' > > > > However this will not expose my custom tasks to Rake (they are > > currently defined in tasks/*.rake). Any thoughts on how I can cleanly > > get my Rake tasks added on the fly? I am currently using Rake 0.7.1. > > Thanks for any help on this. > > > > V/r > > Anthony > > > > -- > > Cell: 808 782-5046 > > Current Location: Melbourne, FL > > _______________________________________________ > > Rubygems-developers mailing list > > Rubygems-developers at rubyforge.org > > http://rubyforge.org/mailman/listinfo/rubygems-developers > > > > > > -- > Dr Nic Williams > http://www.drnicwilliams.com - Ruby/Rails blog > skype: nicwilliams > (m) +31 62 494 8552 > (p) +61 7 3102 3237 (finds me anywhere in the world) > (f) +61 7 3305 7572 (sends fax to my email) > _______________________________________________ > Rubygems-developers mailing list > Rubygems-developers at rubyforge.org > http://rubyforge.org/mailman/listinfo/rubygems-developers > > -- Cell: 808 782-5046 Current Location: Melbourne, FL From halostatue at gmail.com Tue Dec 19 21:59:23 2006 From: halostatue at gmail.com (Austin Ziegler) Date: Tue, 19 Dec 2006 21:59:23 -0500 Subject: [Rubygems-developers] [rubygems] Build broken by drbrain (#1144) In-Reply-To: <1D9002DB-4E8F-4DEB-AF4D-D691B54EA99F@segment7.net> References: <45882010f0b90_37019b63aa151@ubuntu.tmail> <1D9002DB-4E8F-4DEB-AF4D-D691B54EA99F@segment7.net> Message-ID: <9e7db9110612191859j5e995309u5da048e4ad7b8e3a@mail.gmail.com> On 12/19/06, Eric Hodel wrote: > What do we do about this one? How do you invoke (do you invoke?) > configure for cmd.exe? You don't. Configure is based on sh. > Also, this should all get dumped into the results due to the > redirector. Is "1<&2" not correct for cmd.exe? 2>&1 for output will work. Is that what you're doing? -austin -- Austin Ziegler * halostatue at gmail.com * http://www.halostatue.ca/ * austin at halostatue.ca * http://www.halostatue.ca/feed/ * austin at zieglers.ca From anatol.pomozov at gmail.com Tue Dec 19 23:01:45 2006 From: anatol.pomozov at gmail.com (Cerberus) Date: Wed, 20 Dec 2006 07:01:45 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1145) Message-ID: <4588b5a994700_7c019b68821a0@ubuntu.tmail> ------------------------------------------------------------------------ r1145 | drbrain | 2006-12-20 06:51:30 +0300 | 1 line Changed paths: M /trunk/lib/rubygems/installer.rb Fix redirector ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/rake/rake_test_loader Started ......................................................................................EFFFEFF..F...............................................................................F.............................................................................................. Finished in 51.922 seconds. 1) Error: test_self_build(TestGemExtConfigureBuilder): Gem::InstallError: configure failed: sh ./configure --prefix=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3540/prefix 'sh' is not recognized as an internal or external command, operable program or batch file. ./lib/rubygems/installer.rb:600:in `run' ./lib/rubygems/installer.rb:611:in `build' ./test/test_gem_ext_configure_builder.rb:28:in `test_self_build' ./test/test_gem_ext_configure_builder.rb:27:in `test_self_build' 2) Failure: test_self_build_fail(TestGemExtConfigureBuilder) [./test/test_gem_ext_configure_builder.rb:54]: <"configure failed:\n\nsh ./configure --prefix=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3540/prefix\n./configure: ./configure: No such file or directory\n"> expected but was <"configure failed:\n\nsh ./configure --prefix=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3540/prefix\n'sh' is not recognized as an internal or external command,\noperable program or batch file.\n">. 3) Failure: test_self_build_has_makefile(TestGemExtConfigureBuilder) [./test/test_gem_ext_configure_builder.rb:76]: <["make", "ok\n", "make install", "ok\n"]> expected but was <["nmake", "\nMicrosoft (R) Program Maintenance Utility Version 7.10.3077\nCopyright (C) Microsoft Corporation. All rights reserved.\n\nok\n", "nmake install", "\nMicrosoft (R) Program Maintenance Utility Version 7.10.3077\nCopyright (C) Microsoft Corporation. All rights reserved.\n\nok\n"]>. 4) Failure: test_class_build(TestGemExtExtConfBuilder) [./test/test_gem_ext_ext_conf_builder.rb:37]: <["ruby extconf.rb", "creating Makefile\n", "make", "make: Nothing to be done for `all'.\n", "make install", "make: Nothing to be done for `install'.\n"]> expected but was <["ruby extconf.rb", "creating Makefile\n", "nmake", "\nMicrosoft (R) Program Maintenance Utility Version 7.10.3077\nCopyright (C) Microsoft Corporation. All rights reserved.\n\n", "nmake install", "\nMicrosoft (R) Program Maintenance Utility Version 7.10.3077\nCopyright (C) Microsoft Corporation. All rights reserved.\n\n"]>. 5) Error: test_class_build_extconf_bad(TestGemExtExtConfBuilder): Gem::InstallError: make failed: ruby extconf.rbchecking for main() in nonexistent.lib... no creating Makefile nmake Microsoft (R) Program Maintenance Utility Version 7.10.3077 Copyright (C) Microsoft Corporation. All rights reserved. c:\progra~1\ruby\bin\ruby -e "puts 'EXPORTS', 'Init_foo'" > foo-i386-mswin32.def cl -nologo -MD -Zi -O2b2xg- -G6 -I. -Ic:/progra~1/ruby/lib/ruby/1.8/i386-mswin32 -Ic:/progra~1/ruby/lib/ruby/1.8/i386-mswin32 -I. -c -Tcfoo.c foo.c cl -nologo -LD -Fefoo.so foo.obj msvcrt-ruby18.lib oldnames.lib user32.lib advapi32.lib wsock32.lib -link -incremental:no -debug -opt:ref -opt:icf -dll -libpath:"c:/progra~1/ruby/lib" -def:foo-i386-mswin32.def -implib:foo-i386-mswin32.lib -pdb:foo-i386-mswin32.pdb Creating library foo-i386-mswin32.lib and object foo-i386-mswin32.exp foo.obj : error LNK2019: unresolved external symbol _nonexistent referenced in function _Init_foo foo.so : fatal error LNK1120: 1 unresolved externals NMAKE : fatal error U1077: 'cl' : return code '0x2' Stop. ./lib/rubygems/installer.rb:586:in `make' ./lib/rubygems/installer.rb:581:in `make' ./lib/rubygems/installer.rb:627:in `build' ./test/test_gem_ext_ext_conf_builder.rb:54:in `test_class_build_extconf_bad' ./test/test_gem_ext_ext_conf_builder.rb:53:in `test_class_build_extconf_bad' 6) Failure: test_class_build_extconf_fail(TestGemExtExtConfBuilder) [./test/test_gem_ext_ext_conf_builder.rb:113]: <"extconf failed:\n\nruby extconf.rb\nchecking for main() in -lnonexistent... no\nneed libnonexistent\n*** extconf.rb failed ***\nCould not create Makefile due to some reason, probably lack of\nnecessary libraries and/or headers. Check the mkmf.log file for more\ndetails. You may need configuration options.\n\nProvided configuration options:\n\t--with-opt-dir\n\t--without-opt-dir\n\t--with-opt-include\n\t--without-opt-include=${opt-dir}/include\n\t--with-opt-lib\n\t--without-opt-lib=${opt-dir}/lib\n\t--with-make-prog\n\t--without-make-prog\n\t--srcdir=.\n\t--curdir\n\t--ruby=/usr/local/bin/ruby\n\t--with-nonexistentlib\n\t--without-nonexistentlib\n"> expected but was <"extconf failed:\n\nruby extconf.rb\nchecking for main() in nonexistent.lib... no\nneed libnonexistent\n*** extconf.rb failed ***\nCould not create Makefile due to some reason, probably lack of\nnecessary libraries and/or headers. Check the mkmf.log file for more\ndetails. You may need configuration options.\n\nProvided configuration options:\n\t--with-opt-dir\n\t--without-opt-dir\n\t--with-opt-include\n\t--without-opt-include=${opt-dir}/include\n\t--with-opt-lib\n\t--without-opt-lib=${opt-dir}/lib\n\t--with-make-prog\n\t--srcdir=.\n\t--curdir\n\t--ruby=c:/progra~1/ruby/bin/ruby\n\t--with-nonexistentlib\n\t--without-nonexistentlib\n">. 7) Failure: test_class_make(TestGemExtExtConfBuilder) [./test/test_gem_ext_ext_conf_builder.rb:141]: <["make", "make: Nothing to be done for `all'.\n", "make install", "make: Nothing to be done for `install'.\n"]> expected but was <["nmake", "\nMicrosoft (R) Program Maintenance Utility Version 7.10.3077\nCopyright (C) Microsoft Corporation. All rights reserved.\n\n", "nmake install", "\nMicrosoft (R) Program Maintenance Utility Version 7.10.3077\nCopyright (C) Microsoft Corporation. All rights reserved.\n\n"]>. 8) Failure: test_class_build_fail(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:46]: exception expected but none was thrown. 9) Failure: test_source_index_hash(TestRemoteInstaller) [./test/test_remote_installer.rb:87]: <1> expected but was <5>. 270 tests, 960 assertions, 7 failures, 2 errors rake aborted! Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib "c:/PRO...] (See full trace by running task with --trace) -- Cerberus 0.3.2, http://cerberus.rubyforge.org/ From anatol.pomozov at gmail.com Tue Dec 19 23:41:37 2006 From: anatol.pomozov at gmail.com (Cerberus) Date: Wed, 20 Dec 2006 07:41:37 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1146) Message-ID: <4588bf014fd58_81019b68821fb@ubuntu.tmail> ------------------------------------------------------------------------ r1146 | drbrain | 2006-12-20 07:23:58 +0300 | 1 line Changed paths: M /trunk/ChangeLog M /trunk/lib/rubygems/remote_installer.rb Make --force force installation. ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/rake/rake_test_loader Started ......................................................................................EFFFEFF..F...............................................................................F.............................................................................................. Finished in 50.734 seconds. 1) Error: test_self_build(TestGemExtConfigureBuilder): Gem::InstallError: configure failed: sh ./configure --prefix=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3924/prefix 'sh' is not recognized as an internal or external command, operable program or batch file. ./lib/rubygems/installer.rb:600:in `run' ./lib/rubygems/installer.rb:611:in `build' ./test/test_gem_ext_configure_builder.rb:28:in `test_self_build' ./test/test_gem_ext_configure_builder.rb:27:in `test_self_build' 2) Failure: test_self_build_fail(TestGemExtConfigureBuilder) [./test/test_gem_ext_configure_builder.rb:54]: <"configure failed:\n\nsh ./configure --prefix=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3924/prefix\n./configure: ./configure: No such file or directory\n"> expected but was <"configure failed:\n\nsh ./configure --prefix=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3924/prefix\n'sh' is not recognized as an internal or external command,\noperable program or batch file.\n">. 3) Failure: test_self_build_has_makefile(TestGemExtConfigureBuilder) [./test/test_gem_ext_configure_builder.rb:76]: <["make", "ok\n", "make install", "ok\n"]> expected but was <["nmake", "\nMicrosoft (R) Program Maintenance Utility Version 7.10.3077\nCopyright (C) Microsoft Corporation. All rights reserved.\n\nok\n", "nmake install", "\nMicrosoft (R) Program Maintenance Utility Version 7.10.3077\nCopyright (C) Microsoft Corporation. All rights reserved.\n\nok\n"]>. 4) Failure: test_class_build(TestGemExtExtConfBuilder) [./test/test_gem_ext_ext_conf_builder.rb:37]: <["ruby extconf.rb", "creating Makefile\n", "make", "make: Nothing to be done for `all'.\n", "make install", "make: Nothing to be done for `install'.\n"]> expected but was <["ruby extconf.rb", "creating Makefile\n", "nmake", "\nMicrosoft (R) Program Maintenance Utility Version 7.10.3077\nCopyright (C) Microsoft Corporation. All rights reserved.\n\n", "nmake install", "\nMicrosoft (R) Program Maintenance Utility Version 7.10.3077\nCopyright (C) Microsoft Corporation. All rights reserved.\n\n"]>. 5) Error: test_class_build_extconf_bad(TestGemExtExtConfBuilder): Gem::InstallError: make failed: ruby extconf.rbchecking for main() in nonexistent.lib... no creating Makefile nmake Microsoft (R) Program Maintenance Utility Version 7.10.3077 Copyright (C) Microsoft Corporation. All rights reserved. c:\progra~1\ruby\bin\ruby -e "puts 'EXPORTS', 'Init_foo'" > foo-i386-mswin32.def cl -nologo -MD -Zi -O2b2xg- -G6 -I. -Ic:/progra~1/ruby/lib/ruby/1.8/i386-mswin32 -Ic:/progra~1/ruby/lib/ruby/1.8/i386-mswin32 -I. -c -Tcfoo.c foo.c cl -nologo -LD -Fefoo.so foo.obj msvcrt-ruby18.lib oldnames.lib user32.lib advapi32.lib wsock32.lib -link -incremental:no -debug -opt:ref -opt:icf -dll -libpath:"c:/progra~1/ruby/lib" -def:foo-i386-mswin32.def -implib:foo-i386-mswin32.lib -pdb:foo-i386-mswin32.pdb Creating library foo-i386-mswin32.lib and object foo-i386-mswin32.exp foo.obj : error LNK2019: unresolved external symbol _nonexistent referenced in function _Init_foo foo.so : fatal error LNK1120: 1 unresolved externals NMAKE : fatal error U1077: 'cl' : return code '0x2' Stop. ./lib/rubygems/installer.rb:586:in `make' ./lib/rubygems/installer.rb:581:in `make' ./lib/rubygems/installer.rb:627:in `build' ./test/test_gem_ext_ext_conf_builder.rb:54:in `test_class_build_extconf_bad' ./test/test_gem_ext_ext_conf_builder.rb:53:in `test_class_build_extconf_bad' 6) Failure: test_class_build_extconf_fail(TestGemExtExtConfBuilder) [./test/test_gem_ext_ext_conf_builder.rb:113]: <"extconf failed:\n\nruby extconf.rb\nchecking for main() in -lnonexistent... no\nneed libnonexistent\n*** extconf.rb failed ***\nCould not create Makefile due to some reason, probably lack of\nnecessary libraries and/or headers. Check the mkmf.log file for more\ndetails. You may need configuration options.\n\nProvided configuration options:\n\t--with-opt-dir\n\t--without-opt-dir\n\t--with-opt-include\n\t--without-opt-include=${opt-dir}/include\n\t--with-opt-lib\n\t--without-opt-lib=${opt-dir}/lib\n\t--with-make-prog\n\t--without-make-prog\n\t--srcdir=.\n\t--curdir\n\t--ruby=/usr/local/bin/ruby\n\t--with-nonexistentlib\n\t--without-nonexistentlib\n"> expected but was <"extconf failed:\n\nruby extconf.rb\nchecking for main() in nonexistent.lib... no\nneed libnonexistent\n*** extconf.rb failed ***\nCould not create Makefile due to some reason, probably lack of\nnecessary libraries and/or headers. Check the mkmf.log file for more\ndetails. You may need configuration options.\n\nProvided configuration options:\n\t--with-opt-dir\n\t--without-opt-dir\n\t--with-opt-include\n\t--without-opt-include=${opt-dir}/include\n\t--with-opt-lib\n\t--without-opt-lib=${opt-dir}/lib\n\t--with-make-prog\n\t--srcdir=.\n\t--curdir\n\t--ruby=c:/progra~1/ruby/bin/ruby\n\t--with-nonexistentlib\n\t--without-nonexistentlib\n">. 7) Failure: test_class_make(TestGemExtExtConfBuilder) [./test/test_gem_ext_ext_conf_builder.rb:141]: <["make", "make: Nothing to be done for `all'.\n", "make install", "make: Nothing to be done for `install'.\n"]> expected but was <["nmake", "\nMicrosoft (R) Program Maintenance Utility Version 7.10.3077\nCopyright (C) Microsoft Corporation. All rights reserved.\n\n", "nmake install", "\nMicrosoft (R) Program Maintenance Utility Version 7.10.3077\nCopyright (C) Microsoft Corporation. All rights reserved.\n\n"]>. 8) Failure: test_class_build_fail(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:46]: exception expected but none was thrown. 9) Failure: test_source_index_hash(TestRemoteInstaller) [./test/test_remote_installer.rb:87]: <1> expected but was <5>. 270 tests, 960 assertions, 7 failures, 2 errors rake aborted! Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib "c:/PRO...] (See full trace by running task with --trace) -- Cerberus 0.3.2, http://cerberus.rubyforge.org/ From anatol.pomozov at gmail.com Wed Dec 20 04:21:37 2006 From: anatol.pomozov at gmail.com (Cerberus) Date: Wed, 20 Dec 2006 12:21:37 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1147) Message-ID: <458900a131510_ec19b04a0127@ubuntu.tmail> ------------------------------------------------------------------------ r1147 | drbrain | 2006-12-20 12:18:22 +0300 | 3 lines Changed paths: M /trunk M /trunk/ChangeLog M /trunk/lib/rubygems/cmd_manager.rb M /trunk/lib/rubygems/gem_commands.rb M /trunk/lib/rubygems/remote_fetcher.rb M /trunk/test/test_gem_source_info_cache_entry.rb A /trunk/test/test_gem_sources_command.rb M /trunk/test/test_remote_fetcher.rb RemoteFetcher#fetch_size gives a sensible error on bad URIs. Add command to manage sources listed in the source_cache. Fixes bug #1128. ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/rake/rake_test_loader Started ......................................................................................EFFFEFF..F.......................................................................................F.............................................................................................. Finished in 65.219 seconds. 1) Error: test_self_build(TestGemExtConfigureBuilder): Gem::InstallError: configure failed: sh ./configure --prefix=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3784/prefix 'sh' is not recognized as an internal or external command, operable program or batch file. ./lib/rubygems/installer.rb:600:in `run' ./lib/rubygems/installer.rb:611:in `build' ./test/test_gem_ext_configure_builder.rb:28:in `test_self_build' ./test/test_gem_ext_configure_builder.rb:27:in `test_self_build' 2) Failure: test_self_build_fail(TestGemExtConfigureBuilder) [./test/test_gem_ext_configure_builder.rb:54]: <"configure failed:\n\nsh ./configure --prefix=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3784/prefix\n./configure: ./configure: No such file or directory\n"> expected but was <"configure failed:\n\nsh ./configure --prefix=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3784/prefix\n'sh' is not recognized as an internal or external command,\noperable program or batch file.\n">. 3) Failure: test_self_build_has_makefile(TestGemExtConfigureBuilder) [./test/test_gem_ext_configure_builder.rb:76]: <["make", "ok\n", "make install", "ok\n"]> expected but was <["nmake", "\nMicrosoft (R) Program Maintenance Utility Version 7.10.3077\nCopyright (C) Microsoft Corporation. All rights reserved.\n\nok\n", "nmake install", "\nMicrosoft (R) Program Maintenance Utility Version 7.10.3077\nCopyright (C) Microsoft Corporation. All rights reserved.\n\nok\n"]>. 4) Failure: test_class_build(TestGemExtExtConfBuilder) [./test/test_gem_ext_ext_conf_builder.rb:37]: <["ruby extconf.rb", "creating Makefile\n", "make", "make: Nothing to be done for `all'.\n", "make install", "make: Nothing to be done for `install'.\n"]> expected but was <["ruby extconf.rb", "creating Makefile\n", "nmake", "\nMicrosoft (R) Program Maintenance Utility Version 7.10.3077\nCopyright (C) Microsoft Corporation. All rights reserved.\n\n", "nmake install", "\nMicrosoft (R) Program Maintenance Utility Version 7.10.3077\nCopyright (C) Microsoft Corporation. All rights reserved.\n\n"]>. 5) Error: test_class_build_extconf_bad(TestGemExtExtConfBuilder): Gem::InstallError: make failed: ruby extconf.rbchecking for main() in nonexistent.lib... no creating Makefile nmake Microsoft (R) Program Maintenance Utility Version 7.10.3077 Copyright (C) Microsoft Corporation. All rights reserved. c:\progra~1\ruby\bin\ruby -e "puts 'EXPORTS', 'Init_foo'" > foo-i386-mswin32.def cl -nologo -MD -Zi -O2b2xg- -G6 -I. -Ic:/progra~1/ruby/lib/ruby/1.8/i386-mswin32 -Ic:/progra~1/ruby/lib/ruby/1.8/i386-mswin32 -I. -c -Tcfoo.c foo.c cl -nologo -LD -Fefoo.so foo.obj msvcrt-ruby18.lib oldnames.lib user32.lib advapi32.lib wsock32.lib -link -incremental:no -debug -opt:ref -opt:icf -dll -libpath:"c:/progra~1/ruby/lib" -def:foo-i386-mswin32.def -implib:foo-i386-mswin32.lib -pdb:foo-i386-mswin32.pdb Creating library foo-i386-mswin32.lib and object foo-i386-mswin32.exp foo.obj : error LNK2019: unresolved external symbol _nonexistent referenced in function _Init_foo foo.so : fatal error LNK1120: 1 unresolved externals NMAKE : fatal error U1077: 'cl' : return code '0x2' Stop. ./lib/rubygems/installer.rb:586:in `make' ./lib/rubygems/installer.rb:581:in `make' ./lib/rubygems/installer.rb:627:in `build' ./test/test_gem_ext_ext_conf_builder.rb:54:in `test_class_build_extconf_bad' ./test/test_gem_ext_ext_conf_builder.rb:53:in `test_class_build_extconf_bad' 6) Failure: test_class_build_extconf_fail(TestGemExtExtConfBuilder) [./test/test_gem_ext_ext_conf_builder.rb:113]: <"extconf failed:\n\nruby extconf.rb\nchecking for main() in -lnonexistent... no\nneed libnonexistent\n*** extconf.rb failed ***\nCould not create Makefile due to some reason, probably lack of\nnecessary libraries and/or headers. Check the mkmf.log file for more\ndetails. You may need configuration options.\n\nProvided configuration options:\n\t--with-opt-dir\n\t--without-opt-dir\n\t--with-opt-include\n\t--without-opt-include=${opt-dir}/include\n\t--with-opt-lib\n\t--without-opt-lib=${opt-dir}/lib\n\t--with-make-prog\n\t--without-make-prog\n\t--srcdir=.\n\t--curdir\n\t--ruby=/usr/local/bin/ruby\n\t--with-nonexistentlib\n\t--without-nonexistentlib\n"> expected but was <"extconf failed:\n\nruby extconf.rb\nchecking for main() in nonexistent.lib... no\nneed libnonexistent\n*** extconf.rb failed ***\nCould not create Makefile due to some reason, probably lack of\nnecessary libraries and/or headers. Check the mkmf.log file for more\ndetails. You may need configuration options.\n\nProvided configuration options:\n\t--with-opt-dir\n\t--without-opt-dir\n\t--with-opt-include\n\t--without-opt-include=${opt-dir}/include\n\t--with-opt-lib\n\t--without-opt-lib=${opt-dir}/lib\n\t--with-make-prog\n\t--srcdir=.\n\t--curdir\n\t--ruby=c:/progra~1/ruby/bin/ruby\n\t--with-nonexistentlib\n\t--without-nonexistentlib\n">. 7) Failure: test_class_make(TestGemExtExtConfBuilder) [./test/test_gem_ext_ext_conf_builder.rb:141]: <["make", "make: Nothing to be done for `all'.\n", "make install", "make: Nothing to be done for `install'.\n"]> expected but was <["nmake", "\nMicrosoft (R) Program Maintenance Utility Version 7.10.3077\nCopyright (C) Microsoft Corporation. All rights reserved.\n\n", "nmake install", "\nMicrosoft (R) Program Maintenance Utility Version 7.10.3077\nCopyright (C) Microsoft Corporation. All rights reserved.\n\n"]>. 8) Failure: test_class_build_fail(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:46]: exception expected but none was thrown. 9) Failure: test_source_index_hash(TestRemoteInstaller) [./test/test_remote_installer.rb:87]: <1> expected but was <6>. 278 tests, 977 assertions, 7 failures, 2 errors rake aborted! Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib "c:/PRO...] (See full trace by running task with --trace) -- Cerberus 0.3.2, http://cerberus.rubyforge.org/ From anatol.pomozov at gmail.com Wed Dec 20 04:41:23 2006 From: anatol.pomozov at gmail.com (Cerberus) Date: Wed, 20 Dec 2006 12:41:23 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1148) Message-ID: <45890543ab630_a6c19b04a01d1@ubuntu.tmail> ------------------------------------------------------------------------ r1148 | drbrain | 2006-12-20 12:35:25 +0300 | 1 line Changed paths: M /trunk/test/test_gem_ext_configure_builder.rb M /trunk/test/test_gem_ext_ext_conf_builder.rb Make tests less overzealous for win32 ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/rake/rake_test_loader Started ......................................................................................EF..E....F.......................................................................................F.............................................................................................. Finished in 49.157 seconds. 1) Error: test_self_build(TestGemExtConfigureBuilder): Gem::InstallError: configure failed: sh ./configure --prefix=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3824/prefix 'sh' is not recognized as an internal or external command, operable program or batch file. ./lib/rubygems/installer.rb:600:in `run' ./lib/rubygems/installer.rb:611:in `build' ./test/test_gem_ext_configure_builder.rb:30:in `test_self_build' ./test/test_gem_ext_configure_builder.rb:29:in `test_self_build' 2) Failure: test_self_build_fail(TestGemExtConfigureBuilder) [./test/test_gem_ext_configure_builder.rb:57]: <"configure failed:\n\nsh ./configure --prefix=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3824/prefix\n./configure: ./configure: No such file or directory\n"> expected but was <"configure failed:\n\nsh ./configure --prefix=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3824/prefix\n'sh' is not recognized as an internal or external command,\noperable program or batch file.\n">. 3) Error: test_class_build_extconf_bad(TestGemExtExtConfBuilder): Gem::InstallError: make failed: ruby extconf.rbchecking for main() in nonexistent.lib... no creating Makefile nmake Microsoft (R) Program Maintenance Utility Version 7.10.3077 Copyright (C) Microsoft Corporation. All rights reserved. c:\progra~1\ruby\bin\ruby -e "puts 'EXPORTS', 'Init_foo'" > foo-i386-mswin32.def cl -nologo -MD -Zi -O2b2xg- -G6 -I. -Ic:/progra~1/ruby/lib/ruby/1.8/i386-mswin32 -Ic:/progra~1/ruby/lib/ruby/1.8/i386-mswin32 -I. -c -Tcfoo.c foo.c cl -nologo -LD -Fefoo.so foo.obj msvcrt-ruby18.lib oldnames.lib user32.lib advapi32.lib wsock32.lib -link -incremental:no -debug -opt:ref -opt:icf -dll -libpath:"c:/progra~1/ruby/lib" -def:foo-i386-mswin32.def -implib:foo-i386-mswin32.lib -pdb:foo-i386-mswin32.pdb Creating library foo-i386-mswin32.lib and object foo-i386-mswin32.exp foo.obj : error LNK2019: unresolved external symbol _nonexistent referenced in function _Init_foo foo.so : fatal error LNK1120: 1 unresolved externals NMAKE : fatal error U1077: 'cl' : return code '0x2' Stop. ./lib/rubygems/installer.rb:586:in `make' ./lib/rubygems/installer.rb:581:in `make' ./lib/rubygems/installer.rb:627:in `build' ./test/test_gem_ext_ext_conf_builder.rb:63:in `test_class_build_extconf_bad' ./test/test_gem_ext_ext_conf_builder.rb:62:in `test_class_build_extconf_bad' 4) Failure: test_class_build_fail(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:46]: exception expected but none was thrown. 5) Failure: test_source_index_hash(TestRemoteInstaller) [./test/test_remote_installer.rb:87]: <1> expected but was <6>. 278 tests, 984 assertions, 3 failures, 2 errors rake aborted! Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib "c:/PRO...] (See full trace by running task with --trace) -- Cerberus 0.3.2, http://cerberus.rubyforge.org/ From anatol.pomozov at gmail.com Wed Dec 20 05:01:55 2006 From: anatol.pomozov at gmail.com (Cerberus) Date: Wed, 20 Dec 2006 13:01:55 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1149) Message-ID: <45890a136e5a0_b2019b04a0154@ubuntu.tmail> ------------------------------------------------------------------------ r1149 | drbrain | 2006-12-20 12:57:32 +0300 | 3 lines Changed paths: M /trunk/test/test_gem_ext_configure_builder.rb M /trunk/test/test_gem_ext_ext_conf_builder.rb Remove test_class_build_extconf_bad, cross-compiler behavior is not constant. PLATFORM not VERSION, you idiot. ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/rake/rake_test_loader Started ..............................................................................................F.......................................................................................F.............................................................................................. Finished in 56.359 seconds. 1) Failure: test_class_build_fail(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:46]: exception expected but none was thrown. 2) Failure: test_source_index_hash(TestRemoteInstaller) [./test/test_remote_installer.rb:87]: <1> expected but was <6>. 277 tests, 982 assertions, 2 failures, 0 errors rake aborted! Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib "c:/PRO...] (See full trace by running task with --trace) -- Cerberus 0.3.2, http://cerberus.rubyforge.org/ From drbrain at segment7.net Wed Dec 20 05:15:12 2006 From: drbrain at segment7.net (Eric Hodel) Date: Wed, 20 Dec 2006 02:15:12 -0800 Subject: [Rubygems-developers] [rubygems] Build still broken (#1149) In-Reply-To: <45890a136e5a0_b2019b04a0154@ubuntu.tmail> References: <45890a136e5a0_b2019b04a0154@ubuntu.tmail> Message-ID: <1C15B84E-FA40-4DFF-AB1D-BCC300FA9166@segment7.net> On Dec 20, 2006, at 02:01, Cerberus wrote: > ---------------------------------------------------------------------- > -- > r1149 | drbrain | 2006-12-20 12:57:32 +0300 | 3 lines > Changed paths: > M /trunk/test/test_gem_ext_configure_builder.rb > M /trunk/test/test_gem_ext_ext_conf_builder.rb > > Remove test_class_build_extconf_bad, cross-compiler behavior is not > constant. > PLATFORM not VERSION, you idiot. > > ---------------------------------------------------------------------- > -- > > (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) > Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/ > rake/rake_test_loader > Started > ...................................................................... > ........................F............................................. > ..........................................F........................... > ................................................................... > Finished in 56.359 seconds. > > 1) Failure: > test_class_build_fail(TestGemExtRakeBuilder) [./test/ > test_gem_ext_rake_builder.rb:46]: > exception expected but none was thrown. Why doesn't a rake file like: task :extension do abort 'fail' end cause a non-zero exit on win32? > 2) Failure: > test_source_index_hash(TestRemoteInstaller) [./test/ > test_remote_installer.rb:87]: > <1> expected but was > <6>. I have no clue about this one. My guess is a file isn't getting cleaned up correctly. Can you look around Anatol? > 277 tests, 982 assertions, 2 failures, 0 errors > rake aborted! > Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib > "c:/PRO...] > > (See full trace by running task with --trace) > > -- > Cerberus 0.3.2, http://cerberus.rubyforge.org/ > _______________________________________________ > Rubygems-developers mailing list > Rubygems-developers at rubyforge.org > http://rubyforge.org/mailman/listinfo/rubygems-developers -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From drbrain at segment7.net Wed Dec 20 05:17:36 2006 From: drbrain at segment7.net (Eric Hodel) Date: Wed, 20 Dec 2006 02:17:36 -0800 Subject: [Rubygems-developers] Release time? Message-ID: <6652427C-4A3F-44A3-86FC-F5A4A8413896@segment7.net> There's one open bug that should be easy to fix. There's one feature request that could be fixed in a reasonable amount of time. (The other two involve tattle). I think that means its time to release. PS: I have selfish reasons for this release. -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From halostatue at gmail.com Wed Dec 20 08:14:35 2006 From: halostatue at gmail.com (Austin Ziegler) Date: Wed, 20 Dec 2006 08:14:35 -0500 Subject: [Rubygems-developers] Release time? In-Reply-To: <6652427C-4A3F-44A3-86FC-F5A4A8413896@segment7.net> References: <6652427C-4A3F-44A3-86FC-F5A4A8413896@segment7.net> Message-ID: <9e7db9110612200514t7ad6f98cqbd8c7905b3851914@mail.gmail.com> On 12/20/06, Eric Hodel wrote: > There's one open bug that should be easy to fix. > > There's one feature request that could be fixed in a reasonable > amount of time. (The other two involve tattle). > > I think that means its time to release. > > PS: I have selfish reasons for this release. Can we may be get #activate_gem created for this release? Without this, we can't start encouraging people to get away from #require_gem and all that means. -austin -- Austin Ziegler * halostatue at gmail.com * http://www.halostatue.ca/ * austin at halostatue.ca * http://www.halostatue.ca/feed/ * austin at zieglers.ca From jim.weirich at gmail.com Wed Dec 20 08:38:25 2006 From: jim.weirich at gmail.com (Jim Weirich) Date: Wed, 20 Dec 2006 08:38:25 -0500 Subject: [Rubygems-developers] Release time? In-Reply-To: <9e7db9110612200514t7ad6f98cqbd8c7905b3851914@mail.gmail.com> References: <6652427C-4A3F-44A3-86FC-F5A4A8413896@segment7.net> <9e7db9110612200514t7ad6f98cqbd8c7905b3851914@mail.gmail.com> Message-ID: On 12/20/06, Austin Ziegler wrote: > Can we may be get #activate_gem created for this release? Without > this, we can't start encouraging people to get away from #require_gem > and all that means. That's been in SVN for some time. Except that it is just 'gem' rather than activate_gem. I.e. just put gem 'gemname' in your code. -- -- -- Jim Weirich jim at weirichhouse.org http://onestepback.org ----------------------------------------------------------------- "Beware of bugs in the above code; I have only proved it correct, not tried it." -- Donald Knuth (in a memo to Peter van Emde Boas) From halostatue at gmail.com Wed Dec 20 11:50:31 2006 From: halostatue at gmail.com (Austin Ziegler) Date: Wed, 20 Dec 2006 11:50:31 -0500 Subject: [Rubygems-developers] Release time? In-Reply-To: References: <6652427C-4A3F-44A3-86FC-F5A4A8413896@segment7.net> <9e7db9110612200514t7ad6f98cqbd8c7905b3851914@mail.gmail.com> Message-ID: <9e7db9110612200850t1a5ed042m2b724fd7d395bba0@mail.gmail.com> On 12/20/06, Jim Weirich wrote: > On 12/20/06, Austin Ziegler wrote: > > Can we may be get #activate_gem created for this release? Without > > this, we can't start encouraging people to get away from #require_gem > > and all that means. > That's been in SVN for some time. Except that it is just 'gem' rather > than activate_gem. I.e. just put > gem 'gemname' > > in your code. Then, when this release is done, let's trumpet this from the rooftops -- I was using soap4r and it needs "require_gem" and is documented as such to use the SOAP gem. -austin -- Austin Ziegler * halostatue at gmail.com * http://www.halostatue.ca/ * austin at halostatue.ca * http://www.halostatue.ca/feed/ * austin at zieglers.ca From drbrain at segment7.net Wed Dec 20 13:06:20 2006 From: drbrain at segment7.net (Eric Hodel) Date: Wed, 20 Dec 2006 10:06:20 -0800 Subject: [Rubygems-developers] Release time? In-Reply-To: References: <6652427C-4A3F-44A3-86FC-F5A4A8413896@segment7.net> <9e7db9110612200514t7ad6f98cqbd8c7905b3851914@mail.gmail.com> Message-ID: On Dec 20, 2006, at 05:38, Jim Weirich wrote: > On 12/20/06, Austin Ziegler wrote: >> Can we may be get #activate_gem created for this release? Without >> this, we can't start encouraging people to get away from #require_gem >> and all that means. > > That's been in SVN for some time. Except that it is just 'gem' rather > than activate_gem. I.e. just put > > gem 'gemname' > > in your code. Should we add a deprecation warning to require_gem then? -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From chad at chadfowler.com Wed Dec 20 13:49:17 2006 From: chad at chadfowler.com (Chad Fowler) Date: Wed, 20 Dec 2006 13:49:17 -0500 Subject: [Rubygems-developers] Release time? In-Reply-To: References: <6652427C-4A3F-44A3-86FC-F5A4A8413896@segment7.net> <9e7db9110612200514t7ad6f98cqbd8c7905b3851914@mail.gmail.com> Message-ID: On 12/20/06, Eric Hodel wrote: > On Dec 20, 2006, at 05:38, Jim Weirich wrote: > > On 12/20/06, Austin Ziegler wrote: > >> Can we may be get #activate_gem created for this release? Without > >> this, we can't start encouraging people to get away from #require_gem > >> and all that means. > > > > That's been in SVN for some time. Except that it is just 'gem' rather > > than activate_gem. I.e. just put > > > > gem 'gemname' > > > > in your code. > > Should we add a deprecation warning to require_gem then? > I vote yes. From drbrain at segment7.net Wed Dec 20 14:23:41 2006 From: drbrain at segment7.net (Eric Hodel) Date: Wed, 20 Dec 2006 11:23:41 -0800 Subject: [Rubygems-developers] Release time? In-Reply-To: References: <6652427C-4A3F-44A3-86FC-F5A4A8413896@segment7.net> <9e7db9110612200514t7ad6f98cqbd8c7905b3851914@mail.gmail.com> Message-ID: <68AF4EF8-82A4-4A16-9938-E5B7369DB9E4@segment7.net> On Dec 20, 2006, at 10:49, Chad Fowler wrote: > On 12/20/06, Eric Hodel wrote: >> On Dec 20, 2006, at 05:38, Jim Weirich wrote: >>> On 12/20/06, Austin Ziegler wrote: >>>> Can we may be get #activate_gem created for this release? Without >>>> this, we can't start encouraging people to get away from >>>> #require_gem >>>> and all that means. >>> >>> That's been in SVN for some time. Except that it is just 'gem' >>> rather >>> than activate_gem. I.e. just put >>> >>> gem 'gemname' >>> >>> in your code. >> >> Should we add a deprecation warning to require_gem then? >> > > I vote yes. Since there was already a comment marking require_gem as deprecated, this is done. -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From drbrain at segment7.net Thu Dec 21 02:11:26 2006 From: drbrain at segment7.net (Eric Hodel) Date: Wed, 20 Dec 2006 23:11:26 -0800 Subject: [Rubygems-developers] Release time! (Was: Release time?) In-Reply-To: <6652427C-4A3F-44A3-86FC-F5A4A8413896@segment7.net> References: <6652427C-4A3F-44A3-86FC-F5A4A8413896@segment7.net> Message-ID: On Dec 20, 2006, at 02:17, Eric Hodel wrote: > There's one open bug that should be easy to fix. > > There's one feature request that could be fixed in a reasonable > amount of time. (The other two involve tattle). > > I think that means its time to release. There are zero bugs open. There are zero feature requests open. The two Cerberus test failures appear to be platform-specific and test-specific. I think that means its time to push a beta... Can I do that myself, or do I need Jim or Chad to do it? -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From drbrain at segment7.net Thu Dec 21 02:38:59 2006 From: drbrain at segment7.net (Eric Hodel) Date: Wed, 20 Dec 2006 23:38:59 -0800 Subject: [Rubygems-developers] Release time! (Was: Release time?) In-Reply-To: References: <6652427C-4A3F-44A3-86FC-F5A4A8413896@segment7.net> Message-ID: On Dec 20, 2006, at 23:11, Eric Hodel wrote: > On Dec 20, 2006, at 02:17, Eric Hodel wrote: >> There's one open bug that should be easy to fix. >> >> There's one feature request that could be fixed in a reasonable >> amount of time. (The other two involve tattle). >> >> I think that means its time to release. > > There are zero bugs open. > > There are zero feature requests open. * zero easily resolvable feature requests. > The two Cerberus test failures appear to be platform-specific and > test-specific. > > I think that means its time to push a beta... Can I do that myself, > or do I need Jim or Chad to do it? -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From anatol.pomozov at gmail.com Thu Dec 21 06:05:19 2006 From: anatol.pomozov at gmail.com (Anatol Pomozov) Date: Thu, 21 Dec 2006 14:05:19 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1149) In-Reply-To: <45890a136e5a0_b2019b04a0154@ubuntu.tmail> References: <45890a136e5a0_b2019b04a0154@ubuntu.tmail> Message-ID: <3665a1a00612210305r6e65c9b4h3059f3bdec36b1b7@mail.gmail.com> Hi, Eric. First of all thanks for your RubyGems improvements. You and other developers are doing great job. On 12/20/06, Cerberus wrote: > > ------------------------------------------------------------------------ > r1149 | drbrain | 2006-12-20 12:57:32 +0300 | 3 lines > Changed paths: > M /trunk/test/test_gem_ext_configure_builder.rb > M /trunk/test/test_gem_ext_ext_conf_builder.rb > > Remove test_class_build_extconf_bad, cross-compiler behavior is not > constant. > PLATFORM not VERSION, you idiot. > > ------------------------------------------------------------------------ > > (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) > Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1 > /lib/rake/rake_test_loader > Started > ..............................................................................................F.......................................................................................F.............................................................................................. > > Finished in 56.359 seconds. > > 1) Failure: > test_class_build_fail(TestGemExtRakeBuilder) > [./test/test_gem_ext_rake_builder.rb:46]: > exception expected but none was thrown. As I see RubyGems could not correctly define that Rake finished with error This is what RubyGems do unless $?.exitstatus.zero? then raise Gem::InstallError, "#{class_name} failed:\n\n#{results.join"\n"}" end But Rake finishes with return code == 0 on windows even if some error present. I dunno why, better to ask Rake developers. Anyway this is how I define that rake finished with error in Cerberus. def successful? $?.exitstatus == 0 and not @output.include?("rake aborted!") end I dont think that it is very elegant (what if user would print this message from Rake code) but it works both on Windoze and *nix. Any comments on this subject? 2) Failure: > test_source_index_hash(TestRemoteInstaller) > [./test/test_remote_installer.rb:87]: > <1> expected but was > <6>. puts source_hash gives me following output http://gems.rubyforge.org# http://onestepback.org/betagems# code.whytheluckystiff.net# http://gems.example.com#http://gems.rubyonrails.org# looks like test uses also my own gem-repositories (not only http://gems.example.com) 277 tests, 982 assertions, 2 failures, 0 errors > rake aborted! > Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib > "c:/PRO...] > > (See full trace by running task with --trace) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20061221/810f3611/attachment.html From chad at chadfowler.com Thu Dec 21 10:31:36 2006 From: chad at chadfowler.com (Chad Fowler) Date: Thu, 21 Dec 2006 10:31:36 -0500 Subject: [Rubygems-developers] Release time! (Was: Release time?) In-Reply-To: References: <6652427C-4A3F-44A3-86FC-F5A4A8413896@segment7.net> Message-ID: On 12/21/06, Eric Hodel wrote: > > On Dec 20, 2006, at 02:17, Eric Hodel wrote: > > > There's one open bug that should be easy to fix. > > > > There's one feature request that could be fixed in a reasonable > > amount of time. (The other two involve tattle). > > > > I think that means its time to release. > > There are zero bugs open. > > There are zero feature requests open. > > The two Cerberus test failures appear to be platform-specific and > test-specific. > > I think that means its time to push a beta... Can I do that myself, > or do I need Jim or Chad to do it? > Jim has historically done the beta releases of RubyGems on his own server, which doesn't appear to be responding right now. He has a job from which he can't use the internet all day, so I propose we wait until he gets home and has a chance to respond and then if it will be a pain to deploy to the usual place, we do something different. I would prefer to leave the betas outside of the main rubygems repo, but we could always throw them somewhere else temporariy if need be. Chad From anatol.pomozov at gmail.com Thu Dec 21 16:01:37 2006 From: anatol.pomozov at gmail.com (Cerberus) Date: Fri, 22 Dec 2006 00:01:37 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1157) Message-ID: <458af631762a0_c4819b04b415d@ubuntu.tmail> ------------------------------------------------------------------------ r1157 | drbrain | 2006-12-21 11:42:32 +0300 | 1 line Changed paths: M /trunk/Rakefile Switch to running all tests by default ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) Required Gem 'Session' missing. We can attempt to install from the RubyGems Distribution, but installation may require admin privileges on your system. Install now from RubyGems distribution? [Yn]Test cancelled...quitting rake aborted! Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib "c:/PRO...] (See full trace by running task with --trace) -- Cerberus 0.3.2, http://cerberus.rubyforge.org/ From jim at weirichhouse.org Thu Dec 21 16:30:53 2006 From: jim at weirichhouse.org (jim at weirichhouse.org) Date: Thu, 21 Dec 2006 21:30:53 -0000 (UTC) Subject: [Rubygems-developers] Release time! (Was: Release time?) In-Reply-To: References: <6652427C-4A3F-44A3-86FC-F5A4A8413896@segment7.net> Message-ID: <14360.198.177.158.254.1166736653.squirrel@onestepback.org> > On 12/21/06, Eric Hodel wrote: >> I think that means its time to push a beta... Can I do that myself, >> or do I need Jim or Chad to do it? >> > > Jim has historically done the beta releases of RubyGems on his own > server, which doesn't appear to be responding right now. He has a job > from which he can't use the internet all day, so I propose we wait > until he gets home and has a chance to respond and then if it will be > a pain to deploy to the usual place, we do something different. I > would prefer to leave the betas outside of the main rubygems repo, but > we could always throw them somewhere else temporariy if need be. Good news! My box is back up, so putting the beta gem in the normal place won't be a problem. Eric, if you wish to create the beta, that is not a problem. The biggest issue with betas is to make sure the 4 digit version number is properly incremented between different betas. Make sure the change log is up to date ... we don't need it for the beta, but I use that to glean diffs from previous versions for the release notes on a full release. Thanks for all your hard work. -- Jim Weirich From drbrain at segment7.net Fri Dec 22 05:04:40 2006 From: drbrain at segment7.net (Eric Hodel) Date: Fri, 22 Dec 2006 02:04:40 -0800 Subject: [Rubygems-developers] [rubygems] Build still broken (#1157) In-Reply-To: <458af631762a0_c4819b04b415d@ubuntu.tmail> References: <458af631762a0_c4819b04b415d@ubuntu.tmail> Message-ID: <9AA05E5C-2E5F-44E7-9C04-EABF4E651AD1@segment7.net> On Dec 21, 2006, at 13:01, Cerberus wrote: > ---------------------------------------------------------------------- > -- > r1157 | drbrain | 2006-12-21 11:42:32 +0300 | 1 line > Changed paths: > M /trunk/Rakefile > > Switch to running all tests by default > ---------------------------------------------------------------------- > -- > > (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) > > Required Gem 'Session' missing. > We can attempt to install from the RubyGems Distribution, > but installation may require admin privileges on your system. > > Install now from RubyGems distribution? [Yn]Test cancelled...quitting > rake aborted! > Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib > "c:/PRO...] I added functional tests to the default test run, but this has an added dependency of the session gem. Some tests were failing because they didn't get run. -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From drbrain at segment7.net Fri Dec 22 05:22:47 2006 From: drbrain at segment7.net (Eric Hodel) Date: Fri, 22 Dec 2006 02:22:47 -0800 Subject: [Rubygems-developers] [rubygems] Build still broken (#1149) In-Reply-To: <3665a1a00612210305r6e65c9b4h3059f3bdec36b1b7@mail.gmail.com> References: <45890a136e5a0_b2019b04a0154@ubuntu.tmail> <3665a1a00612210305r6e65c9b4h3059f3bdec36b1b7@mail.gmail.com> Message-ID: On Dec 21, 2006, at 03:05, Anatol Pomozov wrote: > On 12/20/06, Cerberus < anatol.pomozov at gmail.com> wrote: > 1) Failure: > test_class_build_fail(TestGemExtRakeBuilder) [./test/ > test_gem_ext_rake_builder.rb:46]: > exception expected but none was thrown. > > As I see RubyGems could not correctly define that Rake finished > with error > This is what RubyGems do > unless $?.exitstatus.zero? then > raise Gem::InstallError, "#{class_name} failed:\n\n# > { results.join "\n"}" > end > > But Rake finishes with return code == 0 on windows even if some > error present. I dunno why, better to ask Rake developers. What does this script print? system 'ruby -e "raise"' p $? I get: $ ruby -ve test.rb ruby 1.8.5 (2006-12-04 patchlevel 2) [i686-darwin8.8.2] -e:1: unhandled exception # $ > Anyway this is how I define that rake finished with error in Cerberus. > > def successful? > $?.exitstatus == 0 and not @output.include?("rake aborted!") > end > > I dont think that it is very elegant (what if user would print this > message from Rake code) but it works both on Windoze and *nix. > > Any comments on this subject? > > 2) Failure: > test_source_index_hash(TestRemoteInstaller) [./test/ > test_remote_installer.rb:87]: > <1> expected but was > <6>. > > puts source_hash > > gives me following output > > http://gems.rubyforge.org# http:// > onestepback.org/betagems# > code.whytheluckystiff.net#http:// > gems.example.com# http:// > gems.rubyonrails.org# > > looks like test uses also my own gem-repositories (not only http:// > gems.example.com ) I think I fixed this one, I was reseting the carefully set up gem paths back to the system ones. If you install the session plugin we'll see... -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From anatol.pomozov at gmail.com Fri Dec 22 05:32:07 2006 From: anatol.pomozov at gmail.com (Anatol Pomozov) Date: Fri, 22 Dec 2006 13:32:07 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1157) In-Reply-To: <9AA05E5C-2E5F-44E7-9C04-EABF4E651AD1@segment7.net> References: <458af631762a0_c4819b04b415d@ubuntu.tmail> <9AA05E5C-2E5F-44E7-9C04-EABF4E651AD1@segment7.net> Message-ID: <3665a1a00612220232l1bcadbe9hc0a2a48a3a568ba9@mail.gmail.com> Hi, Eric. But 'session' gem does not work under windows. Tests have a lot of errors like this one 22) Error: test_generate_old_style_index(TestGenerateYamlIndex): NotImplementedError: the fork() function is unimplemented on this machine C:/Program Files/ruby/lib/ruby/gems/1.8/gems/session-2.4.0/lib/session.rb:293:in `fork' C:/Program Files/ruby/lib/ruby/gems/1.8/gems/session-2.4.0/lib/session.rb:293:in `__fork' C:/Program Files/ruby/lib/ruby/gems/1.8/gems/session-2.4.0/lib/session.rb:254:in `__popen3' C:/Program Files/ruby/lib/ruby/gems/1.8/gems/session-2.4.0/lib/session.rb:205:in `initialize' ./test/functional_generate_yaml_index.rb:82:in `generate_yaml_index' ./test/functional_generate_yaml_index.rb:42:in `test_generate_old_style_index' On 12/22/06, Eric Hodel wrote: > > On Dec 21, 2006, at 13:01, Cerberus wrote: > > > ---------------------------------------------------------------------- > > -- > > r1157 | drbrain | 2006-12-21 11:42:32 +0300 | 1 line > > Changed paths: > > M /trunk/Rakefile > > > > Switch to running all tests by default > > ---------------------------------------------------------------------- > > -- > > > > (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) > > > > Required Gem 'Session' missing. > > We can attempt to install from the RubyGems Distribution, > > but installation may require admin privileges on your system. > > > > Install now from RubyGems distribution? [Yn]Test cancelled...quitting > > rake aborted! > > Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib > > "c:/PRO...] > > I added functional tests to the default test run, but this has an > added dependency of the session gem. Some tests were failing because > they didn't get run. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20061222/5e83ca35/attachment.html From anatol.pomozov at gmail.com Fri Dec 22 05:39:47 2006 From: anatol.pomozov at gmail.com (Anatol Pomozov) Date: Fri, 22 Dec 2006 13:39:47 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1149) In-Reply-To: References: <45890a136e5a0_b2019b04a0154@ubuntu.tmail> <3665a1a00612210305r6e65c9b4h3059f3bdec36b1b7@mail.gmail.com> Message-ID: <3665a1a00612220239x3c3eb28bv31613d47eac438a2@mail.gmail.com> Hi Eric. On 12/22/06, Eric Hodel wrote: > > On Dec 21, 2006, at 03:05, Anatol Pomozov wrote: > > On 12/20/06, Cerberus < anatol.pomozov at gmail.com> wrote: > > 1) Failure: > > test_class_build_fail(TestGemExtRakeBuilder) [./test/ > > test_gem_ext_rake_builder.rb:46]: > > exception expected but none was thrown. > > > > As I see RubyGems could not correctly define that Rake finished > > with error > > This is what RubyGems do > > unless $?.exitstatus.zero? then > > raise Gem::InstallError, "#{class_name} failed:\n\n# > > { results.join "\n"}" > > end > > > > But Rake finishes with return code == 0 on windows even if some > > error present. I dunno why, better to ask Rake developers. > > What does this script print? > > system 'ruby -e "raise"' > p $? > > I get: > > $ ruby -ve test.rb > ruby 1.8.5 (2006-12-04 patchlevel 2) [i686-darwin8.8.2] > -e:1: unhandled exception > # > $ I get following output -e:1: unhandled exception # But if I create file Rakefile with following content (as in test) task :extension do abort 'fail' end and try to invoke it (as in RubyGems code) `rake.bat extension` p $? I would get following output C:\work\opensource\ruby\rubygems\~my>rake-error-outstatus.rb fail # So as you see `rake` returns us zero exitcode even if we have failure during Rake execution. It is Windows. What do you have under *nix? > Anyway this is how I define that rake finished with error in Cerberus. > > > > def successful? > > $?.exitstatus == 0 and not @output.include?("rake aborted!") > > end > > > > I dont think that it is very elegant (what if user would print this > > message from Rake code) but it works both on Windoze and *nix. > > > > Any comments on this subject? > > > > 2) Failure: > > test_source_index_hash(TestRemoteInstaller) [./test/ > > test_remote_installer.rb:87]: > > <1> expected but was > > <6>. > > > > puts source_hash > > > > gives me following output > > > > http://gems.rubyforge.org# http:// > > onestepback.org/betagems# > > code.whytheluckystiff.net#http:// > > gems.example.com# http:// > > gems.rubyonrails.org# > > > > looks like test uses also my own gem-repositories (not only http:// > > gems.example.com ) > > I think I fixed this one, I was reseting the carefully set up gem > paths back to the system ones. > > If you install the session plugin we'll see... > -- anatol (http://feeds.feedburner.com/apomozov-eng) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20061222/30f62650/attachment.html From anatol.pomozov at gmail.com Fri Dec 22 12:21:11 2006 From: anatol.pomozov at gmail.com (Cerberus) Date: Fri, 22 Dec 2006 20:21:11 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1158) Message-ID: <458c1407d59f8_4d019b04b4179@ubuntu.tmail> ------------------------------------------------------------------------ r1158 | drbrain | 2006-12-22 13:22:11 +0300 | 1 line Changed paths: M /trunk/test/test_remote_installer.rb Don't reset our carefully set up gem paths. Might fix Cerberus failure. ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) Required Gem 'Session' missing. We can attempt to install from the RubyGems Distribution, but installation may require admin privileges on your system. Install now from RubyGems distribution? [Yn]Test cancelled...quitting rake aborted! Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib "c:/PRO...] (See full trace by running task with --trace) -- Cerberus 0.3.2, http://cerberus.rubyforge.org/ From drbrain at segment7.net Sat Dec 23 02:28:35 2006 From: drbrain at segment7.net (Eric Hodel) Date: Fri, 22 Dec 2006 23:28:35 -0800 Subject: [Rubygems-developers] [rubygems] Build still broken (#1149) In-Reply-To: <3665a1a00612220239x3c3eb28bv31613d47eac438a2@mail.gmail.com> References: <45890a136e5a0_b2019b04a0154@ubuntu.tmail> <3665a1a00612210305r6e65c9b4h3059f3bdec36b1b7@mail.gmail.com> <3665a1a00612220239x3c3eb28bv31613d47eac438a2@mail.gmail.com> Message-ID: <8A45A299-6D68-420C-8BDC-F0CCB750D1FB@segment7.net> On Dec 22, 2006, at 02:39, Anatol Pomozov wrote: > On 12/22/06, Eric Hodel wrote: > On Dec 21, 2006, at 03:05, Anatol Pomozov wrote: > > On 12/20/06, Cerberus < anatol.pomozov at gmail.com> wrote: > > 1) Failure: > > test_class_build_fail(TestGemExtRakeBuilder) [./test/ > > test_gem_ext_rake_builder.rb:46]: > > exception expected but none was thrown. > > > > As I see RubyGems could not correctly define that Rake finished > > with error > > This is what RubyGems do > > unless $?.exitstatus.zero? then > > raise Gem::InstallError, "#{class_name} failed:\n\n# > > { results.join "\n"}" > > end > > > > But Rake finishes with return code == 0 on windows even if some > > error present. I dunno why, better to ask Rake developers. > > What does this script print? > > system 'ruby -e "raise"' > p $? > > I get: > > $ ruby -ve test.rb > ruby 1.8.5 (2006-12-04 patchlevel 2) [ i686-darwin8.8.2] > -e:1: unhandled exception > # > $ > I get following output > -e:1: unhandled exception > # > > But if I create file Rakefile with following content (as in test) > task :extension do abort 'fail' end > > and try to invoke it (as in RubyGems code) > > `rake.bat extension` > p $? > > I would get following output > > C:\work\opensource\ruby\rubygems\~my>rake-error-outstatus.rb > fail > # > > So as you see `rake` returns us zero exitcode even if we have > failure during Rake execution. > > It is Windows. What do you have under *nix? $ ruby -e '`rake extension`; p $?' fail # > > Anyway this is how I define that rake finished with error in > Cerberus. > > > > def successful? > > $?.exitstatus == 0 and not @output.include?("rake aborted!") > > end > > > > I dont think that it is very elegant (what if user would print this > > message from Rake code) but it works both on Windoze and *nix. > > > > Any comments on this subject? > -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From drbrain at segment7.net Sat Dec 23 02:29:18 2006 From: drbrain at segment7.net (Eric Hodel) Date: Fri, 22 Dec 2006 23:29:18 -0800 Subject: [Rubygems-developers] [rubygems] Build still broken (#1157) In-Reply-To: <3665a1a00612220232l1bcadbe9hc0a2a48a3a568ba9@mail.gmail.com> References: <458af631762a0_c4819b04b415d@ubuntu.tmail> <9AA05E5C-2E5F-44E7-9C04-EABF4E651AD1@segment7.net> <3665a1a00612220232l1bcadbe9hc0a2a48a3a568ba9@mail.gmail.com> Message-ID: On Dec 22, 2006, at 02:32, Anatol Pomozov wrote: > Hi, Eric. > > But 'session' gem does not work under windows. Ok. I cannot think of an easy workaround, so I reverted the change. -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From anatol.pomozov at gmail.com Sat Dec 23 04:04:41 2006 From: anatol.pomozov at gmail.com (Anatol Pomozov) Date: Sat, 23 Dec 2006 12:04:41 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1149) In-Reply-To: <8A45A299-6D68-420C-8BDC-F0CCB750D1FB@segment7.net> References: <45890a136e5a0_b2019b04a0154@ubuntu.tmail> <3665a1a00612210305r6e65c9b4h3059f3bdec36b1b7@mail.gmail.com> <3665a1a00612220239x3c3eb28bv31613d47eac438a2@mail.gmail.com> <8A45A299-6D68-420C-8BDC-F0CCB750D1FB@segment7.net> Message-ID: <3665a1a00612230104j26a35a9cv6d9cee42ddb1a70f@mail.gmail.com> Hi, Eric. And that is the root of the problem. On *nix rake returns 1 on Windows - 0 exit code. Is here any Rake guys who could explain this strange behaviour? I've added patch that cures tests. There you could find another issue. See test/test_gem_ext_rake_builder.rb windows redirector adds all STDERR output to the end of STDOUT output. So order of "fail" words changed. On 12/23/06, Eric Hodel wrote: > > On Dec 22, 2006, at 02:39, Anatol Pomozov wrote: > > On 12/22/06, Eric Hodel wrote: > > On Dec 21, 2006, at 03:05, Anatol Pomozov wrote: > > > On 12/20/06, Cerberus < anatol.pomozov at gmail.com> wrote: > > > 1) Failure: > > > test_class_build_fail(TestGemExtRakeBuilder) [./test/ > > > test_gem_ext_rake_builder.rb:46]: > > > exception expected but none was thrown. > > > > > > As I see RubyGems could not correctly define that Rake finished > > > with error > > > This is what RubyGems do > > > unless $?.exitstatus.zero? then > > > raise Gem::InstallError, "#{class_name} failed:\n\n# > > > { results.join "\n"}" > > > end > > > > > > But Rake finishes with return code == 0 on windows even if some > > > error present. I dunno why, better to ask Rake developers. > > > > What does this script print? > > > > system 'ruby -e "raise"' > > p $? > > > > I get: > > > > $ ruby -ve test.rb > > ruby 1.8.5 (2006-12-04 patchlevel 2) [ i686-darwin8.8.2] > > -e:1: unhandled exception > > # > > $ > > I get following output > > -e:1: unhandled exception > > # > > > > But if I create file Rakefile with following content (as in test) > > task :extension do abort 'fail' end > > > > and try to invoke it (as in RubyGems code) > > > > `rake.bat extension` > > p $? > > > > I would get following output > > > > C:\work\opensource\ruby\rubygems\~my>rake-error-outstatus.rb > > fail > > # > > > > So as you see `rake` returns us zero exitcode even if we have > > failure during Rake execution. > > > > It is Windows. What do you have under *nix? > > $ ruby -e '`rake extension`; p $?' > fail > # > > > > Anyway this is how I define that rake finished with error in > > Cerberus. > > > > > > def successful? > > > $?.exitstatus == 0 and not @output.include?("rake aborted!") > > > end > > > > > > I dont think that it is very elegant (what if user would print this > > > message from Rake code) but it works both on Windoze and *nix. > > > > > > Any comments on this subject? > -- anatol (http://feeds.feedburner.com/apomozov-eng) -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20061223/f1ec7ace/attachment.html -------------- next part -------------- A non-text attachment was scrubbed... Name: fix_test.diff Type: text/x-diff Size: 1020 bytes Desc: not available Url : http://rubyforge.org/pipermail/rubygems-developers/attachments/20061223/f1ec7ace/attachment.bin From anatol.pomozov at gmail.com Sat Dec 23 04:21:42 2006 From: anatol.pomozov at gmail.com (Cerberus) Date: Sat, 23 Dec 2006 12:21:42 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1161) Message-ID: <458cf5264b320_f2019bdee815d@ubuntu.tmail> ------------------------------------------------------------------------ r1161 | drbrain | 2006-12-23 11:44:27 +0300 | 1 line Changed paths: M /trunk/lib/rubygems/command.rb M /trunk/lib/rubygems/gem_commands.rb Make gem help output fit in 80 columns ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/rake/rake_test_loader Started ..............................................................................................F...................................................................................................................................................................................... Finished in 44.469 seconds. 1) Failure: test_class_build_fail(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:46]: exception expected but none was thrown. 277 tests, 984 assertions, 1 failures, 0 errors rake aborted! Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib "c:/PRO...] (See full trace by running task with --trace) -- Cerberus 0.3.2, http://cerberus.rubyforge.org/ From drbrain at segment7.net Sat Dec 23 21:53:26 2006 From: drbrain at segment7.net (Eric Hodel) Date: Sat, 23 Dec 2006 18:53:26 -0800 Subject: [Rubygems-developers] [rubygems] Build still broken (#1149) In-Reply-To: <3665a1a00612230104j26a35a9cv6d9cee42ddb1a70f@mail.gmail.com> References: <45890a136e5a0_b2019b04a0154@ubuntu.tmail> <3665a1a00612210305r6e65c9b4h3059f3bdec36b1b7@mail.gmail.com> <3665a1a00612220239x3c3eb28bv31613d47eac438a2@mail.gmail.com> <8A45A299-6D68-420C-8BDC-F0CCB750D1FB@segment7.net> <3665a1a00612230104j26a35a9cv6d9cee42ddb1a70f@mail.gmail.com> Message-ID: On Dec 23, 2006, at 01:04, Anatol Pomozov wrote: > And that is the root of the problem. On *nix rake returns 1 on > Windows - 0 exit code. Is here any Rake guys who could explain this > strange behaviour? Jim? I'll also play with this on my parents' computer. You are running this under cmd.exe, correct? > I've added patch that cures tests. > > There you could find another issue. See test/ > test_gem_ext_rake_builder.rb > windows redirector adds all STDERR output to the end of STDOUT > output. So order of "fail" words changed. Well, the first three lines are the only really important ones, so I didn't bother to test the rest of them. -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From anatol.pomozov at gmail.com Sun Dec 24 01:02:58 2006 From: anatol.pomozov at gmail.com (Cerberus) Date: Sun, 24 Dec 2006 09:02:58 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1170) Message-ID: <458e1812aba18_3d81bff94a1cb@ubuntu.tmail> ------------------------------------------------------------------------ r1170 | drbrain | 2006-12-24 06:10:27 +0300 | 1 line Changed paths: M /trunk/ChangeLog Bumped version to 0.9.0.8 ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/rake/rake_test_loader Started ..............................................................................................F...................................................................................................................................................................................... Finished in 70.891 seconds. 1) Failure: test_class_build_fail(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:45]: exception expected but none was thrown. 277 tests, 984 assertions, 1 failures, 0 errors rake aborted! Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib "c:/PRO...] (See full trace by running task with --trace) -- Cerberus 0.3.3, http://cerberus.rubyforge.org/ From jim at weirichhouse.org Mon Dec 25 14:59:46 2006 From: jim at weirichhouse.org (Jim Weirich) Date: Mon, 25 Dec 2006 14:59:46 -0500 Subject: [Rubygems-developers] [rubygems] Build still broken (#1149) In-Reply-To: <3665a1a00612210305r6e65c9b4h3059f3bdec36b1b7@mail.gmail.com> References: <45890a136e5a0_b2019b04a0154@ubuntu.tmail> <3665a1a00612210305r6e65c9b4h3059f3bdec36b1b7@mail.gmail.com> Message-ID: <45902DB2.6070405@weirichhouse.org> Anatol Pomozov wrote: > But Rake finishes with return code == 0 on windows even if some error > present. I dunno why, better to ask Rake developers. As far as I can tell, Rake is properly returning the exit status code. However on Windows, rake is invoked via a BAT file setup by RubyGems (on Unix systems, a shell script is used). I suspect the exit code is not being properly propagated through the BAT script. Unfortunately, my BAT programming skills are exactly zero. Can someone verify this and suggest a better BAT script for gems that will propagate exit status. Thanks. -- -- Jim Weirich jim at weirichhouse.org http://onestepback.org -- In theory, practice and theory are the same. -- In practice, they are different. From anatol.pomozov at gmail.com Mon Dec 25 16:02:07 2006 From: anatol.pomozov at gmail.com (Cerberus) Date: Tue, 26 Dec 2006 00:02:07 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1174) Message-ID: <45903c4f8d1d0_d001bffa44168@ubuntu.tmail> ------------------------------------------------------------------------ r1174 | drbrain | 2006-12-25 14:17:26 +0300 | 1 line Changed paths: M /trunk/ChangeLog Add last two commits ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/rake/rake_test_loader Started ..............................................................................................F......................................................................................................................................................................................... Finished in 50.578 seconds. 1) Failure: test_class_build_fail(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:45]: exception expected but none was thrown. 280 tests, 990 assertions, 1 failures, 0 errors rake aborted! Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib "c:/PRO...] (See full trace by running task with --trace) -- Cerberus 0.3.3, http://cerberus.rubyforge.org/ From drbrain at segment7.net Mon Dec 25 17:05:24 2006 From: drbrain at segment7.net (Eric Hodel) Date: Mon, 25 Dec 2006 14:05:24 -0800 Subject: [Rubygems-developers] [rubygems] Build still broken (#1149) In-Reply-To: <45902DB2.6070405@weirichhouse.org> References: <45890a136e5a0_b2019b04a0154@ubuntu.tmail> <3665a1a00612210305r6e65c9b4h3059f3bdec36b1b7@mail.gmail.com> <45902DB2.6070405@weirichhouse.org> Message-ID: On Dec 25, 2006, at 11:59, Jim Weirich wrote: > Anatol Pomozov wrote: >> But Rake finishes with return code == 0 on windows even if some error >> present. I dunno why, better to ask Rake developers. > > As far as I can tell, Rake is properly returning the exit > status code. However on Windows, rake is invoked via a BAT > file setup by RubyGems (on Unix systems, a shell script is > used). I suspect the exit code is not being properly > propagated through the BAT script. Unfortunately, my BAT > programming skills are exactly zero. Can someone verify this > and suggest a better BAT script for gems that will propagate > exit status. Yep, I've verified this. "c:\ruby\bin\ruby" -x "%~f0" %* goto endofruby #!/bin/ruby abort 'boom' __END__ :endofruby Will return an exit code of 0 instead of 1. I'll ask ruby-talk. -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From drbrain at segment7.net Mon Dec 25 19:36:17 2006 From: drbrain at segment7.net (Eric Hodel) Date: Mon, 25 Dec 2006 16:36:17 -0800 Subject: [Rubygems-developers] [rubygems] Build still broken (#1149) In-Reply-To: References: <45890a136e5a0_b2019b04a0154@ubuntu.tmail> <3665a1a00612210305r6e65c9b4h3059f3bdec36b1b7@mail.gmail.com> <45902DB2.6070405@weirichhouse.org> Message-ID: <010F827A-E3DA-4526-80A0-D919A801B49B@segment7.net> On Dec 25, 2006, at 14:05, Eric Hodel wrote: > On Dec 25, 2006, at 11:59, Jim Weirich wrote: >> Anatol Pomozov wrote: >>> But Rake finishes with return code == 0 on windows even if some >>> error >>> present. I dunno why, better to ask Rake developers. >> >> As far as I can tell, Rake is properly returning the exit >> status code. However on Windows, rake is invoked via a BAT >> file setup by RubyGems (on Unix systems, a shell script is >> used). I suspect the exit code is not being properly >> propagated through the BAT script. Unfortunately, my BAT >> programming skills are exactly zero. Can someone verify this >> and suggest a better BAT script for gems that will propagate >> exit status. > > Yep, I've verified this. > > "c:\ruby\bin\ruby" -x "%~f0" %* > goto endofruby > #!/bin/ruby > abort 'boom' > __END__ > :endofruby > > Will return an exit code of 0 instead of 1. I'll ask ruby-talk. Ok, it seems that this is a problem with the ruby one-click installer generating batch files that don't match rubygems'. -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From anatol.pomozov at gmail.com Mon Dec 25 23:22:54 2006 From: anatol.pomozov at gmail.com (Cerberus) Date: Tue, 26 Dec 2006 07:22:54 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1177) Message-ID: <4590a39e48440_8e41bffa621e8@ubuntu.tmail> ------------------------------------------------------------------------ r1177 | jimweirich | 2006-12-26 05:55:24 +0300 | 3 lines Changed paths: M /trunk/ChangeLog M /trunk/lib/rubygems/custom_require.rb Removed SUFFIX_PATTERN (evidently missed during an earlier cleanup of the list of suffixes). ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/rake/rake_test_loader Started ..............................................................................................F......................................................................................................................................................................................... Finished in 81.11 seconds. 1) Failure: test_class_build_fail(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:45]: exception expected but none was thrown. 280 tests, 990 assertions, 1 failures, 0 errors rake aborted! Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib "c:/PRO...] (See full trace by running task with --trace) -- Cerberus 0.3.3, http://cerberus.rubyforge.org/ From anatol.pomozov at gmail.com Tue Dec 26 13:05:31 2006 From: anatol.pomozov at gmail.com (Cerberus) Date: Tue, 26 Dec 2006 21:05:31 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1179) Message-ID: <4591646ba7b98_e281b242dc1d9@ubuntu.tmail> ------------------------------------------------------------------------ r1179 | drbrain | 2006-12-26 13:46:31 +0300 | 1 line Changed paths: M /trunk/lib/rubygems/package.rb Reindent, content-free ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) require_gem is obsolete. Use gem instead. Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/rake/rake_test_loader Started .............................................................................................FF......................................................................................................................................................................................... Finished in 194.797 seconds. 1) Failure: test_class_build(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:35]: <["rake RUBYARCHDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3748/prefix RUBYLIBDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3748/prefix extension", "(in C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3748/ext)\n"]> expected but was <["rake RUBYARCHDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3748/prefix RUBYLIBDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3748/prefix extension", "(in C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3748/ext)\nrequire_gem is obsolete. Use gem instead.\n"]>. 2) Failure: test_class_build_fail(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:45]: exception expected but none was thrown. 280 tests, 990 assertions, 2 failures, 0 errors rake aborted! Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib "c:/PRO...] (See full trace by running task with --trace) -- Cerberus 0.3.3, http://cerberus.rubyforge.org/ From jim at weirichhouse.org Tue Dec 26 15:03:09 2006 From: jim at weirichhouse.org (Jim Weirich) Date: Tue, 26 Dec 2006 15:03:09 -0500 Subject: [Rubygems-developers] [rubygems] Build still broken (#1179) In-Reply-To: <4591646ba7b98_e281b242dc1d9@ubuntu.tmail> References: <4591646ba7b98_e281b242dc1d9@ubuntu.tmail> Message-ID: <45917FFD.7050707@weirichhouse.org> Cerberus wrote: > 1) Failure: > test_class_build(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:35]: > <["rake RUBYARCHDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3748/prefix RUBYLIBDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3748/prefix extension", > "(in C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3748/ext)\n"]> expected but was > <["rake RUBYARCHDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3748/prefix RUBYLIBDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3748/prefix extension", > "(in C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3748/ext)\nrequire_gem is obsolete. Use gem instead.\n"]>. This one caught me too. You need to fix the calling script that invokes rake, as the 'require_gem' warning is messing up the test output. Does "gem pristine fix this yet? -- -- Jim Weirich jim at weirichhouse.org http://onestepback.org -- In theory, practice and theory are the same. -- In practice, they are different. From anatol.pomozov at gmail.com Tue Dec 26 22:43:12 2006 From: anatol.pomozov at gmail.com (Cerberus) Date: Wed, 27 Dec 2006 06:43:12 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1180) Message-ID: <4591ebd089350_4d81b2433614f@ubuntu.tmail> ------------------------------------------------------------------------ r1180 | drbrain | 2006-12-27 06:39:39 +0300 | 1 line Changed paths: M /trunk/test/test_package.rb Sort, reindent, clean ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) require_gem is obsolete. Use gem instead. Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/rake/rake_test_loader Started .........................................................................FF............................................................................................................................................................................................................. Finished in 50.656 seconds. 1) Failure: test_class_build(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:35]: <["rake RUBYARCHDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_2456/prefix RUBYLIBDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_2456/prefix extension", "(in C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_2456/ext)\n"]> expected but was <["rake RUBYARCHDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_2456/prefix RUBYLIBDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_2456/prefix extension", "(in C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_2456/ext)\nrequire_gem is obsolete. Use gem instead.\n"]>. 2) Failure: test_class_build_fail(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:45]: exception expected but none was thrown. 280 tests, 990 assertions, 2 failures, 0 errors rake aborted! Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib "c:/PRO...] (See full trace by running task with --trace) -- Cerberus 0.3.3, http://cerberus.rubyforge.org/ From anatol.pomozov at gmail.com Tue Dec 26 23:02:45 2006 From: anatol.pomozov at gmail.com (Cerberus) Date: Wed, 27 Dec 2006 07:02:45 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1181) Message-ID: <4591f065aba18_f141b242e615b@ubuntu.tmail> ------------------------------------------------------------------------ r1181 | drbrain | 2006-12-27 06:49:11 +0300 | 1 line Changed paths: M /trunk/ChangeLog M /trunk/lib/rubygems/package.rb M /trunk/test/test_format.rb M /trunk/test/test_package.rb Raise a useful exception class when we get a pad gem. ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) require_gem is obsolete. Use gem instead. Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/rake/rake_test_loader Started .........................................................................FF............................................................................................................................................................................................................. Finished in 41.891 seconds. 1) Failure: test_class_build(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:35]: <["rake RUBYARCHDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_2820/prefix RUBYLIBDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_2820/prefix extension", "(in C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_2820/ext)\n"]> expected but was <["rake RUBYARCHDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_2820/prefix RUBYLIBDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_2820/prefix extension", "(in C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_2820/ext)\nrequire_gem is obsolete. Use gem instead.\n"]>. 2) Failure: test_class_build_fail(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:45]: exception expected but none was thrown. 280 tests, 993 assertions, 2 failures, 0 errors rake aborted! Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib "c:/PRO...] (See full trace by running task with --trace) -- Cerberus 0.3.3, http://cerberus.rubyforge.org/ From drbrain at segment7.net Tue Dec 26 23:30:01 2006 From: drbrain at segment7.net (Eric Hodel) Date: Tue, 26 Dec 2006 20:30:01 -0800 Subject: [Rubygems-developers] [rubygems] Build still broken (#1179) In-Reply-To: <45917FFD.7050707@weirichhouse.org> References: <4591646ba7b98_e281b242dc1d9@ubuntu.tmail> <45917FFD.7050707@weirichhouse.org> Message-ID: On Dec 26, 2006, at 12:03, Jim Weirich wrote: > This one caught me too. You need to fix the calling script that > invokes rake, as the 'require_gem' warning is messing up the test > output. Does "gem pristine fix this yet? Yes, "gem pristine --all" will fix that. -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From anatol.pomozov at gmail.com Wed Dec 27 00:03:03 2006 From: anatol.pomozov at gmail.com (Cerberus) Date: Wed, 27 Dec 2006 08:03:03 +0300 Subject: [Rubygems-developers] [rubygems] Build still broken (#1182) Message-ID: <4591fe87a7b98_b41b242e6138@ubuntu.tmail> ------------------------------------------------------------------------ r1182 | drbrain | 2006-12-27 07:57:29 +0300 | 1 line Changed paths: M /trunk/ChangeLog M /trunk/lib/rubygems/installer.rb A /trunk/test/data/broken-1.0.0.gem M /trunk/test/test_format.rb M /trunk/test/test_installer.rb M /trunk/test/test_validator.rb Raise InstallError when unpacking a bad gem. ------------------------------------------------------------------------ (in C:/Documents and Settings/anatol/.cerberus/work/rubygems/sources) require_gem is obsolete. Use gem instead. Loaded suite c:/PROGRA~1/ruby/lib/ruby/gems/1.8/gems/rake-0.7.1/lib/rake/rake_test_loader Started .........................................................................FF..........................................................EE.................................................................................................................................................. Finished in 51.14 seconds. 1) Failure: test_class_build(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:35]: <["rake RUBYARCHDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3496/prefix RUBYLIBDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3496/prefix extension", "(in C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3496/ext)\n"]> expected but was <["rake RUBYARCHDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3496/prefix RUBYLIBDIR=C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3496/prefix extension", "(in C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3496/ext)\nrequire_gem is obsolete. Use gem instead.\n"]>. 2) Failure: test_class_build_fail(TestGemExtRakeBuilder) [./test/test_gem_ext_rake_builder.rb:45]: exception expected but none was thrown. 3) Error: test_install_bad_gem(TestInstaller): Errno::EACCES: Permission denied - C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3496/gemhome/cache/broken-1.0.0.gem c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1281:in `unlink' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1281:in `remove_file' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1289:in `platform_support' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1280:in `remove_file' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1269:in `remove' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:748:in `remove_entry' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1319:in `postorder_traverse' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1323:in `postorder_traverse' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1318:in `postorder_traverse' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1317:in `postorder_traverse' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1318:in `postorder_traverse' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1317:in `postorder_traverse' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1318:in `postorder_traverse' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1317:in `postorder_traverse' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:746:in `remove_entry' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:604:in `rm_r' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:600:in `rm_r' ./test/gemutilities.rb:78:in `teardown' 4) Error: test_install_with_message(TestInstaller): Errno::EACCES: Permission denied - C:/DOCUME~1/anatol/LOCALS~1/Temp/test_rubygems_3496/gemhome/cache/broken-1.0.0.gem c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1281:in `unlink' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1281:in `remove_file' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1289:in `platform_support' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1280:in `remove_file' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1269:in `remove' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:748:in `remove_entry' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1319:in `postorder_traverse' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1323:in `postorder_traverse' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1318:in `postorder_traverse' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1317:in `postorder_traverse' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1318:in `postorder_traverse' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1317:in `postorder_traverse' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1318:in `postorder_traverse' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:1317:in `postorder_traverse' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:746:in `remove_entry' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:604:in `rm_r' c:/PROGRA~1/ruby/lib/ruby/1.8/fileutils.rb:600:in `rm_r' ./test/gemutilities.rb:78:in `teardown' 281 tests, 995 assertions, 2 failures, 2 errors rake aborted! Command failed with status (1): [c:/PROGRA~1/ruby/bin/ruby -w -Ilib "c:/PRO...] (See full trace by running task with --trace) -- Cerberus 0.3.3, http://cerberus.rubyforge.org/ From rubygems at freeze.org Wed Dec 27 00:20:23 2006 From: rubygems at freeze.org (Jim Freeze) Date: Tue, 26 Dec 2006 23:20:23 -0600 Subject: [Rubygems-developers] error when irb not installed Message-ID: <65247773-74A3-444F-B1A6-632861E71130@freeze.org> Should I be getting an error when installing a gem on a machine that irb is not installed on? It appears that the install completes, but during the doc install, gem crashes when irb is not installed. Install required dependency text-format? [Yn] y Install required dependency text-hyphen? [Yn] y Successfully installed commandline-0.7.10 Successfully installed text-format-1.0.0 Successfully installed text-hyphen-1.0.0 /usr/lib/site_ruby/1.8/rubygems/custom_require.rb:35:in `require': No such file to load -- irb/slex (LoadError) from /usr/lib/site_ruby/1.8/rubygems/doc_manager.rb:71:in `load_rdoc' from /usr/lib/site_ruby/1.8/rubygems/doc_manager.rb:41:in `generate_ri' from /usr/lib/site_ruby/1.8/rubygems/gem_commands.rb:283:in `execute' from /usr/lib/site_ruby/1.8/rubygems/gem_commands.rb:282:in `each' from /usr/lib/site_ruby/1.8/rubygems/gem_commands.rb:282:in `execute' from /usr/lib/site_ruby/1.8/rubygems/gem_commands.rb:220:in `each' from /usr/lib/site_ruby/1.8/rubygems/gem_commands.rb:220:in `execute' from /usr/lib/site_ruby/1.8/rubygems/command.rb:69:in `invoke' from /usr/lib/site_ruby/1.8/rubygems/cmd_manager.rb:117:in `process_args' from /usr/lib/site_ruby/1.8/rubygems/cmd_manager.rb:88:in `run' from /usr/lib/site_ruby/1.8/rubygems/gem_runner.rb:29:in `run' from /usr/bin/gem:23 Jim Freeze -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20061226/78bdefca/attachment-0001.html From drbrain at segment7.net Wed Dec 27 02:38:44 2006 From: drbrain at segment7.net (Eric Hodel) Date: Tue, 26 Dec 2006 23:38:44 -0800 Subject: [Rubygems-developers] error when irb not installed In-Reply-To: <65247773-74A3-444F-B1A6-632861E71130@freeze.org> References: <65247773-74A3-444F-B1A6-632861E71130@freeze.org> Message-ID: <2682B9D8-8428-4371-B089-706C662933B7@segment7.net> On Dec 26, 2006, at 21:20, Jim Freeze wrote: > Should I be getting an error when installing a gem on > a machine that irb is not installed on? Yes. Rubygems requires all of ruby to be installed. > It appears that the install completes, but during the doc install, > gem crashes when irb is not installed. > > /usr/lib/site_ruby/1.8/rubygems/custom_require.rb:35:in `require': > No such file to load -- irb/slex (LoadError) > from /usr/lib/site_ruby/1.8/rubygems/doc_manager.rb:71:in > `load_rdoc' RDoc depends on irb. $ cat -n lib/rubygems/doc_manager.rb | grep 71 71 require 'rdoc/rdoc' -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From rubygems at freeze.org Wed Dec 27 08:08:36 2006 From: rubygems at freeze.org (Jim Freeze) Date: Wed, 27 Dec 2006 07:08:36 -0600 Subject: [Rubygems-developers] error when irb not installed In-Reply-To: <2682B9D8-8428-4371-B089-706C662933B7@segment7.net> References: <65247773-74A3-444F-B1A6-632861E71130@freeze.org> <2682B9D8-8428-4371-B089-706C662933B7@segment7.net> Message-ID: <2AB6ADBB-1F6D-4389-AC57-F8AFC3B0F676@freeze.org> On Dec 27, 2006, at 1:38 AM, Eric Hodel wrote: > On Dec 26, 2006, at 21:20, Jim Freeze wrote: > >> Should I be getting an error when installing a gem on >> a machine that irb is not installed on? > > Yes. Rubygems requires all of ruby to be installed. Maybe it should more gracefully fail. I was using CentOS at the time, and had installed ruby with 'yum install ruby'. It is an additional install 'yum install irb' to get irb. :( Jim Freeze -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20061227/ba83a9be/attachment.html From drbrain at segment7.net Wed Dec 27 17:40:47 2006 From: drbrain at segment7.net (Eric Hodel) Date: Wed, 27 Dec 2006 14:40:47 -0800 Subject: [Rubygems-developers] error when irb not installed In-Reply-To: <2AB6ADBB-1F6D-4389-AC57-F8AFC3B0F676@freeze.org> References: <65247773-74A3-444F-B1A6-632861E71130@freeze.org> <2682B9D8-8428-4371-B089-706C662933B7@segment7.net> <2AB6ADBB-1F6D-4389-AC57-F8AFC3B0F676@freeze.org> Message-ID: <179C055C-2F33-4077-983D-8C6818C38E7D@segment7.net> On Dec 27, 2006, at 05:08, Jim Freeze wrote: > On Dec 27, 2006, at 1:38 AM, Eric Hodel wrote: >> On Dec 26, 2006, at 21:20, Jim Freeze wrote: >>> Should I be getting an error when installing a gem on >>> a machine that irb is not installed on? >> >> Yes. Rubygems requires all of ruby to be installed. > > Maybe it should more gracefully fail. I was using CentOS at the > time, and had installed ruby with 'yum install ruby'. It is an > additional > install 'yum install irb' to get irb. :( Ruby's `make install` installs everything. There is no make target (or other built-in command) that will only install this standard library file or that standard library file for ruby. Adding hacky code to detect a hacky ruby install isn't maintainable. Today irb is missing, but tomorrow it might be that they've cut out thread.rb or OpenURI or ... RubyGems is set up by default to give you RDoc and ri data that you can use right away. Its fairly common knowledge that gem_server allows you to look at the RDoc for all of your gems. Attempting to degrade gracefully is going to confuse users. Instead of saying: "since you installed ruby from the ABC distribution you need to first install X, Y, and Z, then run `gem rdoc` for all your gems." and maintaining a table of distributions and commands necessary to install all of ruby and hacks to figure out what to do if some part of the standard library is missing. We can say: "irb is part of ruby. If you don't have irb with your ruby it is broken. Figure out how to install irb for ABC distribution then RubyGems will work. Be sure to file a bug with your distribution for shipping a broken ruby, too." and leave things as they are. The error you got is pretty clear: No such file to load -- irb/slex (LoadError) I would file a bug with CentOS. My /usr/local/lib/ruby/1.8 is all of 6.7M, so installing the standard library as it ships is not onerous. -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From rubygems at freeze.org Wed Dec 27 17:52:22 2006 From: rubygems at freeze.org (Jim Freeze) Date: Wed, 27 Dec 2006 16:52:22 -0600 Subject: [Rubygems-developers] error when irb not installed In-Reply-To: <179C055C-2F33-4077-983D-8C6818C38E7D@segment7.net> References: <65247773-74A3-444F-B1A6-632861E71130@freeze.org> <2682B9D8-8428-4371-B089-706C662933B7@segment7.net> <2AB6ADBB-1F6D-4389-AC57-F8AFC3B0F676@freeze.org> <179C055C-2F33-4077-983D-8C6818C38E7D@segment7.net> Message-ID: On Dec 27, 2006, at 4:40 PM, Eric Hodel wrote: > Ruby's `make install` installs everything. There is no make target > (or other built-in command) that will only install this standard > library file or that standard library file for ruby. ... > > "irb is part of ruby. If you don't have irb with your ruby it is > broken. Figure out how to install irb for ABC distribution then > RubyGems will work. Be sure to file a bug with your distribution for > shipping a broken ruby, too." ... > > I would file a bug with CentOS. My /usr/local/lib/ruby/1.8 is all of > 6.7M, so installing the standard library as it ships is not onerous. Yeah, I'm not sure why they left it out. Seems like they had to go out of their way make is a separate install. Jim Freeze -------------- next part -------------- An HTML attachment was scrubbed... URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20061227/33415d76/attachment.html From thewoolleyman at gmail.com Wed Dec 27 17:52:39 2006 From: thewoolleyman at gmail.com (Chad Woolley) Date: Wed, 27 Dec 2006 15:52:39 -0700 Subject: [Rubygems-developers] error when irb not installed In-Reply-To: <179C055C-2F33-4077-983D-8C6818C38E7D@segment7.net> References: <65247773-74A3-444F-B1A6-632861E71130@freeze.org> <2682B9D8-8428-4371-B089-706C662933B7@segment7.net> <2AB6ADBB-1F6D-4389-AC57-F8AFC3B0F676@freeze.org> <179C055C-2F33-4077-983D-8C6818C38E7D@segment7.net> Message-ID: I agree. Packagers should just install the full ruby install. Debian had these same problems, I don't know if they fixed it, because I always install from source now. I know the argument is that there's some stuff you may not want on a server install for security reasons, but that should be a special package for people who actually care (in other words, sysadmins who should probably be building from source anyway). For the other 99% of us who want a full "just works" install on every machine, the full install should be the default. -- Chad On 12/27/06, Eric Hodel wrote: > On Dec 27, 2006, at 05:08, Jim Freeze wrote: > > On Dec 27, 2006, at 1:38 AM, Eric Hodel wrote: > >> On Dec 26, 2006, at 21:20, Jim Freeze wrote: > >>> Should I be getting an error when installing a gem on > >>> a machine that irb is not installed on? > >> > >> Yes. Rubygems requires all of ruby to be installed. > > > > Maybe it should more gracefully fail. I was using CentOS at the > > time, and had installed ruby with 'yum install ruby'. It is an > > additional > > install 'yum install irb' to get irb. :( > > Ruby's `make install` installs everything. There is no make target > (or other built-in command) that will only install this standard > library file or that standard library file for ruby. > > Adding hacky code to detect a hacky ruby install isn't maintainable. > Today irb is missing, but tomorrow it might be that they've cut out > thread.rb or OpenURI or ... > > RubyGems is set up by default to give you RDoc and ri data that you > can use right away. Its fairly common knowledge that gem_server > allows you to look at the RDoc for all of your gems. Attempting to > degrade gracefully is going to confuse users. > > Instead of saying: > > "since you installed ruby from the ABC distribution you need to first > install X, Y, and Z, then run `gem rdoc` for all your gems." > > and maintaining a table of distributions and commands necessary to > install all of ruby and hacks to figure out what to do if some part > of the standard library is missing. > > We can say: > > "irb is part of ruby. If you don't have irb with your ruby it is > broken. Figure out how to install irb for ABC distribution then > RubyGems will work. Be sure to file a bug with your distribution for > shipping a broken ruby, too." > > and leave things as they are. > > The error you got is pretty clear: > > No such file to load -- irb/slex (LoadError) > > I would file a bug with CentOS. My /usr/local/lib/ruby/1.8 is all of > 6.7M, so installing the standard library as it ships is not onerous. > > -- > Eric Hodel - drbrain at segment7.net - http://blog.segment7.net > > I LIT YOUR GEM ON FIRE! > > _______________________________________________ > Rubygems-developers mailing list > Rubygems-developers at rubyforge.org > http://rubyforge.org/mailman/listinfo/rubygems-developers > From eloy.de.enige at gmail.com Thu Dec 28 07:18:22 2006 From: eloy.de.enige at gmail.com (Eloy Duran) Date: Thu, 28 Dec 2006 13:18:22 +0100 Subject: [Rubygems-developers] GUI Interfacing with rubygems? Message-ID: Hello, I'm trying to create an Mac OS X interface to rubygems which should be part of an RubyCocoa IDE. I've looked at another application (rubyslippers) which has a GUI for rubygems, but it seems to me that it just uses the 'gem' command line tool. Which is an option, but I would prefer to be able to use some public Gem api if available. Right now I did some testing with the internals of rubygems like: require 'rubygems/remote_installer' require 'yaml' query = 'event' result = Gem::RemoteInstaller.new().search(query) which returns an array of gems that have 'event' in their name, but it doesn't for instance group them like the results shown on the command line. So for a gem like 'eventmachine' instead of 1 result with multiple versions I would get multiple 'eventmachine' results for the different versions available. Also it only searches based on the title of the gem, not the description... So to wrap it up, I could of course revert to wrapping the 'gem' command line tool, and maybe even scraping the rubyforge.org search results page (because it also searches on the description). But I rather wouldn't. Could someone tell me if such an interface exists and maybe point me to some documentation or an example?? Kind regards, Eloy From thewoolleyman at gmail.com Thu Dec 28 14:42:18 2006 From: thewoolleyman at gmail.com (Chad Woolley) Date: Thu, 28 Dec 2006 12:42:18 -0700 Subject: [Rubygems-developers] GUI Interfacing with rubygems? In-Reply-To: References: Message-ID: Hey Eloy, I've done this in my GemInstaller project: http://rubyforge.org/projects/geminstaller/ This is almost ready for release. I didn't know about rubyslippers until now, I'm going to look at it, because I've run into a few hitches. Maybe it will give me some ideas. The biggest problems I've run into are: * There's no really cohesive API for RubyGems. Whether you use the command line or the api, you have to do parsing out text and acting on it (parsing versions/platforms/etc). If you want to use it programatically, you have to hit a few classes. I've got proxies for GemRunner and SourceIndex, and some mixin hacks for StreamUI so I can intercept stdin and stdout. * Specifying platforms for multiplatform gems. Currently, there is no way to specify platform via a command line option or API. It's hardcoded to always present a list if theres a binary (non-ruby) gem, and ask for the user to make a choice via stdin (which I had to work around with the mixin hacks for StreamUI). There is talk of changing this, but the developers want to gather data on what platforms people use first. I'd really like to just get a command line option for platform added ASAP - hint hint :) * Dealing with sudo. This is currently my only remaining blocker for releasing geminstaller. Must people have rubygems installed as root, so commands that modify the gem repository must be run via sudo. I would like to have a command line option --sudo which uses sudo to run gems, but this doesn't work if you are calling into the API programatically. On the other hand, I could use the command line interface, but this is also tricky in ruby, to properly deal with stderr, stdin, and timeouts if the gem command is expecting different stdin than you think it is. Also, if you use the command line, you have to parse errors out of stdout/stderr as opposed to just catching exceptions. If anyone has ideas here, please let me know. Like I said, I'll look into RubySlippers and see if it gives me any ideas to deal with these problems. Please feel free to contact me directly if you'd like to collaborate, I'm thewoolleyman at gmail.com. Thanks, -- Chad W. On 12/28/06, Eloy Duran wrote: > Hello, > > I'm trying to create an Mac OS X interface to rubygems which should > be part of an RubyCocoa IDE. > I've looked at another application (rubyslippers) which has a GUI for rubygems, > but it seems to me that it just uses the 'gem' command line tool. > Which is an option, but I would prefer to be able to use some public Gem api > if available. > > Right now I did some testing with the internals of rubygems like: > > > require 'rubygems/remote_installer' > require 'yaml' > > query = 'event' > result = Gem::RemoteInstaller.new().search(query) > > > which returns an array of gems that have 'event' in their name, > but it doesn't for instance group them like the results shown on the > command line. > So for a gem like 'eventmachine' instead of 1 result with multiple versions > I would get multiple 'eventmachine' results for the different versions > available. > Also it only searches based on the title of the gem, not the description... > > So to wrap it up, I could of course revert to wrapping the 'gem' > command line tool, > and maybe even scraping the rubyforge.org search results page > (because it also searches on the description). > But I rather wouldn't. Could someone tell me if such an interface > exists and maybe > point me to some documentation or an example?? > > Kind regards, > Eloy > _______________________________________________ > Rubygems-developers mailing list > Rubygems-developers at rubyforge.org > http://rubyforge.org/mailman/listinfo/rubygems-developers > From drbrain at segment7.net Thu Dec 28 14:51:14 2006 From: drbrain at segment7.net (Eric Hodel) Date: Thu, 28 Dec 2006 11:51:14 -0800 Subject: [Rubygems-developers] GUI Interfacing with rubygems? In-Reply-To: References: Message-ID: <95393AB7-8342-44FD-86B2-1179506BCBB9@segment7.net> On Dec 28, 2006, at 04:18, Eloy Duran wrote: > I'm trying to create an Mac OS X interface to rubygems which should > be part of an RubyCocoa IDE. > > I've looked at another application (rubyslippers) which has a GUI > for rubygems, but it seems to me that it just uses the 'gem' > command line tool. Which is an option, but I would prefer to be > able to use some public Gem api if available. > > Right now I did some testing with the internals of rubygems like: > > > require 'rubygems/remote_installer' > require 'yaml' You shouldn't need to require yaml. > query = 'event' > result = Gem::RemoteInstaller.new().search(query) This method no longer exists, you'll now find it over on Gem::SourceInfoCache where it makes more sense. > > > which returns an array of gems that have 'event' in their name, but > it doesn't for instance group them like the results shown on the > command line. > > So for a gem like 'eventmachine' instead of 1 result with multiple > versions I would get multiple 'eventmachine' results for the > different versions available. You can convert a list of matches into the results you want with about five lines of code. Internally, having them grouped the way the command line does makes things much more difficult. > Also it only searches based on the title of the gem, not the > description... RubyGems provides no built-in way of searching descriptions. > So to wrap it up, I could of course revert to wrapping the 'gem' > command line tool, and maybe even scraping the rubyforge.org search > results page (because it also searches on the description). rubyforge.org does not search the gem descriptions. > But I rather wouldn't. Could someone tell me if such an interface > exists and maybe point me to some documentation or an example?? I've done a bunch of work to make RubyGems more scriptable for 0.9.1. Try out the recent beta (0.9.0.8). Unfortunately, too much code lives in lib/rubygems/gem_commands.rb that should live elsewhere. I'm doing what I can to correct this, which will make everyone's lives better. -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From thewoolleyman at gmail.com Thu Dec 28 15:02:11 2006 From: thewoolleyman at gmail.com (Chad Woolley) Date: Thu, 28 Dec 2006 13:02:11 -0700 Subject: [Rubygems-developers] GUI Interfacing with rubygems? In-Reply-To: References: Message-ID: FYI, it looks like this is how the rubyslippers developer is getting around the sudo problem (from his Readme): "On my Linux system I changed permissions on my gem directory to allow the nonprovileged user to install gems." I agree that's what people should do, but i was trying to be nice to people who don't know unix that well. Guess I should just punt, since I have a --sudo option working from the command line, but not when geminstaller is invoked programatically (as will be the case from a Rails app) -- Chad W On 12/28/06, Chad Woolley wrote: > Hey Eloy, > > I've done this in my GemInstaller project: > http://rubyforge.org/projects/geminstaller/ > > This is almost ready for release. I didn't know about rubyslippers > until now, I'm going to look at it, because I've run into a few > hitches. Maybe it will give me some ideas. The biggest problems I've > run into are: > > * There's no really cohesive API for RubyGems. Whether you use the > command line or the api, you have to do parsing out text and acting on > it (parsing versions/platforms/etc). If you want to use it > programatically, you have to hit a few classes. I've got proxies for > GemRunner and SourceIndex, and some mixin hacks for StreamUI so I can > intercept stdin and stdout. > > * Specifying platforms for multiplatform gems. Currently, there is no > way to specify platform via a command line option or API. It's > hardcoded to always present a list if theres a binary (non-ruby) gem, > and ask for the user to make a choice via stdin (which I had to work > around with the mixin hacks for StreamUI). There is talk of changing > this, but the developers want to gather data on what platforms people > use first. I'd really like to just get a command line option for > platform added ASAP - hint hint :) > > * Dealing with sudo. This is currently my only remaining blocker for > releasing geminstaller. Must people have rubygems installed as root, > so commands that modify the gem repository must be run via sudo. I > would like to have a command line option --sudo which uses sudo to run > gems, but this doesn't work if you are calling into the API > programatically. On the other hand, I could use the command line > interface, but this is also tricky in ruby, to properly deal with > stderr, stdin, and timeouts if the gem command is expecting different > stdin than you think it is. Also, if you use the command line, you > have to parse errors out of stdout/stderr as opposed to just catching > exceptions. If anyone has ideas here, please let me know. > > Like I said, I'll look into RubySlippers and see if it gives me any > ideas to deal with these problems. Please feel free to contact me > directly if you'd like to collaborate, I'm thewoolleyman at gmail.com. > > Thanks, > -- Chad W. > > On 12/28/06, Eloy Duran wrote: > > Hello, > > > > I'm trying to create an Mac OS X interface to rubygems which should > > be part of an RubyCocoa IDE. > > I've looked at another application (rubyslippers) which has a GUI for rubygems, > > but it seems to me that it just uses the 'gem' command line tool. > > Which is an option, but I would prefer to be able to use some public Gem api > > if available. > > > > Right now I did some testing with the internals of rubygems like: > > > > > > require 'rubygems/remote_installer' > > require 'yaml' > > > > query = 'event' > > result = Gem::RemoteInstaller.new().search(query) > > > > > > which returns an array of gems that have 'event' in their name, > > but it doesn't for instance group them like the results shown on the > > command line. > > So for a gem like 'eventmachine' instead of 1 result with multiple versions > > I would get multiple 'eventmachine' results for the different versions > > available. > > Also it only searches based on the title of the gem, not the description... > > > > So to wrap it up, I could of course revert to wrapping the 'gem' > > command line tool, > > and maybe even scraping the rubyforge.org search results page > > (because it also searches on the description). > > But I rather wouldn't. Could someone tell me if such an interface > > exists and maybe > > point me to some documentation or an example?? > > > > Kind regards, > > Eloy > > _______________________________________________ > > Rubygems-developers mailing list > > Rubygems-developers at rubyforge.org > > http://rubyforge.org/mailman/listinfo/rubygems-developers > > > From drbrain at segment7.net Thu Dec 28 15:04:14 2006 From: drbrain at segment7.net (Eric Hodel) Date: Thu, 28 Dec 2006 12:04:14 -0800 Subject: [Rubygems-developers] GUI Interfacing with rubygems? In-Reply-To: References: Message-ID: <2F5C7A4A-DDF2-4269-9662-A4136694DAC8@segment7.net> On Dec 28, 2006, at 11:42, Chad Woolley wrote: > Hey Eloy, > > I've done this in my GemInstaller project: > http://rubyforge.org/projects/geminstaller/ > > This is almost ready for release. I didn't know about rubyslippers > until now, I'm going to look at it, because I've run into a few > hitches. Maybe it will give me some ideas. The biggest problems I've > run into are: > > * There's no really cohesive API for RubyGems. Whether you use the > command line or the api, you have to do parsing out text and acting on > it (parsing versions/platforms/etc). If you want to use it > programatically, you have to hit a few classes. I've done a ton of work on this for the 0.9.1 release. Methods have shifted around and the RemoteInstaller has been drastically cut down in capabilities. Those capabilities have all moved to other locations where they make more sense. (I did all this to perform automatic sandbox gem installs for my secret project.) > I've got proxies for GemRunner and SourceIndex, and some mixin > hacks for StreamUI so I can intercept stdin and stdout. Why do you need a proxy for SourceIndex? Its pretty much self- contained. > * Specifying platforms for multiplatform gems. Currently, there is no > way to specify platform via a command line option or API. It's > hardcoded to always present a list if theres a binary (non-ruby) gem, > and ask for the user to make a choice via stdin (which I had to work > around with the mixin hacks for StreamUI). There is talk of changing > this, but the developers want to gather data on what platforms people > use first. I'd really like to just get a command line option for > platform added ASAP - hint hint :) This is on the table for 0.9.2. It was too big of a change to make for 0.9.1. Chad and Jim have a side project that'll get us information to help with this too. > * Dealing with sudo. This is currently my only remaining blocker for > releasing geminstaller. Must people have rubygems installed as root, > so commands that modify the gem repository must be run via sudo. I > would like to have a command line option --sudo which uses sudo to run > gems, but this doesn't work if you are calling into the API > programatically. On the other hand, I could use the command line > interface, but this is also tricky in ruby, to properly deal with > stderr, stdin, and timeouts if the gem command is expecting different > stdin than you think it is. 0.9.1 will not download gems from the server if everything is up-to- date. This will give fewer instances where gems will need to read from the console (of course, there's always the platform problem). (In 0.9.2 I'd like to pull most of RemoteInstaller's functionality into a DependencyInstaller which will make network-free installs much cleaner.) > Also, if you use the command line, you have to parse errors out of > stdout/stderr as opposed to just catching exceptions. If anyone > has ideas here, please let me know. 0.9.1 raises many, many more exceptions (and useful ones, at that) and outputs to stdout much less (especially for extension building). -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From anthonyeden at gmail.com Thu Dec 28 15:15:29 2006 From: anthonyeden at gmail.com (Anthony Eden) Date: Thu, 28 Dec 2006 15:15:29 -0500 Subject: [Rubygems-developers] GUI Interfacing with rubygems? In-Reply-To: References: Message-ID: Why not include an option in the tool which allows a user to execute the chmod automatically with a single action? Explain what it will do and then let them verify that they are ok with it. V/r Anthony On 12/28/06, Chad Woolley wrote: > FYI, it looks like this is how the rubyslippers developer is getting > around the sudo problem (from his Readme): > > "On my Linux system I changed permissions on my gem directory to allow > the nonprovileged user to install gems." > > I agree that's what people should do, but i was trying to be nice to > people who don't know unix that well. Guess I should just punt, since > I have a --sudo option working from the command line, but not when > geminstaller is invoked programatically (as will be the case from a > Rails app) > > -- Chad W > > On 12/28/06, Chad Woolley wrote: > > Hey Eloy, > > > > I've done this in my GemInstaller project: > > http://rubyforge.org/projects/geminstaller/ > > > > This is almost ready for release. I didn't know about rubyslippers > > until now, I'm going to look at it, because I've run into a few > > hitches. Maybe it will give me some ideas. The biggest problems I've > > run into are: > > > > * There's no really cohesive API for RubyGems. Whether you use the > > command line or the api, you have to do parsing out text and acting on > > it (parsing versions/platforms/etc). If you want to use it > > programatically, you have to hit a few classes. I've got proxies for > > GemRunner and SourceIndex, and some mixin hacks for StreamUI so I can > > intercept stdin and stdout. > > > > * Specifying platforms for multiplatform gems. Currently, there is no > > way to specify platform via a command line option or API. It's > > hardcoded to always present a list if theres a binary (non-ruby) gem, > > and ask for the user to make a choice via stdin (which I had to work > > around with the mixin hacks for StreamUI). There is talk of changing > > this, but the developers want to gather data on what platforms people > > use first. I'd really like to just get a command line option for > > platform added ASAP - hint hint :) > > > > * Dealing with sudo. This is currently my only remaining blocker for > > releasing geminstaller. Must people have rubygems installed as root, > > so commands that modify the gem repository must be run via sudo. I > > would like to have a command line option --sudo which uses sudo to run > > gems, but this doesn't work if you are calling into the API > > programatically. On the other hand, I could use the command line > > interface, but this is also tricky in ruby, to properly deal with > > stderr, stdin, and timeouts if the gem command is expecting different > > stdin than you think it is. Also, if you use the command line, you > > have to parse errors out of stdout/stderr as opposed to just catching > > exceptions. If anyone has ideas here, please let me know. > > > > Like I said, I'll look into RubySlippers and see if it gives me any > > ideas to deal with these problems. Please feel free to contact me > > directly if you'd like to collaborate, I'm thewoolleyman at gmail.com. > > > > Thanks, > > -- Chad W. > > > > On 12/28/06, Eloy Duran wrote: > > > Hello, > > > > > > I'm trying to create an Mac OS X interface to rubygems which should > > > be part of an RubyCocoa IDE. > > > I've looked at another application (rubyslippers) which has a GUI for rubygems, > > > but it seems to me that it just uses the 'gem' command line tool. > > > Which is an option, but I would prefer to be able to use some public Gem api > > > if available. > > > > > > Right now I did some testing with the internals of rubygems like: > > > > > > > > > require 'rubygems/remote_installer' > > > require 'yaml' > > > > > > query = 'event' > > > result = Gem::RemoteInstaller.new().search(query) > > > > > > > > > which returns an array of gems that have 'event' in their name, > > > but it doesn't for instance group them like the results shown on the > > > command line. > > > So for a gem like 'eventmachine' instead of 1 result with multiple versions > > > I would get multiple 'eventmachine' results for the different versions > > > available. > > > Also it only searches based on the title of the gem, not the description... > > > > > > So to wrap it up, I could of course revert to wrapping the 'gem' > > > command line tool, > > > and maybe even scraping the rubyforge.org search results page > > > (because it also searches on the description). > > > But I rather wouldn't. Could someone tell me if such an interface > > > exists and maybe > > > point me to some documentation or an example?? > > > > > > Kind regards, > > > Eloy > > > _______________________________________________ > > > Rubygems-developers mailing list > > > Rubygems-developers at rubyforge.org > > > http://rubyforge.org/mailman/listinfo/rubygems-developers > > > > > > _______________________________________________ > Rubygems-developers mailing list > Rubygems-developers at rubyforge.org > http://rubyforge.org/mailman/listinfo/rubygems-developers > -- Cell: 808 782-5046 Current Location: Melbourne, FL From eloy.de.enige at gmail.com Thu Dec 28 17:20:06 2006 From: eloy.de.enige at gmail.com (Eloy Duran) Date: Thu, 28 Dec 2006 23:20:06 +0100 Subject: [Rubygems-developers] GUI Interfacing with rubygems? In-Reply-To: References: Message-ID: Wow, lots of responses I will read them carefully tomorrow. But thanks a lot! Cheers, Eloy From thewoolleyman at gmail.com Thu Dec 28 20:47:00 2006 From: thewoolleyman at gmail.com (Chad Woolley) Date: Thu, 28 Dec 2006 18:47:00 -0700 Subject: [Rubygems-developers] GUI Interfacing with rubygems? In-Reply-To: References: Message-ID: On 12/28/06, Anthony Eden wrote: > Why not include an option in the tool which allows a user to execute > the chmod automatically with a single action? Explain what it will do > and then let them verify that they are ok with it. > > V/r > Anthony Great idea, and that's what I'll probably do. Initially, it will be some instructions in the docs :) The other option is to include directions on invoking the command line programatically with `` or IO.popen, since I already support a --sudo option from the executable wrapper script (by recursively calling my executable again with sudo). From thewoolleyman at gmail.com Thu Dec 28 21:05:43 2006 From: thewoolleyman at gmail.com (Chad Woolley) Date: Thu, 28 Dec 2006 19:05:43 -0700 Subject: [Rubygems-developers] GUI Interfacing with rubygems? In-Reply-To: <2F5C7A4A-DDF2-4269-9662-A4136694DAC8@segment7.net> References: <2F5C7A4A-DDF2-4269-9662-A4136694DAC8@segment7.net> Message-ID: On 12/28/06, Eric Hodel wrote: > On Dec 28, 2006, at 11:42, Chad Woolley wrote: > > * There's no really cohesive API for RubyGems. Whether you use the > > command line or the api, you have to do parsing out text and acting on > > it (parsing versions/platforms/etc). If you want to use it > > programatically, you have to hit a few classes. > > I've done a ton of work on this for the 0.9.1 release. Methods have > shifted around and the RemoteInstaller has been drastically cut down > in capabilities. Those capabilities have all moved to other > locations where they make more sense. (I did all this to perform > automatic sandbox gem installs for my secret project.) > Great news. Too bad I'll still have to support the old release for a while :) > > I've got proxies for GemRunner and SourceIndex, and some mixin > > hacks for StreamUI so I can intercept stdin and stdout. > > Why do you need a proxy for SourceIndex? Its pretty much self- > contained. > Because I'm using Rspec and mock-intensive Behavior-Driven Development approach. I follow the rule of thumb to "Only Mock Types You Own" [1]. In other words, wrap any third-party code you interface with in a proxy interface, and only proxy the methods you need to use, then test the heck out of it with "Enemy Unit Tests" [2]. That way, if the behavior of the third party code changes (which it very might well have with your refactorings!), you know exactly what isn't behaving the way it used to. These ideas are from: 1. "Mock Roles, Not Objects" - http://www.jmock.org/oopsla2004.pdf 2. "Enemy Unit Tests" - the GroboUtils project - http://groboutils.sourceforge.net/testing-junit/art_eut.html > > * Specifying platforms for multiplatform gems. Currently, there is no > > way to specify platform via a command line option or API. It's > > hardcoded to always present a list if theres a binary (non-ruby) gem, > > and ask for the user to make a choice via stdin (which I had to work > > around with the mixin hacks for StreamUI). There is talk of changing > > this, but the developers want to gather data on what platforms people > > use first. I'd really like to just get a command line option for > > platform added ASAP - hint hint :) > > This is on the table for 0.9.2. It was too big of a change to make > for 0.9.1. Chad and Jim have a side project that'll get us > information to help with this too. Yep, they mentioned Tattle in the earlier thread where I asked. I just hope that the Tattle data gathering process doesn't unnecessarily delay the addition of the --platform option. Actually, embedding hooks/autoinstall for Tattle in tools like GemInstaller, RubySlippers, and Eloy's project seems like a great way to get Tattle in widespread use more quickly. > > > * Dealing with sudo. This is currently my only remaining blocker for > > releasing geminstaller. Must people have rubygems installed as root, > > so commands that modify the gem repository must be run via sudo. I > > would like to have a command line option --sudo which uses sudo to run > > gems, but this doesn't work if you are calling into the API > > programatically. On the other hand, I could use the command line > > interface, but this is also tricky in ruby, to properly deal with > > stderr, stdin, and timeouts if the gem command is expecting different > > stdin than you think it is. > > 0.9.1 will not download gems from the server if everything is up-to- > date. This will give fewer instances where gems will need to read > from the console (of course, there's always the platform problem). > > (In 0.9.2 I'd like to pull most of RemoteInstaller's functionality > into a DependencyInstaller which will make network-free installs much > cleaner.) > > > Also, if you use the command line, you have to parse errors out of > > stdout/stderr as opposed to just catching exceptions. If anyone > > has ideas here, please let me know. > > 0.9.1 raises many, many more exceptions (and useful ones, at that) > and outputs to stdout much less (especially for extension building). Yes, but you still need sudo support if a gem actually needs to be installed, and AFAIK this means invoking the command line tool programatically. Also, one (maybe bad) suggestion would be to eliminate the use of stderr altogether, since it's relatively difficult to capture stderr output under Ruby in a cross-platform (works on windows) way. IO.popen doesn't allow you to capture stderr, and output = `gem #{args_string} 2>&1` is the only other way that I know works on windows, but this approach makes it difficult to handle stdin. To see my as-yet-unsuccessful attempts to solve this issue, look at the history of gem_command_line_proxy.rb and gem_command_line_proxy_spec.rb in the geminstaller source. > Eric Hodel - drbrain at segment7.net - http://blog.segment7.net Thanks for the responses! -- Chad W. From drbrain at segment7.net Fri Dec 29 03:17:23 2006 From: drbrain at segment7.net (Eric Hodel) Date: Fri, 29 Dec 2006 00:17:23 -0800 Subject: [Rubygems-developers] rubygems 0.9.0.9 beta time? Message-ID: <4D500F49-AB0A-41ED-B314-82BFF57407A2@segment7.net> I saw Jim made a flurry of cleanups, and I made some changes to 'gem pristine' to make the require_gem -> gem transition to flow more smoothly. Since its holiday time this will keep the pending 0.9.1 release in the public eye a little better, and hopefully we can release 0.9.1 in the following week. Are there any issues pending that I don't know about or have forgotten? -- Eric Hodel - drbrain at segment7.net - http://blog.segment7.net I LIT YOUR GEM ON FIRE! From jim at weirichhouse.org Fri Dec 29 08:16:21 2006 From: jim at weirichhouse.org (Jim Weirich) Date: Fri, 29 Dec 2006 08:16:21 -0500 Subject: [Rubygems-developers] rubygems 0.9.0.9 beta time? In-Reply-To: <4D500F49-AB0A-41ED-B314-82BFF57407A2@segment7.net> References: <4D500F49-AB0A-41ED-B314-82BFF57407A2@segment7.net> Message-ID: <45951525.4080705@weirichhouse.org> Eric Hodel wrote: > I saw Jim made a flurry of cleanups, and I made some changes to 'gem > pristine' to make the require_gem -> gem transition to flow more > smoothly. > > Since its holiday time this will keep the pending 0.9.1 release in > the public eye a little better, and hopefully we can release 0.9.1 in > the following week. > > Are there any issues pending that I don't know about or have forgotten? I am talking with Michael Studman about a patch to the Specification hash method. Evidently, some gems have bad values in the specification that is cause problems with the "hash" method and the "==" operator. The problem is manifest on JRuby more often than on standard Ruby just because JRuby uses a different hash algorithm that happens to compare the bad specs more often than the C version of Ruby. We have beefed up the hash and == operators in Specification, but Michael believes there are still some problems in this area. His patch changes the hash function so that the distribution of specs in the hash are different, making it less likely that the bad specs collide in the same hash bucket. But that's a solution that only "probably" fixes the issue. I'm working with him to identify the exact problem to come up with a better fix. BTW, I've asked Michael to join this list so the issue gets better visibility. -- -- Jim Weirich jim at weirichhouse.org http://onestepback.org -- In theory, practice and theory are the same. -- In practice, they are different.