From drbrain at segment7.net  Mon Dec  5 19:39:27 2005
From: drbrain at segment7.net (Eric Hodel)
Date: Mon, 5 Dec 2005 16:39:27 -0800
Subject: [Rubygems-developers] $SAFE = 1 patch
Message-ID: <7734EEE9-AAF0-4AF8-91F4-198B13E4FDB0@segment7.net>

This patch lets Rubygems run with $SAFE = 1.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: SAFE.patch
Type: application/octet-stream
Size: 3568 bytes
Desc: not available
Url : http://rubyforge.org/pipermail/rubygems-developers/attachments/20051205/b8d075e8/SAFE-0001.obj
-------------- next part --------------

-- 
Eric Hodel - drbrain at segment7.net - http://segment7.net
This implementation is HODEL-HASH-9600 compliant

http://trackmap.robotcoop.com



From jim.weirich at gmail.com  Tue Dec  6 09:24:06 2005
From: jim.weirich at gmail.com (Jim Weirich)
Date: Tue, 6 Dec 2005 09:24:06 -0500
Subject: [Rubygems-developers] $SAFE = 1 patch
In-Reply-To: <7734EEE9-AAF0-4AF8-91F4-198B13E4FDB0@segment7.net>
References: <7734EEE9-AAF0-4AF8-91F4-198B13E4FDB0@segment7.net>
Message-ID: <d4dd9080512060624m66fccd1fj9057ebe8b150337f@mail.gmail.com>

On 12/5/05, Eric Hodel <drbrain at segment7.net> wrote:
>
> This patch lets Rubygems run with $SAFE = 1.


Excellent!  And the change comes with a patch for the tests too!  Extra
points for that.

But ... I added the $SAFE=1 to the test (per the patch file) and ran the
tests (without the rest of your patch installed) ... and all the tests still
pass.  I expected to see failures because of $SAFE=1 mode.  Is the test not
testing what we think?

On a slightly different note, shouldn't we do more than just untaint the
string.  I would have thought that at least some minimal validity checking
would be appropriate ... after all, that's the reason the strings are
tainted in the first place.

--
--
-- 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/20051206/69d905d2/attachment.htm

From drbrain at segment7.net  Tue Dec  6 14:36:14 2005
From: drbrain at segment7.net (Eric Hodel)
Date: Tue, 6 Dec 2005 11:36:14 -0800
Subject: [Rubygems-developers] $SAFE = 1 patch
In-Reply-To: <d4dd9080512060624m66fccd1fj9057ebe8b150337f@mail.gmail.com>
References: <7734EEE9-AAF0-4AF8-91F4-198B13E4FDB0@segment7.net>
	<d4dd9080512060624m66fccd1fj9057ebe8b150337f@mail.gmail.com>
Message-ID: <0581AAC4-455B-460F-B388-24FB67492211@segment7.net>

On Dec 6, 2005, at 6:24 AM, Jim Weirich wrote:

> On 12/5/05, Eric Hodel <drbrain at segment7.net> wrote: This patch  
> lets Rubygems run with $SAFE = 1.
>
> Excellent!  And the change comes with a patch for the tests too!   
> Extra points for that.
>
> But ... I added the $SAFE=1 to the test (per the patch file) and  
> ran the tests (without the rest of your patch installed) ... and  
> all the tests still pass.  I expected to see failures because of  
> $SAFE=1 mode.  Is the test not testing what we think?

Strange, I get failures with just the change to test/gemenvironment.rb:

[11:31] drbrain at kaa$ rake
(in /Users/drbrain/tmp/rubygems)
./lib/rubygems/custom_require.rb:99:in `[]': Insecure operation - []  
(SecurityError)
         from ./lib/rubygems/custom_require.rb:99:in `matching_file'
         from ./lib/rubygems/custom_require.rb:81:in `find'
         from ./lib/rubygems/custom_require.rb:80:in `find'
         from ./lib/rubygems/custom_require.rb:25:in `require'
         from ./test/gemenvironment.rb:8
         from ./lib/rubygems/custom_require.rb:21:in `require'
         from ./test/test_gemloadpaths.rb:6
         from /usr/local/lib/ruby/gems/1.8/gems/rake-0.6.2/lib/rake/ 
rake_test_loader.rb:5
         from /usr/local/lib/ruby/gems/1.8/gems/rake-0.6.2/lib/rake/ 
rake_test_loader.rb:5
rake aborted!
Command failed with status (1): [/usr/local/bin/ruby183 -Ilib "/usr/ 
local/l...]

[ ~/tmp/rubygems ]
[11:31] drbrain at kaa$ cvs diff
Index: test/gemenvironment.rb
===================================================================
RCS file: /var/cvs/rubygems/rubygems/test/gemenvironment.rb,v
retrieving revision 1.3
diff -u -r1.3 gemenvironment.rb
--- test/gemenvironment.rb      17 Mar 2005 04:04:15 -0000      1.3
+++ test/gemenvironment.rb      6 Dec 2005 19:22:04 -0000
@@ -1,5 +1,6 @@
# Create a test environment for gems.
+$SAFE = 1
require 'rubygems'
require 'rubygems/installer'
require 'rubygems/builder'

> On a slightly different note, shouldn't we do more than just  
> untaint the string.  I would have thought that at least some  
> minimal validity checking would be appropriate ... after all,  
> that's the reason the strings are tainted in the first place.

I don't know.  All the paths seem to come from the gem directory or  
the gemspec files, so I thought it was ok to trust them.  I thought  
it would be strange to distrust the gem directory...

-- 
Eric Hodel - drbrain at segment7.net - http://segment7.net
This implementation is HODEL-HASH-9600 compliant

http://trackmap.robotcoop.com



From jim.weirich at gmail.com  Tue Dec  6 16:00:44 2005
From: jim.weirich at gmail.com (Jim Weirich)
Date: Tue, 6 Dec 2005 16:00:44 -0500
Subject: [Rubygems-developers] $SAFE = 1 patch
In-Reply-To: <0581AAC4-455B-460F-B388-24FB67492211@segment7.net>
References: <7734EEE9-AAF0-4AF8-91F4-198B13E4FDB0@segment7.net>
	<d4dd9080512060624m66fccd1fj9057ebe8b150337f@mail.gmail.com>
	<0581AAC4-455B-460F-B388-24FB67492211@segment7.net>
Message-ID: <d4dd9080512061300r37aafbccw5c0349072ed09330@mail.gmail.com>

On 12/6/05, Eric Hodel <drbrain at segment7.net> wrote:> Strange, I get
failures with just the change to test/gemenvironment.rb:

Mystery solved ... Eric is running 1.8.3, I am on 1.8.2.

I have commited the SAFE patch ... enjoy.

--
--
-- 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/20051206/2426433a/attachment.htm

From anatol.pomozov at gmail.com  Sat Dec 10 08:48:20 2005
From: anatol.pomozov at gmail.com (Anatol Pomozov)
Date: Sat, 10 Dec 2005 14:48:20 +0100
Subject: [Rubygems-developers] Gem problem behind Proxy with authorization.
Message-ID: <3665a1a00512100548o4b03b9b7u48ccb9fd32d30314@mail.gmail.com>

Hi. I am again picking up question about Gem that working behind proxy with
authorization.
Problem related to
http://rubyforge.org/tracker/index.php?func=detail&aid=1289&group_id=126&atid=575
Seems that is most requestable bug in Gem project.

I am sitting behind passworded proxy and only this problem give me no rest.
After installing DamageControl that requires
more then 10 gems I decided to investigate this problem by myself.

First of all Open-uri lib does not support http_proxy with login/password
encoded in URL.
So my EnvVariable HTTP_PROXY=http://lgoin:passwd at www.proxy.com:8080/ does
not work.
I patched open-uri to solve this problem. To be honest I patch open-uri lib
from RubyCVS version that will be in Ruby 1.8.4 version.

Here you can find patch and here full *.rb version(
http://pomozov.info/downloads/gems/open-uri.patch and
http://pomozov.info/downloads/gems/open-uri.rb). I have replaced following
file "C:\Program Files\ruby\lib\ruby\site_ruby\1.8\rubygems\open-uri.rb"
with patched version

And following code(insied patched open-uri)
    io.status = [resp.code, resp.message]
    p io.status
give us
    ["200", "OK"]

So now socket connection seems ok.

Also I have tried to dump data that comes from socket (
http://pomozov.info/downloads/gems/dump.Z) and I found that it is absolutely
the same files as http://gems.rubyforge.org/yaml.Z
So Gem *could download* index file. Proxy wall passed!!!! I feel solution is
somewhere close to me.
But I try to run gem and I get error

C:\Documents and Settings\anatol>gem list -r --backtrace

*** REMOTE GEMS ***
Updating Gem source index for: http://gems.rubyforge.org
ERROR:  HTTP Response 407

No stacktrace, no error. What I am doing wrong?? How could I get more info?
How to figure place where error occurred.
Please help me I want to solve this problem very much. I want to use Gems.
--
anatol (http://pomozov.info)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20051210/4439cd42/attachment.htm

From anatol.pomozov at gmail.com  Sun Dec 11 09:24:19 2005
From: anatol.pomozov at gmail.com (Anatol Pomozov)
Date: Sun, 11 Dec 2005 15:24:19 +0100
Subject: [Rubygems-developers] Why RubyGem lib contains open-uri library??
Message-ID: <3665a1a00512110624g19f52294p4287986ecf100c54@mail.gmail.com>

Hi, all. I am trying to solve problem with proxy that I wrote yesterday.

I am reading the code and have one question: why RubyGem have own version
open-uri lib?? Standart ruby distibution shipped with this library (at least
on Windows).

Could anybody shine some light on this question??

--
anatol (http://pomozov.info)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20051211/2a3a8279/attachment.htm

From jim.weirich at gmail.com  Sun Dec 11 13:18:18 2005
From: jim.weirich at gmail.com (Jim Weirich)
Date: Sun, 11 Dec 2005 13:18:18 -0500
Subject: [Rubygems-developers] Why RubyGem lib contains open-uri
	library??
In-Reply-To: <3665a1a00512110624g19f52294p4287986ecf100c54@mail.gmail.com>
References: <3665a1a00512110624g19f52294p4287986ecf100c54@mail.gmail.com>
Message-ID: <d4dd9080512111018t27893875ga624f1f3389de4ad@mail.gmail.com>

On 12/11/05, Anatol Pomozov <anatol.pomozov at gmail.com> wrote:
> Hi, all. I am trying to solve problem with proxy that I wrote yesterday.
>
> I am reading the code and have one question: why RubyGem have own version
> open-uri lib?? Standart ruby distibution shipped with this library (at least
> on Windows).
>
> Could anybody shine some light on this question??

When we first started using open-uri, there was a bug in the library,
so we had to supply a patched version of it.

--
--
-- 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 anatol.pomozov at gmail.com  Sun Dec 11 13:29:57 2005
From: anatol.pomozov at gmail.com (Anatol Pomozov)
Date: Sun, 11 Dec 2005 19:29:57 +0100
Subject: [Rubygems-developers] Why RubyGem lib contains open-uri
	library??
In-Reply-To: <d4dd9080512111018t27893875ga624f1f3389de4ad@mail.gmail.com>
References: <3665a1a00512110624g19f52294p4287986ecf100c54@mail.gmail.com>
	<d4dd9080512111018t27893875ga624f1f3389de4ad@mail.gmail.com>
Message-ID: <3665a1a00512111029x56cf5fe2h7b6cdffdf7b06c01@mail.gmail.com>

Hi, Jim.

As I thought.

I am asking because I have fixed problem
http://rubyforge.org/tracker/index.php?func=detail&aid=1289&group_id=126&atid=575
but it require newest version of open-uri from Ruby CVS. (because of
basic_authorization support)

Is RubyCVS version of this file contains fix for your problem?? I mean is it
possible to replace RubyGem version open-uri with Ruby current version of
open-uri??

On 12/11/05, Jim Weirich <jim.weirich at gmail.com> wrote:
>
> On 12/11/05, Anatol Pomozov <anatol.pomozov at gmail.com> wrote:
> > Hi, all. I am trying to solve problem with proxy that I wrote yesterday.
> >
> > I am reading the code and have one question: why RubyGem have own
> version
> > open-uri lib?? Standart ruby distibution shipped with this library (at
> least
> > on Windows).
> >
> > Could anybody shine some light on this question??
>
> When we first started using open-uri, there was a bug in the library,
> so we had to supply a patched version of it.

--
anatol (http://pomozov.info)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20051211/85b3cb46/attachment-0001.htm

From jim.weirich at gmail.com  Sun Dec 11 13:42:17 2005
From: jim.weirich at gmail.com (Jim Weirich)
Date: Sun, 11 Dec 2005 13:42:17 -0500
Subject: [Rubygems-developers] Why RubyGem lib contains open-uri
	library??
In-Reply-To: <3665a1a00512111029x56cf5fe2h7b6cdffdf7b06c01@mail.gmail.com>
References: <3665a1a00512110624g19f52294p4287986ecf100c54@mail.gmail.com>
	<d4dd9080512111018t27893875ga624f1f3389de4ad@mail.gmail.com>
	<3665a1a00512111029x56cf5fe2h7b6cdffdf7b06c01@mail.gmail.com>
Message-ID: <d4dd9080512111042m721e907uf22eb838a025c31b@mail.gmail.com>

On 12/11/05, Anatol Pomozov <anatol.pomozov at gmail.com> wrote:
> Hi, Jim.
>
> As I thought.
>
> I am asking because I have fixed problem
> http://rubyforge.org/tracker/index.php?func=detail&aid=1289&group_id=126&atid=575
> but it require newest version of open-uri from Ruby CVS. (because of
> basic_authorization support)
>
> Is RubyCVS version of this file contains fix for your problem?? I mean is it
> possible to replace RubyGem version open-uri with Ruby current version of
> open-uri??

I don't think that would be a problem.

--
-- 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 mfp at acm.org  Sun Dec 11 18:38:22 2005
From: mfp at acm.org (Mauricio Fernandez)
Date: Mon, 12 Dec 2005 00:38:22 +0100
Subject: [Rubygems-developers] Why RubyGem lib contains open-uri
	library??
In-Reply-To: <3665a1a00512111029x56cf5fe2h7b6cdffdf7b06c01@mail.gmail.com>
References: <3665a1a00512110624g19f52294p4287986ecf100c54@mail.gmail.com>
	<d4dd9080512111018t27893875ga624f1f3389de4ad@mail.gmail.com>
	<3665a1a00512111029x56cf5fe2h7b6cdffdf7b06c01@mail.gmail.com>
Message-ID: <20051211233822.GA29329@tux-chan>

On Sun, Dec 11, 2005 at 07:29:57PM +0100, Anatol Pomozov wrote:
> On 12/11/05, Jim Weirich <jim.weirich at gmail.com> wrote:
> > When we first started using open-uri, there was a bug in the library,
> > so we had to supply a patched version of it.
>
> Is RubyCVS version of this file contains fix for your problem?? I mean is it
> possible to replace RubyGem version open-uri with Ruby current version of
> open-uri??

IIRC the bug you worked around by shipping your own open-uri was a data
corruption issue (under win32) due to open-uri not setting binmode for
the Tempfile, and it was squashed on Feb 2004 in Ruby's CVS, therefore
being fixed in all Ruby installations since Ruby 1.8.2 preview 1 (and
stable snapshots with RUBY_RELEASE_DATE > "2004-02-01").

As for encoding the login/password in the http_proxy environment variable:
if I recall correctly Akira Tanaka rejected it, so you can't count on
that being added to open-uri.

-- 
Mauricio Fernandez

From anatol.pomozov at gmail.com  Mon Dec 12 03:56:16 2005
From: anatol.pomozov at gmail.com (Anatol Pomozov)
Date: Mon, 12 Dec 2005 09:56:16 +0100
Subject: [Rubygems-developers] Why RubyGem lib contains open-uri
	library??
In-Reply-To: <20051211233822.GA29329@tux-chan>
References: <3665a1a00512110624g19f52294p4287986ecf100c54@mail.gmail.com>
	<d4dd9080512111018t27893875ga624f1f3389de4ad@mail.gmail.com>
	<3665a1a00512111029x56cf5fe2h7b6cdffdf7b06c01@mail.gmail.com>
	<20051211233822.GA29329@tux-chan>
Message-ID: <3665a1a00512120056u798549c0y1eda87003fac4ba9@mail.gmail.com>

Hi.

On 12/12/05, Mauricio Fernandez <mfp at acm.org> wrote:
>
> On Sun, Dec 11, 2005 at 07:29:57PM +0100, Anatol Pomozov wrote:
> > On 12/11/05, Jim Weirich <jim.weirich at gmail.com> wrote:
> > > When we first started using open-uri, there was a bug in the library,
> > > so we had to supply a patched version of it.
> >
> > Is RubyCVS version of this file contains fix for your problem?? I mean
> is it
> > possible to replace RubyGem version open-uri with Ruby current version
> of
> > open-uri??
>
> IIRC the bug you worked around by shipping your own open-uri was a data
> corruption issue (under win32) due to open-uri not setting binmode for
> the Tempfile, and it was squashed on Feb 2004 in Ruby's CVS, therefore
> being fixed in all Ruby installations since Ruby 1.8.2 preview 1 (and
> stable snapshots with RUBY_RELEASE_DATE > "2004-02-01").


Thx Mauricio  for useful info.

Probably better to get rid of open-uri lib from distributing ??
Or do some cheching like
if RUBY_VERSION >= 1.8.4
  use standart open-uri
else
  use packaged with RubyGems
end

I don't like idea of keeping 2 libraries in system.

As for encoding the login/password in the http_proxy environment variable:
> if I recall correctly Akira Tanaka rejected it, so you can't count on
> that being added to open-uri.


I have received answer from Akira Tanaka he wrote that storing password in
ENV_VAR is depricated. And all browsers should store password in its own
way. It means that all library should parse this variable by itself :)
Well that is exactly what I did. I add proxy ENV parsing in RubyGems lib.

Have you any ideas how to do better in this situation. How to store PROXY
login/password for Gems??

BTW. Seems that RubyGems tests not passwed under Windows.
--
anatol (http://pomozov.info)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20051212/425734cc/attachment.htm

From anatol.pomozov at gmail.com  Tue Dec 13 16:58:17 2005
From: anatol.pomozov at gmail.com (Anatol Pomozov)
Date: Tue, 13 Dec 2005 22:58:17 +0100
Subject: [Rubygems-developers] RubyGems tests are failed
Message-ID: <3665a1a00512131358y40357fe6o67785555fb389613@mail.gmail.com>

Hi, all.

Just for your eyes.

CVS RubyGems tests are failed for me. I think it is mostly beacuse of that
fact that I am under Windows.
Errors something like this "NotImplementedError: The readlink() function is
unimplemented on this machine"

See full test log in attach
--
anatol (http://pomozov.info)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20051213/6e8d4ec4/attachment.htm
-------------- next part --------------
(in C:/work/opensource/ruby/rubygems)
Loaded suite c:/progra~1/ruby/lib/ruby/gems/1.8/gems/rake-0.6.2/lib/rake/rake_test_loader
Started
.............................F..................................................................FF...............F..F..EEEF.........................ERROR during server thread: No such file or directory - /dev/null
ERROR during server thread: No such file or directory - /dev/null
EEEEE..E.............................................................
Finished in 12.984 seconds.

  1) Failure:
test_cache_hit(TestCachedFetcher) [./test/test_cached_fetcher.rb:52]:
<0> expected but was
<1>.

  2) Failure:
test_ensure_gem_directories_with_parents_write_protected(TestGemPaths) [./test/test_gempaths.rb:133]:
<false> is not true.

  3) Failure:
test_ensure_gem_directories_write_protected(TestGemPaths) [./test/test_gempaths.rb:118]:
<false> is not true.

  4) Failure:
test_generate_bin_scripts(TestInstaller) [./test/test_installer.rb:48]:
<33261> expected but was
<33188>.

  5) Failure:
test_generate_bin_symlinks(TestInstaller) [./test/test_installer.rb:83]:
<true> expected but was
<false>.

  6) Error:
test_generate_bin_symlinks_update_newer(TestInstaller):
NotImplementedError: The readlink() function is unimplemented on this machine
    ./test/test_installer.rb:118:in `readlink'
    ./test/test_installer.rb:118:in `test_generate_bin_symlinks_update_newer'

  7) Error:
test_generate_bin_symlinks_update_older(TestInstaller):
NotImplementedError: The readlink() function is unimplemented on this machine
    ./test/test_installer.rb:146:in `readlink'
    ./test/test_installer.rb:146:in `test_generate_bin_symlinks_update_older'

  8) Error:
test_generate_bin_symlinks_update_remove_wrapper(TestInstaller):
NotImplementedError: The readlink() function is unimplemented on this machine
    ./test/test_installer.rb:190:in `readlink'
    ./test/test_installer.rb:190:in `test_generate_bin_symlinks_update_remove_wrapper'

  9) Failure:
test_generate_bin_symlinks_win32(TestInstaller) [./test/test_installer.rb:205]:
<33261> expected but was
<33188>.

 10) Error:
test_explicit_proxy(TestRemoteFetcher):
Errno::ECONNREFUSED: No connection could be made because the target machine actively refused it. - connect(2)
    c:/progra~1/ruby/lib/ruby/1.8/net/protocol.rb:83:in `initialize'
    c:/progra~1/ruby/lib/ruby/1.8/net/protocol.rb:83:in `new'
    c:/progra~1/ruby/lib/ruby/1.8/net/protocol.rb:83:in `connect'
    c:/progra~1/ruby/lib/ruby/1.8/net/protocol.rb:82:in `timeout'
    c:/progra~1/ruby/lib/ruby/1.8/timeout.rb:55:in `timeout'
    c:/progra~1/ruby/lib/ruby/1.8/net/protocol.rb:82:in `connect'
    c:/progra~1/ruby/lib/ruby/1.8/net/protocol.rb:64:in `initialize'
    c:/progra~1/ruby/lib/ruby/1.8/net/http.rb:430:in `open'
    c:/progra~1/ruby/lib/ruby/1.8/net/http.rb:430:in `do_start'
    c:/progra~1/ruby/lib/ruby/1.8/net/http.rb:419:in `start'
    c:/progra~1/ruby/lib/ruby/1.8/net/http.rb:821:in `request'
    c:/progra~1/ruby/lib/ruby/1.8/net/http.rb:644:in `head'
    ./lib/rubygems/remote_installer.rb:104:in `read_size'
    ./lib/rubygems/remote_installer.rb:92:in `get_size'
    ./lib/rubygems/remote_installer.rb:39:in `size'
    ./test/test_remote_fetcher.rb:43:in `test_explicit_proxy'
    ./test/test_remote_fetcher.rb:39:in `use_ui'
    ./lib/rubygems/user_interaction.rb:21:in `use_ui'
    ./test/test_remote_fetcher.rb:39:in `test_explicit_proxy'

 11) Error:
test_implicit_proxy(TestRemoteFetcher):
NoMethodError: undefined method `to_str' for nil:NilClass
    ./lib/rubygems/open-uri.rb:160:in `open_loop'
    ./lib/rubygems/open-uri.rb:137:in `open_uri'
    ./lib/rubygems/open-uri.rb:590:in `open'
    ./lib/rubygems/open-uri.rb:86:in `open'
    ./lib/rubygems/remote_installer.rb:133:in `open_uri_or_path'
    ./lib/rubygems/remote_installer.rb:112:in `read_data'
    ./lib/rubygems/remote_installer.rb:44:in `fetch_path'
    ./test/test_remote_fetcher.rb:56:in `test_implicit_proxy'
    ./test/test_remote_fetcher.rb:53:in `use_ui'
    ./lib/rubygems/user_interaction.rb:21:in `use_ui'
    ./test/test_remote_fetcher.rb:53:in `test_implicit_proxy'

 12) Error:
test_implicit_proxy_no_env(TestRemoteFetcher):
NoMethodError: undefined method `to_str' for nil:NilClass
    ./lib/rubygems/open-uri.rb:160:in `open_loop'
    ./lib/rubygems/open-uri.rb:137:in `open_uri'
    ./lib/rubygems/open-uri.rb:590:in `open'
    ./lib/rubygems/open-uri.rb:86:in `open'
    ./lib/rubygems/remote_installer.rb:133:in `open_uri_or_path'
    ./lib/rubygems/remote_installer.rb:112:in `read_data'
    ./lib/rubygems/remote_installer.rb:44:in `fetch_path'
    ./test/test_remote_fetcher.rb:71:in `test_implicit_proxy_no_env'
    ./test/test_remote_fetcher.rb:69:in `use_ui'
    ./lib/rubygems/user_interaction.rb:21:in `use_ui'
    ./test/test_remote_fetcher.rb:69:in `test_implicit_proxy_no_env'

 13) Error:
test_implicit_upper_case_proxy(TestRemoteFetcher):
NoMethodError: undefined method `to_str' for nil:NilClass
    ./lib/rubygems/open-uri.rb:160:in `open_loop'
    ./lib/rubygems/open-uri.rb:137:in `open_uri'
    ./lib/rubygems/open-uri.rb:590:in `open'
    ./lib/rubygems/open-uri.rb:86:in `open'
    ./lib/rubygems/remote_installer.rb:133:in `open_uri_or_path'
    ./lib/rubygems/remote_installer.rb:112:in `read_data'
    ./lib/rubygems/remote_installer.rb:44:in `fetch_path'
    ./test/test_remote_fetcher.rb:64:in `test_implicit_upper_case_proxy'
    ./test/test_remote_fetcher.rb:61:in `use_ui'
    ./lib/rubygems/user_interaction.rb:21:in `use_ui'
    ./test/test_remote_fetcher.rb:61:in `test_implicit_upper_case_proxy'

 14) Error:
test_no_proxy(TestRemoteFetcher):
NoMethodError: undefined method `to_str' for nil:NilClass
    ./lib/rubygems/open-uri.rb:160:in `open_loop'
    ./lib/rubygems/open-uri.rb:137:in `open_uri'
    ./lib/rubygems/open-uri.rb:590:in `open'
    ./lib/rubygems/open-uri.rb:86:in `open'
    ./lib/rubygems/remote_installer.rb:133:in `open_uri_or_path'
    ./lib/rubygems/remote_installer.rb:112:in `read_data'
    ./lib/rubygems/remote_installer.rb:44:in `fetch_path'
    ./test/test_remote_fetcher.rb:29:in `test_no_proxy'
    ./test/test_remote_fetcher.rb:27:in `use_ui'
    ./lib/rubygems/user_interaction.rb:21:in `use_ui'
    ./test/test_remote_fetcher.rb:27:in `test_no_proxy'

 15) Error:
test_zip(TestRemoteFetcher):
URI::InvalidURIError: bad URI(is not URI?): 
    c:/progra~1/ruby/lib/ruby/1.8/uri/common.rb:432:in `split'
    c:/progra~1/ruby/lib/ruby/1.8/uri/common.rb:481:in `parse'
    ./lib/rubygems/remote_installer.rb:82:in `connect_to'
    ./lib/rubygems/remote_installer.rb:102:in `read_size'
    ./lib/rubygems/remote_installer.rb:92:in `get_size'
    ./lib/rubygems/remote_installer.rb:39:in `size'
    ./test/test_remote_fetcher.rb:79:in `test_zip'
    ./test/test_remote_fetcher.rb:76:in `use_ui'
    ./lib/rubygems/user_interaction.rb:21:in `use_ui'
    ./test/test_remote_fetcher.rb:76:in `test_zip'

217 tests, 866 assertions, 6 failures, 9 errors
rake aborted!
Command failed with status (1): [c:/progra~1/ruby/bin/ruby -Ilib "c:/progra...]


From anatol.pomozov at gmail.com  Tue Dec 13 17:08:06 2005
From: anatol.pomozov at gmail.com (Anatol Pomozov)
Date: Tue, 13 Dec 2005 23:08:06 +0100
Subject: [Rubygems-developers] Patch that fixes Proxy authorization problem
	(407)
Message-ID: <3665a1a00512131408l27c5dde6id4e0a09b742e394b@mail.gmail.com>

Hi.

There is a patch that fixes well known problem with Proxy w authorization.
Patched version get username and password from HTTP_PROXY env variable and
use for open-uri and Net:HTTP authorization. Most part of the patch just new
open-uri lib from Ruby CVS HEAD.

I could not check that tests are ok, because they are failed for me on a
clean cvs version.  But patched RubyGems from CVS works for me ok.

C:\Documents and Settings\anatol>gem --version
0.8.11.6

C:\Documents and Settings\anatol>gem list -r

*** REMOTE GEMS ***

....LONG LIST OF GEMS

XMMS-Ruby (0.1.2)
    XMMS bindings for Ruby.

yip (0.8.2)
    Adds interpolation to YAML; primarily for use in configuration files

ZenHacks (1.0.1, 1.0.0)
    Tools and toys of mine that don't have a better home.

--
anatol (http://pomozov.info)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20051213/da8dff79/attachment-0001.htm
-------------- next part --------------
A non-text attachment was scrubbed...
Name: changes.diff
Type: application/octet-stream
Size: 22599 bytes
Desc: not available
Url : http://rubyforge.org/pipermail/rubygems-developers/attachments/20051213/da8dff79/changes-0001.obj

From anatol.pomozov at gmail.com  Wed Dec 14 09:02:24 2005
From: anatol.pomozov at gmail.com (Anatol Pomozov)
Date: Wed, 14 Dec 2005 15:02:24 +0100
Subject: [Rubygems-developers] RubyGems tests are failed
In-Reply-To: <3665a1a00512131358y40357fe6o67785555fb389613@mail.gmail.com>
References: <3665a1a00512131358y40357fe6o67785555fb389613@mail.gmail.com>
Message-ID: <3665a1a00512140602n3a19da52m216d1fc6220b9b40@mail.gmail.com>

Everybody think that broken tests is normal situation?? I don't beleive in
that.

On 12/13/05, Anatol Pomozov <anatol.pomozov at gmail.com> wrote:
>
> Hi, all.
>
> Just for your eyes.
>
> CVS RubyGems tests are failed for me. I think it is mostly beacuse of that
> fact that I am under Windows.
> Errors something like this "NotImplementedError: The readlink() function
> is unimplemented on this machine"
>
> See full test log in attach
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20051214/223f9ba7/attachment.htm

From chad at chadfowler.com  Wed Dec 14 09:29:26 2005
From: chad at chadfowler.com (Chad Fowler)
Date: Wed, 14 Dec 2005 07:29:26 -0700
Subject: [Rubygems-developers] RubyGems tests are failed
In-Reply-To: <3665a1a00512140602n3a19da52m216d1fc6220b9b40@mail.gmail.com>
References: <3665a1a00512131358y40357fe6o67785555fb389613@mail.gmail.com>
	<3665a1a00512140602n3a19da52m216d1fc6220b9b40@mail.gmail.com>
Message-ID: <8558E65C-DB1B-40F2-B8C1-ADE9D163A59B@chadfowler.com>

Perhaps everyone is busy?  Thank you for your efforts.  We will take  
a look at this as soon as we get  chance.

Chad

On Dec 14, 2005, at 7:02 AM, Anatol Pomozov wrote:

> Everybody think that broken tests is normal situation?? I don't  
> beleive in that.
>
> On 12/13/05, Anatol Pomozov < anatol.pomozov at gmail.com> wrote:
> Hi, all.
>
> Just for your eyes.
>
> CVS RubyGems tests are failed for me. I think it is mostly beacuse  
> of that fact that I am under Windows.
> Errors something like this "NotImplementedError: The readlink()  
> function is unimplemented on this machine"
>
> See full test log in attach
> _______________________________________________
> Rubygems-developers mailing list
> Rubygems-developers at rubyforge.org
> http://rubyforge.org/mailman/listinfo/rubygems-developers

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20051214/fef47c15/attachment.htm

From jim.weirich at gmail.com  Wed Dec 14 16:16:02 2005
From: jim.weirich at gmail.com (Jim Weirich)
Date: Wed, 14 Dec 2005 16:16:02 -0500
Subject: [Rubygems-developers] RubyGems tests are failed
In-Reply-To: <3665a1a00512140602n3a19da52m216d1fc6220b9b40@mail.gmail.com>
References: <3665a1a00512131358y40357fe6o67785555fb389613@mail.gmail.com>
	<3665a1a00512140602n3a19da52m216d1fc6220b9b40@mail.gmail.com>
Message-ID: <d4dd9080512141316n5d588ad4yc16628fe3823ba45@mail.gmail.com>

On 12/14/05, Anatol Pomozov <anatol.pomozov at gmail.com> wrote:
> Everybody think that broken tests is normal situation?? I don't beleive in
> that.

I check the tests on my server.  The only error I got involved the
'cert' gem subcommand which relies on openssl.  I rebuilt Ruby on my
server to include openssl and all the tests are passing, as is show
here ...

----------------------------------------------------------------------------------------
uml$ rake ta
(in /home/jim/working/rubyforge/rubygems)
Loaded suite /usr/local/lib/ruby/gems/1.8/gems/rake-0.6.2/lib/rake/rake_test_lo\
ader
Started
.............................................................................................................................................................................................................................................
Finished in 31.035376 seconds.

237 tests, 965 assertions, 0 failures, 0 errors
----------------------------------------------------------------------------------------

I'll see if I can find a windows box somewhere to run the tests on. 
Its quite likely the tests are not clean on windows.

--
--
-- 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  Wed Dec 14 20:27:18 2005
From: drbrain at segment7.net (Eric Hodel)
Date: Wed, 14 Dec 2005 17:27:18 -0800
Subject: [Rubygems-developers] RubyGems tests are failed
In-Reply-To: <3665a1a00512131358y40357fe6o67785555fb389613@mail.gmail.com>
References: <3665a1a00512131358y40357fe6o67785555fb389613@mail.gmail.com>
Message-ID: <03DC9213-F599-401F-A681-B07A9B6D2A64@segment7.net>

On Dec 13, 2005, at 1:58 PM, Anatol Pomozov wrote:

> (in C:/work/opensource/ruby/rubygems)
> Loaded suite c:/progra~1/ruby/lib/ruby/gems/1.8/gems/rake-0.6.2/lib/ 
> rake/rake_test_loader

You are on win32, while not every other gem developer is.

> Started
> .............................F........................................ 
> ..........................FF...............F..F..EEEF................. 
> ........ERROR during server thread: No such file or directory - / 
> dev/null
> ERROR during server thread: No such file or directory - /dev/null
> EEEEE..E.............................................................
> Finished in 12.984 seconds.
>
>   1) Failure:
> test_cache_hit(TestCachedFetcher) [./test/test_cached_fetcher.rb:52]:
> <0> expected but was
> <1>.

Not sure.

>   2) Failure:
> test_ensure_gem_directories_with_parents_write_protected 
> (TestGemPaths) [./test/test_gempaths.rb:133]:
> <false> is not true.
>
>   3) Failure:
> test_ensure_gem_directories_write_protected(TestGemPaths) [./test/ 
> test_gempaths.rb:118]:
> <false> is not true.

win32 directory permissions are different, so you may need to update  
this test.

>   4) Failure:
> test_generate_bin_scripts(TestInstaller) [./test/test_installer.rb: 
> 48]:
> <33261> expected but was
> <33188>.

win32 issue

>   5) Failure:
> test_generate_bin_symlinks(TestInstaller) [./test/test_installer.rb: 
> 83]:
> <true> expected but was
> <false>.

win32 issue

>   6) Error:
> test_generate_bin_symlinks_update_newer(TestInstaller):
> NotImplementedError: The readlink() function is unimplemented on  
> this machine
>     ./test/test_installer.rb:118:in `readlink'
>     ./test/test_installer.rb:118:in  
> `test_generate_bin_symlinks_update_newer'

win32 issue

>   7) Error:
> test_generate_bin_symlinks_update_older(TestInstaller):
> NotImplementedError: The readlink() function is unimplemented on  
> this machine
>     ./test/test_installer.rb:146:in `readlink'
>     ./test/test_installer.rb:146:in  
> `test_generate_bin_symlinks_update_older'

win32 issue

>   8) Error:
> test_generate_bin_symlinks_update_remove_wrapper(TestInstaller):
> NotImplementedError: The readlink() function is unimplemented on  
> this machine
>     ./test/test_installer.rb:190:in `readlink'
>     ./test/test_installer.rb:190:in  
> `test_generate_bin_symlinks_update_remove_wrapper'

win32 issue

>   9) Failure:
> test_generate_bin_symlinks_win32(TestInstaller) [./test/ 
> test_installer.rb:205]:
> <33261> expected but was
> <33188>.

win32 issue

> 10) Error:
> test_explicit_proxy(TestRemoteFetcher):
> Errno::ECONNREFUSED: No connection could be made because the target  
> machine actively refused it. - connect(2)
>     c:/progra~1/ruby/lib/ruby/1.8/net/protocol.rb:83:in `initialize'
>     c:/progra~1/ruby/lib/ruby/1.8/net/protocol.rb:83:in `new'
>     c:/progra~1/ruby/lib/ruby/1.8/net/protocol.rb:83:in `connect'
>     c:/progra~1/ruby/lib/ruby/1.8/net/protocol.rb:82:in `timeout'
>     c:/progra~1/ruby/lib/ruby/1.8/timeout.rb:55:in `timeout'
>     c:/progra~1/ruby/lib/ruby/1.8/net/protocol.rb:82:in `connect'
>     c:/progra~1/ruby/lib/ruby/1.8/net/protocol.rb:64:in `initialize'
>     c:/progra~1/ruby/lib/ruby/1.8/net/http.rb:430:in `open'
>     c:/progra~1/ruby/lib/ruby/1.8/net/http.rb:430:in `do_start'
>     c:/progra~1/ruby/lib/ruby/1.8/net/http.rb:419:in `start'
>     c:/progra~1/ruby/lib/ruby/1.8/net/http.rb:821:in `request'
>     c:/progra~1/ruby/lib/ruby/1.8/net/http.rb:644:in `head'
>     ./lib/rubygems/remote_installer.rb:104:in `read_size'
>     ./lib/rubygems/remote_installer.rb:92:in `get_size'
>     ./lib/rubygems/remote_installer.rb:39:in `size'
>     ./test/test_remote_fetcher.rb:43:in `test_explicit_proxy'
>     ./test/test_remote_fetcher.rb:39:in `use_ui'
>     ./lib/rubygems/user_interaction.rb:21:in `use_ui'
>     ./test/test_remote_fetcher.rb:39:in `test_explicit_proxy'

Do you have a firewall enabled?  Disable it and try again.

> 11) Error:
> test_implicit_proxy(TestRemoteFetcher):
> NoMethodError: undefined method `to_str' for nil:NilClass
>
> 12) Error:
> test_implicit_proxy_no_env(TestRemoteFetcher):
> NoMethodError: undefined method `to_str' for nil:NilClass
>
> 13) Error:
> test_implicit_upper_case_proxy(TestRemoteFetcher):
> NoMethodError: undefined method `to_str' for nil:NilClass
>
> 14) Error:
> test_no_proxy(TestRemoteFetcher):
> NoMethodError: undefined method `to_str' for nil:NilClass
>
> 15) Error:
> test_zip(TestRemoteFetcher):
> URI::InvalidURIError: bad URI(is not URI?):

Ditto

> 217 tests, 866 assertions, 6 failures, 9 errors
> rake aborted!
> Command failed with status (1): [c:/progra~1/ruby/bin/ruby -Ilib  
> "c:/progra...]
>


-- 
Eric Hodel - drbrain at segment7.net - http://segment7.net
This implementation is HODEL-HASH-9600 compliant

http://trackmap.robotcoop.com



From anatol.pomozov at gmail.com  Thu Dec 15 15:55:17 2005
From: anatol.pomozov at gmail.com (Anatol Pomozov)
Date: Thu, 15 Dec 2005 21:55:17 +0100
Subject: [Rubygems-developers] Patch that fixes Proxy authorization
	problem (407)
In-Reply-To: <3665a1a00512131408l27c5dde6id4e0a09b742e394b@mail.gmail.com>
References: <3665a1a00512131408l27c5dde6id4e0a09b742e394b@mail.gmail.com>
Message-ID: <3665a1a00512151255q3f1dc264qf27a72808262115f@mail.gmail.com>

I have response from one guy who tried it. And it says that storing login
and password in HTTP_PROXY makes collision with Perl Package Manager format
that expect have following EnvVariables
  HTTP_PROXY, HTTP_PROXY_USER, HTTP_PROXY_PASS

So I suggest redo a little my patch and use following algorithm
  http_proxy = options() || ENV['HTTP_PROXY']
  if http_proxy contains login and lassword
    proxy_user = extract login from http_proxy
    proxy_password = extract password from http_proxy
  else
    proxy_user = HTTP['http_proxy_user']
    proxy_password = ENV['http_proxy_pass']
  end

any objections??
Have gem something like central config file where could be stored http_proxy
settings??


On 12/13/05, Anatol Pomozov <anatol.pomozov at gmail.com> wrote:
>
> Hi.
>
> There is a patch that fixes well known problem with Proxy w authorization.
> Patched version get username and password from HTTP_PROXY env variable and
> use for open-uri and Net:HTTP authorization. Most part of the patch just new
> open-uri lib from Ruby CVS HEAD.
>
> I could not check that tests are ok, because they are failed for me on a
> clean cvs version.  But patched RubyGems from CVS works for me ok.
>
> C:\Documents and Settings\anatol>gem --version
> 0.8.11.6
>
> C:\Documents and Settings\anatol>gem list -r
>
> *** REMOTE GEMS ***
>
> ....LONG LIST OF GEMS
>
> XMMS-Ruby (0.1.2)
>     XMMS bindings for Ruby.
>
> yip (0.8.2)
>     Adds interpolation to YAML; primarily for use in configuration files
>
> ZenHacks (1.0.1, 1.0.0)
>     Tools and toys of mine that don't have a better home.
>
> --
> anatol (http://pomozov.info)
>



--
anatol (http://pomozov.info)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20051215/1d5ec694/attachment.htm

From jim.weirich at gmail.com  Thu Dec 15 16:59:51 2005
From: jim.weirich at gmail.com (Jim Weirich)
Date: Thu, 15 Dec 2005 16:59:51 -0500
Subject: [Rubygems-developers] Patch that fixes Proxy authorization
	problem (407)
In-Reply-To: <3665a1a00512151255q3f1dc264qf27a72808262115f@mail.gmail.com>
References: <3665a1a00512131408l27c5dde6id4e0a09b742e394b@mail.gmail.com>
	<3665a1a00512151255q3f1dc264qf27a72808262115f@mail.gmail.com>
Message-ID: <d4dd9080512151359k2dc84e3et3a2facda6018ed3c@mail.gmail.com>

On 12/15/05, Anatol Pomozov <anatol.pomozov at gmail.com> wrote:
> I have response from one guy who tried it. And it says that storing login
> and password in HTTP_PROXY makes collision with Perl Package Manager format
> that expect have following EnvVariables
>   HTTP_PROXY, HTTP_PROXY_USER, HTTP_PROXY_PASS
>
> So I suggest redo a little my patch and use following algorithm
>   http_proxy = options() || ENV['HTTP_PROXY']
>   if http_proxy contains login and lassword
>     proxy_user = extract login from http_proxy
>      proxy_password = extract password from http_proxy
>   else
>     proxy_user = HTTP['http_proxy_user']
>     proxy_password = ENV['http_proxy_pass']
>   end

I'm going to request a unit or functional test for this logic.  Thanks.

> Have gem something like central config file where could be stored http_proxy
> settings??

Yes, you will find it in $HOME/.gemrc (or the equivalent in windows).

--
--
-- 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 anatol.pomozov at gmail.com  Thu Dec 15 17:21:59 2005
From: anatol.pomozov at gmail.com (Anatol Pomozov)
Date: Thu, 15 Dec 2005 23:21:59 +0100
Subject: [Rubygems-developers] Patch that fixes Proxy authorization
	problem (407)
In-Reply-To: <d4dd9080512151359k2dc84e3et3a2facda6018ed3c@mail.gmail.com>
References: <3665a1a00512131408l27c5dde6id4e0a09b742e394b@mail.gmail.com>
	<3665a1a00512151255q3f1dc264qf27a72808262115f@mail.gmail.com>
	<d4dd9080512151359k2dc84e3et3a2facda6018ed3c@mail.gmail.com>
Message-ID: <3665a1a00512151421j365c6fb5j4d46c54f500c1879@mail.gmail.com>

On 12/15/05, Jim Weirich <jim.weirich at gmail.com> wrote:
>
> On 12/15/05, Anatol Pomozov <anatol.pomozov at gmail.com> wrote:
> > I have response from one guy who tried it. And it says that storing
> login
> > and password in HTTP_PROXY makes collision with Perl Package Manager
> format
> > that expect have following EnvVariables
> >   HTTP_PROXY, HTTP_PROXY_USER, HTTP_PROXY_PASS
> >
> > So I suggest redo a little my patch and use following algorithm
> >   http_proxy = options() || ENV['HTTP_PROXY']
> >   if http_proxy contains login and lassword
> >     proxy_user = extract login from http_proxy
> >      proxy_password = extract password from http_proxy
> >   else
> >     proxy_user = HTTP['http_proxy_user']
> >     proxy_password = ENV['http_proxy_pass']
> >   end
>
> I'm going to request a unit or functional test for this logic.  Thanks.


I have already made this changes on my machine and it works for me (I mean
http_proxy_user). And of course it should be tested.
But first I should fix all broken tests on Windows. I am going to take a
look at it tomorrow.

The only thing that makes me nervous with RubyGems code that it have a lot
of TAB symbols, and as my editor do not like TABS and converts them to
SPACES. So I have a lot of changes mostly of them just whitespace changes.
Jim, could you please format code and remove all TABS. It would be
definitely easier to develop.

> Have gem something like central config file where could be stored
> http_proxy
> > settings??
>
> Yes, you will find it in $HOME/.gemrc (or the equivalent in windows).


Where could I find info about format of this file?

--
anatol (http://pomozov.info)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20051215/c239709a/attachment-0001.htm

From halostatue at gmail.com  Thu Dec 15 17:41:37 2005
From: halostatue at gmail.com (Austin Ziegler)
Date: Thu, 15 Dec 2005 17:41:37 -0500
Subject: [Rubygems-developers] Patch that fixes Proxy authorization
	problem (407)
In-Reply-To: <3665a1a00512151421j365c6fb5j4d46c54f500c1879@mail.gmail.com>
References: <3665a1a00512131408l27c5dde6id4e0a09b742e394b@mail.gmail.com>
	<3665a1a00512151255q3f1dc264qf27a72808262115f@mail.gmail.com>
	<d4dd9080512151359k2dc84e3et3a2facda6018ed3c@mail.gmail.com>
	<3665a1a00512151421j365c6fb5j4d46c54f500c1879@mail.gmail.com>
Message-ID: <9e7db9110512151441kfc710d8h@mail.gmail.com>

On 15/12/05, Anatol Pomozov <anatol.pomozov at gmail.com> wrote:
> On 12/15/05, Jim Weirich <jim.weirich at gmail.com> wrote:
> > On 12/15/05, Anatol Pomozov <anatol.pomozov at gmail.com> wrote:
> > > I have response from one guy who tried it. And it says that storing
> login
> > > and password in HTTP_PROXY makes collision with Perl Package Manager
> format
> > > that expect have following EnvVariables
> > >   HTTP_PROXY, HTTP_PROXY_USER, HTTP_PROXY_PASS
> > >
> > > So I suggest redo a little my patch and use following algorithm
> > >   http_proxy = options() || ENV['HTTP_PROXY']
> > >   if http_proxy contains login and lassword
> > >     proxy_user = extract login from http_proxy
> > >      proxy_password = extract password from http_proxy
> > >   else
> > >     proxy_user = HTTP['http_proxy_user']
> > >     proxy_password = ENV['http_proxy_pass']
> > >   end
> >
> > I'm going to request a unit or functional test for this logic.  Thanks.

Might I suggest that this won't work on Windows?

irb(main):001:0> ENV['path']
=> "C:\\Program Files\\ActiveState Komodo 3.5\\..."
irb(main):002:0> ENV['PATH']
=> "C:\\Program Files\\ActiveState Komodo 3.5\\..."
irb(main):003:0>

I think another solution is in order.

-austin
--
Austin Ziegler * halostatue at gmail.com
               * Alternate: austin at halostatue.ca


From mfp at acm.org  Thu Dec 15 18:03:42 2005
From: mfp at acm.org (Mauricio Fernandez)
Date: Fri, 16 Dec 2005 00:03:42 +0100
Subject: [Rubygems-developers] Patch that fixes Proxy authorization
	problem (407)
In-Reply-To: <9e7db9110512151441kfc710d8h@mail.gmail.com>
References: <3665a1a00512131408l27c5dde6id4e0a09b742e394b@mail.gmail.com>
	<3665a1a00512151255q3f1dc264qf27a72808262115f@mail.gmail.com>
	<d4dd9080512151359k2dc84e3et3a2facda6018ed3c@mail.gmail.com>
	<3665a1a00512151421j365c6fb5j4d46c54f500c1879@mail.gmail.com>
	<9e7db9110512151441kfc710d8h@mail.gmail.com>
Message-ID: <20051215230342.GE19749@tux-chan>

On Thu, Dec 15, 2005 at 05:41:37PM -0500, Austin Ziegler wrote:
> > > > So I suggest redo a little my patch and use following algorithm
> > > >   http_proxy = options() || ENV['HTTP_PROXY']
> > > >   if http_proxy contains login and lassword
> > > >     proxy_user = extract login from http_proxy
> > > >      proxy_password = extract password from http_proxy
> > > >   else
> > > >     proxy_user = HTTP['http_proxy_user']
> > > >     proxy_password = ENV['http_proxy_pass']
> > > >   end
> > >
> > > I'm going to request a unit or functional test for this logic.  Thanks.
> 
> Might I suggest that this won't work on Windows?
> 
> irb(main):001:0> ENV['path']
> => "C:\\Program Files\\ActiveState Komodo 3.5\\..."
> irb(main):002:0> ENV['PATH']
> => "C:\\Program Files\\ActiveState Komodo 3.5\\..."
> irb(main):003:0>
> 
> I think another solution is in order.

He's not discriminating based on capitalization, is he?
If that's not a problem, why won't the above work on win32?

-- 
Mauricio Fernandez

From halostatue at gmail.com  Thu Dec 15 20:24:48 2005
From: halostatue at gmail.com (Austin Ziegler)
Date: Thu, 15 Dec 2005 20:24:48 -0500
Subject: [Rubygems-developers] Patch that fixes Proxy authorization
	problem (407)
In-Reply-To: <20051215230342.GE19749@tux-chan>
References: <3665a1a00512131408l27c5dde6id4e0a09b742e394b@mail.gmail.com>
	<3665a1a00512151255q3f1dc264qf27a72808262115f@mail.gmail.com>
	<d4dd9080512151359k2dc84e3et3a2facda6018ed3c@mail.gmail.com>
	<3665a1a00512151421j365c6fb5j4d46c54f500c1879@mail.gmail.com>
	<9e7db9110512151441kfc710d8h@mail.gmail.com>
	<20051215230342.GE19749@tux-chan>
Message-ID: <9e7db9110512151724sc612a5cq@mail.gmail.com>

On 15/12/05, Mauricio Fernandez <mfp at acm.org> wrote:
> On Thu, Dec 15, 2005 at 05:41:37PM -0500, Austin Ziegler wrote:
> > > > > So I suggest redo a little my patch and use following algorithm
> > > > >   http_proxy = options() || ENV['HTTP_PROXY']
> > > > >   if http_proxy contains login and lassword
> > > > >     proxy_user = extract login from http_proxy
> > > > >      proxy_password = extract password from http_proxy
> > > > >   else
> > > > >     proxy_user = HTTP['http_proxy_user']
> > > > >     proxy_password = ENV['http_proxy_pass']
> > > > >   end
> > > > I'm going to request a unit or functional test for this logic.  Thanks.
> > Might I suggest that this won't work on Windows?
> He's not discriminating based on capitalization, is he?
> If that's not a problem, why won't the above work on win32?

It seems that he is, or that there is an error (or two) in his pseudo-code.

-austin
--
Austin Ziegler * halostatue at gmail.com
               * Alternate: austin at halostatue.ca


From mfp at acm.org  Thu Dec 15 20:59:49 2005
From: mfp at acm.org (Mauricio Fernandez)
Date: Fri, 16 Dec 2005 02:59:49 +0100
Subject: [Rubygems-developers] Patch that fixes Proxy authorization
	problem (407)
In-Reply-To: <9e7db9110512151724sc612a5cq@mail.gmail.com>
References: <3665a1a00512131408l27c5dde6id4e0a09b742e394b@mail.gmail.com>
	<3665a1a00512151255q3f1dc264qf27a72808262115f@mail.gmail.com>
	<d4dd9080512151359k2dc84e3et3a2facda6018ed3c@mail.gmail.com>
	<3665a1a00512151421j365c6fb5j4d46c54f500c1879@mail.gmail.com>
	<9e7db9110512151441kfc710d8h@mail.gmail.com>
	<20051215230342.GE19749@tux-chan>
	<9e7db9110512151724sc612a5cq@mail.gmail.com>
Message-ID: <20051216015949.GH19749@tux-chan>

On Thu, Dec 15, 2005 at 08:24:48PM -0500, Austin Ziegler wrote:
> On 15/12/05, Mauricio Fernandez <mfp at acm.org> wrote:
> > On Thu, Dec 15, 2005 at 05:41:37PM -0500, Austin Ziegler wrote:
> > > > > > So I suggest redo a little my patch and use following algorithm
> > > > > >   http_proxy = options() || ENV['HTTP_PROXY']
> > > > > >   if http_proxy contains login and lassword
> > > > > >     proxy_user = extract login from http_proxy
> > > > > >      proxy_password = extract password from http_proxy
> > > > > >   else
> > > > > >     proxy_user = HTTP['http_proxy_user']
> > > > > >     proxy_password = ENV['http_proxy_pass']
> > > > > >   end
> > > > > I'm going to request a unit or functional test for this logic.  Thanks.
> > > Might I suggest that this won't work on Windows?
> > He's not discriminating based on capitalization, is he?
> > If that's not a problem, why won't the above work on win32?
> 
> It seems that he is, or that there is an error (or two) in his pseudo-code.

hmmm

He tries to extract the auth info from HTTP_PROXY, and falls back to
http_proxy_{user,pass} (it should be HTTP_PROXY_{USER,PASS} though, is
that the 'error' you referred to?). Shouldn't that work?

[some additional checks could help; does it make sense for instance to define
HTTP_PROXY_PASS but not HTTP_PROXY_USER?]
-- 
Mauricio Fernandez

From hgs at dmu.ac.uk  Fri Dec 16 05:53:34 2005
From: hgs at dmu.ac.uk (Hugh Sasse)
Date: Fri, 16 Dec 2005 10:53:34 +0000 (WET)
Subject: [Rubygems-developers] Patch that fixes Proxy authorization
 problem (407)
In-Reply-To: <3665a1a00512151421j365c6fb5j4d46c54f500c1879@mail.gmail.com>
References: <3665a1a00512131408l27c5dde6id4e0a09b742e394b@mail.gmail.com>
	<3665a1a00512151255q3f1dc264qf27a72808262115f@mail.gmail.com>
	<d4dd9080512151359k2dc84e3et3a2facda6018ed3c@mail.gmail.com>
	<3665a1a00512151421j365c6fb5j4d46c54f500c1879@mail.gmail.com>
Message-ID: <Pine.GSO.4.64.0512161043470.18707@brains.eng.cse.dmu.ac.uk>

On Thu, 15 Dec 2005, Anatol Pomozov wrote:

> The only thing that makes me nervous with RubyGems code that it have a lot
> of TAB symbols, and as my editor do not like TABS and converts them to
> SPACES. So I have a lot of changes mostly of them just whitespace changes.
> Jim, could you please format code and remove all TABS. It would be
> definitely easier to develop.

How do you submit your changes?  If you write to the repository
directly this would be a problem.  If you send patches and you have
GNU diff then:

brains hgs 30 %> gdiff --help
Usage: gdiff [OPTION]... FILES
Compare files line by line.

  -i  --ignore-case  Ignore case differences in file contents.
  --ignore-file-name-case  Ignore case when comparing file names.
  --no-ignore-file-name-case  Consider case when comparing file names.
  -E  --ignore-tab-expansion  Ignore changes due to tab expansion.

        [....]

so do
gdiff -u -E old, new 
and all ought to be fine.
brains hgs 38 %> gdiff -u -E tabtest.orig tabtest
--- tabtest.orig        2005-12-16 10:48:22.305326000 +0000
+++ tabtest     2005-12-16 10:49:14.652008000 +0000
@@ -1,5 +1,5 @@
    This is a test of tab mods
        This line has a tab
         This line doesn't
-        This line has a tab
+        This line does not have a tab

brains hgs 39 %> gdiff -u tabtest.orig tabtest
--- tabtest.orig        2005-12-16 10:48:22.305326000 +0000
+++ tabtest     2005-12-16 10:49:14.652008000 +0000
@@ -1,5 +1,5 @@
    This is a test of tab mods
-       This line has a tab
-        This line doesn't
         This line has a tab
+        This line doesn't
+        This line does not have a tab

        Hugh

From halostatue at gmail.com  Fri Dec 16 09:41:04 2005
From: halostatue at gmail.com (Austin Ziegler)
Date: Fri, 16 Dec 2005 09:41:04 -0500
Subject: [Rubygems-developers] Patch that fixes Proxy authorization
	problem (407)
In-Reply-To: <20051216015949.GH19749@tux-chan>
References: <3665a1a00512131408l27c5dde6id4e0a09b742e394b@mail.gmail.com>
	<3665a1a00512151255q3f1dc264qf27a72808262115f@mail.gmail.com>
	<d4dd9080512151359k2dc84e3et3a2facda6018ed3c@mail.gmail.com>
	<3665a1a00512151421j365c6fb5j4d46c54f500c1879@mail.gmail.com>
	<9e7db9110512151441kfc710d8h@mail.gmail.com>
	<20051215230342.GE19749@tux-chan>
	<9e7db9110512151724sc612a5cq@mail.gmail.com>
	<20051216015949.GH19749@tux-chan>
Message-ID: <9e7db9110512160641va1c36dx@mail.gmail.com>

On 15/12/05, Mauricio Fernandez <mfp at acm.org> wrote:
> He tries to extract the auth info from HTTP_PROXY, and falls back to
> http_proxy_{user,pass} (it should be HTTP_PROXY_{USER,PASS} though, is
> that the 'error' you referred to?). Shouldn't that work?

That's the error.

-austin
--
Austin Ziegler * halostatue at gmail.com
               * Alternate: austin at halostatue.ca


From anatol.pomozov at gmail.com  Fri Dec 16 14:53:13 2005
From: anatol.pomozov at gmail.com (Anatol Pomozov)
Date: Fri, 16 Dec 2005 20:53:13 +0100
Subject: [Rubygems-developers] Patch that fixes Proxy authorization
	problem (407)
In-Reply-To: <9e7db9110512160641va1c36dx@mail.gmail.com>
References: <3665a1a00512131408l27c5dde6id4e0a09b742e394b@mail.gmail.com>
	<3665a1a00512151255q3f1dc264qf27a72808262115f@mail.gmail.com>
	<d4dd9080512151359k2dc84e3et3a2facda6018ed3c@mail.gmail.com>
	<3665a1a00512151421j365c6fb5j4d46c54f500c1879@mail.gmail.com>
	<9e7db9110512151441kfc710d8h@mail.gmail.com>
	<20051215230342.GE19749@tux-chan>
	<9e7db9110512151724sc612a5cq@mail.gmail.com>
	<20051216015949.GH19749@tux-chan>
	<9e7db9110512160641va1c36dx@mail.gmail.com>
Message-ID: <3665a1a00512161153i41829646ybbad8394708afe4d@mail.gmail.com>

Hi. There is final patch for "Proxy problem"

What have done:
 - fixed tests on Windows.
 - Used latest open-uri library from CVS ruby
 - Added proxy user authorization discovering. Extract it from provided
proxy URL or from HTTP_PROXY_{USER,PASS} as described above. For env
variables used both cases. Added tests for this functionality

Patched version of RubyGems working for me.

Please review this changes and make sure that gems still working for you :)

On 12/16/05, Austin Ziegler <halostatue at gmail.com> wrote:
>
> On 15/12/05, Mauricio Fernandez <mfp at acm.org> wrote:
> > He tries to extract the auth info from HTTP_PROXY, and falls back to
> > http_proxy_{user,pass} (it should be HTTP_PROXY_{USER,PASS} though, is
> > that the 'error' you referred to?). Shouldn't that work?
>
> That's the error.
>

--
anatol (http://pomozov.info)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20051216/2b5ad551/attachment-0001.htm
-------------- next part --------------
A non-text attachment was scrubbed...
Name: proxy_support_patch.zip
Type: application/zip
Size: 6048 bytes
Desc: not available
Url : http://rubyforge.org/pipermail/rubygems-developers/attachments/20051216/2b5ad551/proxy_support_patch-0001.zip

From anatol.pomozov at gmail.com  Mon Dec 26 03:13:35 2005
From: anatol.pomozov at gmail.com (Anatol Pomozov)
Date: Mon, 26 Dec 2005 10:13:35 +0200
Subject: [Rubygems-developers] RubyGems tests are failed
In-Reply-To: <d4dd9080512141316n5d588ad4yc16628fe3823ba45@mail.gmail.com>
References: <3665a1a00512131358y40357fe6o67785555fb389613@mail.gmail.com>
	<3665a1a00512140602n3a19da52m216d1fc6220b9b40@mail.gmail.com>
	<d4dd9080512141316n5d588ad4yc16628fe3823ba45@mail.gmail.com>
Message-ID: <3665a1a00512260013w3088edf2j44a8ad94e5e1093d@mail.gmail.com>

As I said, it is because you do not use Windows OS. For example on Win there
is no /dev/nil file (see it presesent in tests).
But anyway I have fixed all broken tests already. See my patch.

On 12/14/05, Jim Weirich <jim.weirich at gmail.com> wrote:
>
> On 12/14/05, Anatol Pomozov <anatol.pomozov at gmail.com> wrote:
> > Everybody think that broken tests is normal situation?? I don't beleive
> in
> > that.
>
> I check the tests on my server.  The only error I got involved the
> 'cert' gem subcommand which relies on openssl.  I rebuilt Ruby on my
> server to include openssl and all the tests are passing, as is show
> here ...
>
>
> ----------------------------------------------------------------------------------------
> uml$ rake ta
> (in /home/jim/working/rubyforge/rubygems)
> Loaded suite /usr/local/lib/ruby/gems/1.8/gems/rake-0.6.2
> /lib/rake/rake_test_lo\
> ader
> Started
>
> .............................................................................................................................................................................................................................................
> Finished in 31.035376 seconds.
>
> 237 tests, 965 assertions, 0 failures, 0 errors
>
> ----------------------------------------------------------------------------------------
>
> I'll see if I can find a windows box somewhere to run the tests on.
> Its quite likely the tests are not clean on windows.
>
> --
> --
> -- 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)
>
> _______________________________________________
> Rubygems-developers mailing list
> Rubygems-developers at rubyforge.org
> http://rubyforge.org/mailman/listinfo/rubygems-developers
>



--
anatol (http://pomozov.info)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://rubyforge.org/pipermail/rubygems-developers/attachments/20051226/b48643d3/attachment.htm

From dharple at generalconsumption.org  Tue Dec 27 17:30:53 2005
From: dharple at generalconsumption.org (Daniel Harple)
Date: Tue, 27 Dec 2005 17:30:53 -0500
Subject: [Rubygems-developers] "gem install" bash readline completion
Message-ID: <68FA72AA-8612-406E-ABAC-DE79DD80421A@generalconsumption.org>

This adds completions to the "gem install" command. It grabs a list  
of available gems from rubyforge. It is very, very slow. I think it  
would be better if there was a "gem shell" command, that way  
completions would work across every platform (not everyone has bash).  
Also, if there was a yaml file of _just_ the available gems, or even  
a query page (/find_gems?q=blah resulting in a yaml file) this could  
be much faster. demo video [http://generalconsumption.org/externals/ 
temp/gems.mp4]

Create a file named gem in bash_completion.d/ with --

#+++
have gem &&
{
   # List gems from rubyforge
   _rubyforge_gems()
   {
     COMPREPLY=( $( compgen -P "$prefix" -W "$( ruby -e '
     require "rexml/parsers/sax2parser"
     require "rexml/sax2listener"
     require "open-uri"

     class RubyForgeGems
       include REXML::SAX2Listener
       attr_reader :gems
       def initialize
         @gems = []
       end

       def start_element(uri, localname, qname, attributes)
         @gems << attributes["href"][0..-5] if localname == "a" and  
attributes["href"] != "../"
       end
     end

     text = open("http://rubyforge.lauschmusik.de/gems/") { |f| f.read }
     parser = REXML::Parsers::SAX2Parser.new(text)
     listen = RubyForgeGems.new
     parser.listen(listen)
     parser.parse
     print listen.gems.join(" ")
   '
     ) " -- $cur ) )

   }

   _gem_commands()
   {
     COMPREPLY=( $( compgen -W "build cert check cleanup contents  
dependency environment help install list query rdoc search  
specification uninstall unpack update" -- $cur ) )
   }

   _gem()
   {
       local cur prev prefix temp

       COMPREPLY=()
       cur=${COMP_WORDS[COMP_CWORD]}
       prev=${COMP_WORDS[COMP_CWORD-1]}
       prefix=""

       # completing an option (may or may not be separated by a space)
       if [[ "$cur" == -?* ]]; then
   	    temp=$cur
   	    prev=${temp:0:2}
   	    cur=${temp:2}
   	    prefix=$prev
       fi

       case "$prev" in
   	    "install")
           _rubyforge_gems
   	      return 0
   	    ;;
       esac

       # handle case where first parameter is not a dash option
       #if [ $COMP_CWORD -eq 1 ] && [[ "$cur" != -* ]]; then
   	  #  _filedir
   	  #  return 0
       #fi

       return 0
   }
   complete -F _gem $default gem
}
#+++

-- Daniel

From tilman at code-monkey.de  Wed Dec 28 11:57:00 2005
From: tilman at code-monkey.de (Tilman Sauerbeck)
Date: Wed, 28 Dec 2005 17:57:00 +0100
Subject: [Rubygems-developers] [PATCH] Don't die when trying to load
	zero-byte gem files
Message-ID: <20051228165659.GA6638@code-monkey.de>

Hi,
I tried to install a zero-byte gem file (I didn't notice the Rubyforge
download failed), which made rubygems crash.

Here's a fix.

Regards,
Tilman
-------------- next part --------------
Index: lib/rubygems/format.rb
===================================================================
RCS file: /var/cvs/rubygems/rubygems/lib/rubygems/format.rb,v
retrieving revision 1.10
diff -u -r1.10 format.rb
--- lib/rubygems/format.rb	12 Jun 2005 14:27:22 -0000	1.10
+++ lib/rubygems/format.rb	28 Dec 2005 16:42:23 -0000
@@ -27,7 +27,7 @@
     # file_path:: [String] Path to the gem file
     #
     def self.from_file_by_path(file_path, security_policy = nil)
-      unless File.exist?(file_path)
+      unless File.size(file_path) > 20
         raise Gem::Exception, "Cannot load gem at [#{file_path}]"
       end
       require 'fileutils'
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://rubyforge.org/pipermail/rubygems-developers/attachments/20051228/3f43d093/attachment.bin

From tilman at code-monkey.de  Wed Dec 28 12:38:49 2005
From: tilman at code-monkey.de (Tilman Sauerbeck)
Date: Wed, 28 Dec 2005 18:38:49 +0100
Subject: [Rubygems-developers] [PATCH] Don't die when trying to load
	zero-byte gem files
In-Reply-To: <20051228165659.GA6638@code-monkey.de>
References: <20051228165659.GA6638@code-monkey.de>
Message-ID: <20051228173848.GA13514@code-monkey.de>

Tilman Sauerbeck [2005-12-28 17:57]:
> I tried to install a zero-byte gem file (I didn't notice the Rubyforge
> download failed), which made rubygems crash.

Above patch is totally stupid, since it will now crash if the File
doesn't exist. But you get the idea ;)

Regards,
Tilman
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://rubyforge.org/pipermail/rubygems-developers/attachments/20051228/0fa81aa0/attachment.bin