From djberg96 at gmail.com  Thu Dec  2 10:19:28 2010
From: djberg96 at gmail.com (Daniel Berger)
Date: Thu, 2 Dec 2010 08:19:28 -0700
Subject: [Win32utils-devel] Fwd: [win32utils-help][95414] RE:
	EventLog.open("Application") fails
In-Reply-To: <20101202150640.D82C6185835E@rubyforge.org>
References: <20101202150640.D82C6185835E@rubyforge.org>
Message-ID: <AANLkTiksmkW_6JwFaYS8YDWC9CAFSAsRZB36Dms7Sw=Y@mail.gmail.com>

I haven't had the chance to dig into this yet, but I suspect a
language/code page issue.

Regards,

Dan


---------- Forwarded message ----------
From: Little Snake <noreply at rubyforge.org>
Date: Thu, Dec 2, 2010 at 8:06 AM
Subject: [win32utils-help][95414] RE: EventLog.open("Application") fails
To: noreply at rubyforge.org



Read and respond to this message at:
http://rubyforge.org/forum/message.php?msg_id=95414
By: Little Snake

Hi,

The OS is Windows XP (polish language version) with service pack 3

the ruby version is:
ruby 1.8.7 (2008-08-11 patchlevel 72) [i386-cygwin]
(the latest from the repository)
as shown above it is version for cygwin (ver. 1.7.7-1).

As I looked at the problem, I've traced it to windows-pr eventlog.rb line 33
where the OpenEventLog function is created, and then to API::initialize
in windows-api api.rb line 316 (which I'm sure, is obvious for you) and there
I was stuck, for the next step was win32-api api.c file which I'd rather not
analyze myself at the moment ? I'm too much of a noob at the area :-(.

But I run the test program for win32-api and surprise: it failed:

$ ruby test_win32_api.rb
Loaded suite test_win32_api
Started
F.....F...........

?1) Failure:
test_call(TC_Win32_API) [test_win32_api.rb:37]:
<"\lib\ruby\gems\1.8\gems\win32-api-1.4.6\test"> expected but was
<"C:\cygwin\lib\ruby\gems\1.8\gems\win32-api-1.4.6\test">.

diff:
- \lib\ruby\gems\1.8\gems\win32-api-1.4.6\test
+ C:\cygwin\lib\ruby\gems\1.8\gems\win32-api-1.4.6\test
? +++++++++

?2) Failure:
test_constructor_expected_failures(TC_Win32_API) [test_win32_api.rb:95]:
<Win32::API::LoadLibraryError> exception expected but was
Class: <Win32::API::Error>
Message: <"Attempt to format message failed (error = '1815')">
---Backtrace---
test_win32_api.rb:95:in `initialize'
test_win32_api.rb:95:in `new'
test_win32_api.rb:95:in `test_constructor_expected_failures'
test_win32_api.rb:95:in `test_constructor_expected_failures'
---------------

Finished in 0.103 seconds.

18 tests, 40 assertions, 2 failures, 0 errors, 0 pendings, 0 omissions,
0 notifications
88.8889% passed

The first failure is obvious and irrelevant, but the second ?

To double check I installed another instance of ruby from rubyinstaller for
windows (the two instances do not interfere).
The version according to ruby -v is:
ruby 1.8.7 (2010-06-23 patchlevel 299) [i386-mingw32]

The result of the same program was slightly different, but again it was
a failure:

C:/prg/Ruby187/lib/ruby/gems/1.8/gems/windows-api-0.4.0/lib/windows/api.rb:336:in
`initialize': Attempt to format message failed (error = '1815')
(Win32::API::Error)
? ? ? ?from C:/prg/Ruby187/lib/ruby/gems/1.8/gems/windows-api-0.4.0/lib/windows
/api.rb:336:in `new'
? ? ? ?from C:/prg/Ruby187/lib/ruby/gems/1.8/gems/windows-api-0.4.0/lib/windows
/api.rb:336:in `initialize'
? ? ? ?from C:/prg/Ruby187/lib/ruby/gems/1.8/gems/windows-pr-1.1.2/lib/windows/
eventlog.rb:39:in `new'
? ? ? ?from C:/prg/Ruby187/lib/ruby/gems/1.8/gems/windows-pr-1.1.2/lib/windows/
eventlog.rb:39
? ? ? ?from C:/prg/Ruby187/lib/ruby/site_ruby/1.8/rubygems/custom_require.rb:31:in
`gem_original_require'
? ? ? ?from C:/prg/Ruby187/lib/ruby/site_ruby/1.8/rubygems/custom_require.rb:31:in
`require'
? ? ? ?from C:/prg/Ruby187/lib/ruby/gems/1.8/gems/win32-eventlog-0.5.2/lib/win3
2/eventlog.rb:2
? ? ? ?from C:/prg/Ruby187/lib/ruby/site_ruby/1.8/rubygems/custom_require.rb:36:in
`gem_original_require'
? ? ? ?from C:/prg/Ruby187/lib/ruby/site_ruby/1.8/rubygems/custom_require.rb:36:in
`require'
? ? ? ?from wlogex.rb:4

Then again ruby test_win32_api.rb
C:/prg/Ruby187/lib/ruby/site_ruby/1.8/rubygems.rb:779:in
`report_activate_error':
Could not find RubyGem test-unit (>= 0) (Gem::LoadError)
? ? ? ?from C:/prg/Ruby187/lib/ruby/site_ruby/1.8/rubygems.rb:214:in
`activate'
? ? ? ?from C:/prg/Ruby187/lib/ruby/site_ruby/1.8/rubygems.rb:1082:in `gem'
? ? ? ?from test_win32_api.rb:8

C:\prg\Ruby187\lib\ruby\gems\1.8\gems\win32-api-1.4.6-x86-mingw32\test>ruby
test_win32_api.rb
Loaded suite test_win32_api
Started
......F...........

?1) Failure:
test_constructor_expected_failures(TC_Win32_API) [test_win32_api.rb:95]:
<Win32::API::LoadLibraryError> exception expected but was
Class: <Win32::API::Error>
Message: <"Attempt to format message failed (error = '1815')">
---Backtrace---
test_win32_api.rb:95:in `initialize'
test_win32_api.rb:95:in `new'
test_win32_api.rb:95:in `test_constructor_expected_failures'
test_win32_api.rb:95:in `test_constructor_expected_failures'
---------------

Finished in 0.078066 seconds.

18 tests, 40 assertions, 1 failures, 0 errors, 0 pendings, 0 omissions,
0 notifications
94.4444% passed
And again the "Attempt to format message failed? error.

I would be very gratefull for any help in resolving the issue. ?What I dreamed
for was a nice script checking eventlogs on a number of windows servers, and
I really long for such a thing ;-).


Thank you for your time.


Little Snake


______________________________________________________________________
You are receiving this email because you elected to monitor this forum.
To stop monitoring this forum, login to RubyForge and visit:
http://rubyforge.org/forum/monitor.php?forum_id=320&group_id=85&stop=1

From billagee at gmail.com  Tue Dec 14 15:13:52 2010
From: billagee at gmail.com (Bill Agee)
Date: Tue, 14 Dec 2010 12:13:52 -0800
Subject: [Win32utils-devel] How to initialize and pass a VARIANT to
	IAccessible::get_accName?
Message-ID: <AANLkTi=C5SzLsx6oBq1pU7Kk=XAo4EYwzw3eukA2czTq@mail.gmail.com>

Hi all,

First off, thanks for Win32Utils!  The libraries have been really beneficial
for me.

I have a question (hope this is an appropriate place for it):

I'm trying to write some demo scripts that use the IAccessible interface:

http://msdn.microsoft.com/en-us/library/dd318466%28v=vs.85%29.aspx

I've had some success so far (for example, I can successfully call the
"IAccessible::get_accChildCount" method).

But I'm stuck on IAccessible::get_accName.

In fact, any method like get_accName that takes a VARIANT (note: not a
VARIANT*) seems to be failing for me.

Does anyone have experience using win32-api to define and call a function
that takes a VARIANT?

In the case of get_accName, I can't get it to return successfully...

Either ruby.exe crashes (when I pass a 16-byte packed string as the VARIANT)
or I get the HRESULT 'The parameter is incorrect.' (when I use a >=28-byte
packed string as the VARIANT).


I'm wondering if the problem is one of:

- the Win32::API function prototype I'm using for the VARIANT param is
wrong, or:

- the way I'm initializing the VARIANT before passing it is incorrect

I'd be really grateful if anyone could weigh in on how to achieve this. :)

Right now I'm treating the VARIANT as a pointer in the get_accName prototype
(note the full example program this came from is at the end of the email):

  Get_accName = Win32::API::Function.new(table[10], 'PPP', 'L')

The C prototype for the function (from OleAcc.h) is:

  /* [id][propget][hidden] */ HRESULT ( STDMETHODCALLTYPE *get_accName )(
      __RPC__in IAccessible * This,
      /* [optional][in] */ VARIANT varChild,
      /* [retval][out] */ __RPC__deref_out_opt BSTR *pszName);


I combed through the windows-pr code looking for examples of functions that
take a VARIANT param, but so far I've only found functions that take a
pointer-to-VARIANT.

Also, I suspect this code, which initializes my VARIANT, might be causing my
problem (it was cobbled together from snippets on the win32utils-devel list
and the pr-win32ole source):

  # Initialize VARIANT that holds CHILDID_SELF
  self_var = 0.chr * 16
  VariantInit(self_var)
  self_var[0, 2] = [VT_I4].pack('S')
  self_var[2, 2] = [0].pack('S')
  self_var[4, 2] = [0].pack('S')
  self_var[6, 2] = [0].pack('S')
  self_var[8, 4] = [CHILDID_SELF].pack('L')
  self_var[12, 4] = [0].pack('L')


Here's a complete script that shows the get_accName problem - please let me
know if there are any questions I could answer that might help resolve the
issue.  If it would help I also have a working C demo program where
get_accName is used.

# print_desktop_name.rb

require 'rubygems'
require 'win32/api'
require 'windows/com'
require 'windows/com/variant'
require 'windows/error'
require 'windows/msvcrt/buffer'
require 'windows/unicode'
require 'windows/window'
require 'windows/window/menu'

include Windows::COM
include Windows::COM::Variant
include Windows::Error
include Windows::MSVCRT::Buffer
include Windows::Unicode
include Windows::Window
include Windows::Window::Menu

IID_IAccessible = [0x618736e0, 0x3c3d, 0x11cf, 0x81, 0x0c, 0x00, 0xaa, 0x00,
0x38, 0x9b, 0x71].pack('LSSC8')
CHILDID_SELF = 0

CoInitialize(nil) # Initialize COM

desktop_hwnd = GetDesktopWindow.call() # Get the desktop's HWND

# Use the HWND to get the desktop's IAccessible object
AccessibleObjectFromWindow = Win32::API.new('AccessibleObjectFromWindow',
                                            'LLPP', 'L', 'oleacc')
desktop_iacc = 0.chr * 4
hr = AccessibleObjectFromWindow.call(desktop_hwnd,
                                     OBJID_WINDOW,
                                     IID_IAccessible,
                                     desktop_iacc)
raise "Failed to instantiate IAcc object!" if (hr != S_OK)

# Get the memory address of the IAcc pointer - need to use it later as a
# 'this' pointer to pass to get_accName
desktop_iacc_ptr = desktop_iacc.unpack('L').first

# Get the IAccessibleVtbl C interface. It contains 28 functions (see
OleAcc.h).
lpVtbl = 0.chr * 4
table = 0.chr * (4 * 28)

# Unpack the IAccessibleVtbl:
memcpy(lpVtbl, desktop_iacc_ptr, 4)
memcpy(table, lpVtbl.unpack('L').first, 4 * 28)
table = table.unpack('L*')

# Define get_accName
Get_accName = Win32::API::Function.new(table[10], 'PPP', 'L')

# Create buffer for the BSTR that will receive the IAcc object name
name_bstr = 0.chr * 4

# Initialize CHILDID_SELF variant
self_var = 0.chr * 16
VariantInit(self_var)
self_var[0, 2] = [VT_I4].pack('S')
self_var[2, 2] = [0].pack('S')
self_var[4, 2] = [0].pack('S')
self_var[6, 2] = [0].pack('S')
self_var[8, 4] = [CHILDID_SELF].pack('L')
self_var[12, 4] = [0].pack('L')

# This segfaults on XP and Win7:
# (using ruby 1.8.7 i386-mingw32, 2010-08-16 patchlevel 302)
hr = Get_accName.call(desktop_iacc_ptr, self_var, name_bstr)
puts "HRESULT from get_accName is: " + hr.to_s
puts "HRESULT message is: '#{get_last_error(hr)}'"
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://rubyforge.org/pipermail/win32utils-devel/attachments/20101214/3005e52c/attachment-0001.html>

From phasis at gmail.com  Tue Dec 14 21:36:06 2010
From: phasis at gmail.com (Heesob Park)
Date: Wed, 15 Dec 2010 11:36:06 +0900
Subject: [Win32utils-devel] How to initialize and pass a VARIANT to
	IAccessible::get_accName?
In-Reply-To: <AANLkTi=C5SzLsx6oBq1pU7Kk=XAo4EYwzw3eukA2czTq@mail.gmail.com>
References: <AANLkTi=C5SzLsx6oBq1pU7Kk=XAo4EYwzw3eukA2czTq@mail.gmail.com>
Message-ID: <AANLkTi=QjMT_w7m622gTbGqSifEtRRELxnXfag48TjKE@mail.gmail.com>

Hi,

2010/12/15 Bill Agee <billagee at gmail.com>:
> Hi all,
>
> First off, thanks for Win32Utils!? The libraries have been really beneficial
> for me.
>
> I have a question (hope this is an appropriate place for it):
>
> I'm trying to write some demo scripts that use the IAccessible interface:
>
> http://msdn.microsoft.com/en-us/library/dd318466%28v=vs.85%29.aspx
>
> I've had some success so far (for example, I can successfully call the
> "IAccessible::get_accChildCount" method).
>
> But I'm stuck on IAccessible::get_accName.
>
> In fact, any method like get_accName that takes a VARIANT (note: not a
> VARIANT*) seems to be failing for me.
>
> Does anyone have experience using win32-api to define and call a function
> that takes a VARIANT?
>
> In the case of get_accName, I can't get it to return successfully...
>
> Either ruby.exe crashes (when I pass a 16-byte packed string as the VARIANT)
> or I get the HRESULT 'The parameter is incorrect.' (when I use a >=28-byte
> packed string as the VARIANT).
>
>
> I'm wondering if the problem is one of:
>
> - the Win32::API function prototype I'm using for the VARIANT param is
> wrong, or:
>
> - the way I'm initializing the VARIANT before passing it is incorrect
>
> I'd be really grateful if anyone could weigh in on how to achieve this. :)
>
> Right now I'm treating the VARIANT as a pointer in the get_accName prototype
> (note the full example program this came from is at the end of the email):
>
> ? Get_accName = Win32::API::Function.new(table[10], 'PPP', 'L')
>
> The C prototype for the function (from OleAcc.h) is:
>
> ? /* [id][propget][hidden] */ HRESULT ( STDMETHODCALLTYPE *get_accName )(
> ????? __RPC__in IAccessible * This,
> ????? /* [optional][in] */ VARIANT varChild,
> ????? /* [retval][out] */ __RPC__deref_out_opt BSTR *pszName);
>
>
> I combed through the windows-pr code looking for examples of functions that
> take a VARIANT param, but so far I've only found functions that take a
> pointer-to-VARIANT.
>
> Also, I suspect this code, which initializes my VARIANT, might be causing my
> problem (it was cobbled together from snippets on the win32utils-devel list
> and the pr-win32ole source):
>
> ? # Initialize VARIANT that holds CHILDID_SELF
> ? self_var = 0.chr * 16
> ? VariantInit(self_var)
> ? self_var[0, 2] = [VT_I4].pack('S')
> ? self_var[2, 2] = [0].pack('S')
> ? self_var[4, 2] = [0].pack('S')
> ? self_var[6, 2] = [0].pack('S')
> ? self_var[8, 4] = [CHILDID_SELF].pack('L')
> ? self_var[12, 4] = [0].pack('L')
>
>
> Here's a complete script that shows the get_accName problem - please let me
> know if there are any questions I could answer that might help resolve the
> issue.? If it would help I also have a working C demo program where
> get_accName is used.
>
> # print_desktop_name.rb
>
> require 'rubygems'
> require 'win32/api'
> require 'windows/com'
> require 'windows/com/variant'
> require 'windows/error'
> require 'windows/msvcrt/buffer'
> require 'windows/unicode'
> require 'windows/window'
> require 'windows/window/menu'
>
> include Windows::COM
> include Windows::COM::Variant
> include Windows::Error
> include Windows::MSVCRT::Buffer
> include Windows::Unicode
> include Windows::Window
> include Windows::Window::Menu
>
> IID_IAccessible = [0x618736e0, 0x3c3d, 0x11cf, 0x81, 0x0c, 0x00, 0xaa, 0x00,
> 0x38, 0x9b, 0x71].pack('LSSC8')
> CHILDID_SELF = 0
>
> CoInitialize(nil) # Initialize COM
>
> desktop_hwnd = GetDesktopWindow.call() # Get the desktop's HWND
>
> # Use the HWND to get the desktop's IAccessible object
> AccessibleObjectFromWindow = Win32::API.new('AccessibleObjectFromWindow',
> ??????????????????????????????????????????? 'LLPP', 'L', 'oleacc')
> desktop_iacc = 0.chr * 4
> hr = AccessibleObjectFromWindow.call(desktop_hwnd,
> ???????????????????????????????????? OBJID_WINDOW,
> ???????????????????????????????????? IID_IAccessible,
> ???????????????????????????????????? desktop_iacc)
> raise "Failed to instantiate IAcc object!" if (hr != S_OK)
>
> # Get the memory address of the IAcc pointer - need to use it later as a
> # 'this' pointer to pass to get_accName
> desktop_iacc_ptr = desktop_iacc.unpack('L').first
>
> # Get the IAccessibleVtbl C interface. It contains 28 functions (see
> OleAcc.h).
> lpVtbl = 0.chr * 4
> table = 0.chr * (4 * 28)
>
> # Unpack the IAccessibleVtbl:
> memcpy(lpVtbl, desktop_iacc_ptr, 4)
> memcpy(table, lpVtbl.unpack('L').first, 4 * 28)
> table = table.unpack('L*')
>
> # Define get_accName
> Get_accName = Win32::API::Function.new(table[10], 'PPP', 'L')
>
> # Create buffer for the BSTR that will receive the IAcc object name
> name_bstr = 0.chr * 4
>
> # Initialize CHILDID_SELF variant
> self_var = 0.chr * 16
> VariantInit(self_var)
> self_var[0, 2] = [VT_I4].pack('S')
> self_var[2, 2] = [0].pack('S')
> self_var[4, 2] = [0].pack('S')
> self_var[6, 2] = [0].pack('S')
> self_var[8, 4] = [CHILDID_SELF].pack('L')
> self_var[12, 4] = [0].pack('L')
>
> # This segfaults on XP and Win7:
> # (using ruby 1.8.7 i386-mingw32, 2010-08-16 patchlevel 302)
> hr = Get_accName.call(desktop_iacc_ptr, self_var, name_bstr)
> puts "HRESULT from get_accName is: " + hr.to_s
> puts "HRESULT message is: '#{get_last_error(hr)}'"
>
After some debugging, I found that passing VARIANT parameter is same
to passing four DWORD parameters  with each VARIANT content element.

Get_accName = Win32::API::Function.new(table[10], 'PPP', 'L')
should be
Get_accName = Win32::API::Function.new(table[10], 'PLLLLP', 'L')

hr = Get_accName.call(desktop_iacc_ptr, self_var, name_bstr)
should be
hr = Get_accName.call(desktop_iacc_ptr,self_var[0,4].unpack('L').first,self_var[4,4].unpack('L').first,self_var[8,4].unpack('L').first,self_var[12,4].unpack('L').first,name_bstr)

Regards,
Park Heesob

From billagee at gmail.com  Wed Dec 15 00:38:16 2010
From: billagee at gmail.com (Bill Agee)
Date: Tue, 14 Dec 2010 21:38:16 -0800
Subject: [Win32utils-devel] How to initialize and pass a VARIANT to
	IAccessible::get_accName?
In-Reply-To: <AANLkTi=QjMT_w7m622gTbGqSifEtRRELxnXfag48TjKE@mail.gmail.com>
References: <AANLkTi=C5SzLsx6oBq1pU7Kk=XAo4EYwzw3eukA2czTq@mail.gmail.com>
	<AANLkTi=QjMT_w7m622gTbGqSifEtRRELxnXfag48TjKE@mail.gmail.com>
Message-ID: <AANLkTikE+Fu2GqE_sSKL41_TtyMjqAw3ExyoVDYtihd=@mail.gmail.com>

On Tue, Dec 14, 2010 at 6:36 PM, Heesob Park <phasis at gmail.com> wrote:


> After some debugging, I found that passing VARIANT parameter is same
> to passing four DWORD parameters  with each VARIANT content element.
>
> Get_accName = Win32::API::Function.new(table[10], 'PPP', 'L')
> should be
> Get_accName = Win32::API::Function.new(table[10], 'PLLLLP', 'L')
>
> hr = Get_accName.call(desktop_iacc_ptr, self_var, name_bstr)
> should be
> hr =
> Get_accName.call(desktop_iacc_ptr,self_var[0,4].unpack('L').first,self_var[4,4].unpack('L').first,self_var[8,4].unpack('L').first,self_var[12,4].unpack('L').first,name_bstr)
>


Outstanding!  The code works great now.  With your changes in place,
get_accName return S_OK, and the BSTR contains the expected string
('Desktop') when I print it with wprintf:

  Wprintf = Windows::API.new('wprintf', 'PP', 'I', 'msvcrt')

  format_str = multi_to_wide("The object name is: '%s'\n")
  format_str_ptr = [format_str].pack('p').unpack('L').first
  Wprintf.call(format_str_ptr, name_bstr.unpack('L').first)

This is huge; being able to pass a VARIANT means that the IAccessible and
IUIAutomation interfaces are now wide open. :)

Thanks so much!

Bill
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://rubyforge.org/pipermail/win32utils-devel/attachments/20101214/713374a8/attachment.html>

From noreply at rubyforge.org  Sat Dec 25 05:02:54 2010
From: noreply at rubyforge.org (noreply at rubyforge.org)
Date: Sat, 25 Dec 2010 05:02:54 -0500 (EST)
Subject: [Win32utils-devel] [ win32utils-Bugs-28801 ] unclosed handle in
	Process.setpriority/getpriority
Message-ID: <20101225100254.AE7D31858317@rubyforge.org>

Bugs item #28801, was opened at 2010-12-25 11:02
You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=411&aid=28801&group_id=85

Category: win32-process
Group: Code
Status: Open
Resolution: None
Priority: 3
Submitted By: Rafa? Michalski (royaltm)
Assigned to: Nobody (None)
Summary: unclosed handle in Process.setpriority/getpriority 

Initial Comment:
Using Process.setpriority/getpriority was crashing my ruby program.

I think the problem was with unclosed handle from OpenProcess() function.

my patch for lib/win32/process.rb (0.6.4) follows, however the bug exists also in previous releases.


----------------------------------------------------------------------

You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=411&aid=28801&group_id=85

From noreply at rubyforge.org  Sun Dec 26 12:59:08 2010
From: noreply at rubyforge.org (noreply at rubyforge.org)
Date: Sun, 26 Dec 2010 12:59:08 -0500 (EST)
Subject: [Win32utils-devel] [ win32utils-Bugs-28801 ] unclosed handle in
	Process.setpriority/getpriority
Message-ID: <20101226175908.E00981858300@rubyforge.org>

Bugs item #28801, was opened at 2010-12-25 03:02
You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=411&aid=28801&group_id=85

Category: win32-process
Group: Code
Status: Open
Resolution: None
Priority: 3
Submitted By: Rafa? Michalski (royaltm)
Assigned to: Nobody (None)
Summary: unclosed handle in Process.setpriority/getpriority 

Initial Comment:
Using Process.setpriority/getpriority was crashing my ruby program.

I think the problem was with unclosed handle from OpenProcess() function.

my patch for lib/win32/process.rb (0.6.4) follows, however the bug exists also in previous releases.


----------------------------------------------------------------------

>Comment By: Daniel Berger (djberg96)
Date: 2010-12-26 10:59

Message:
Yeah, I need to clean up getpriority and setpriority in general to more strictly match the current spec.

I think I originally avoided CloseHandle because I seem to recall it causing issues because it would affect the current process somehow. Maybe I'm imagining things, though. I'll definitely take a look.

Thanks,

Dan

----------------------------------------------------------------------

You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=411&aid=28801&group_id=85

From noreply at rubyforge.org  Mon Dec 27 03:02:38 2010
From: noreply at rubyforge.org (noreply at rubyforge.org)
Date: Mon, 27 Dec 2010 03:02:38 -0500 (EST)
Subject: [Win32utils-devel] [ win32utils-Bugs-28801 ] unclosed handle in
	Process.setpriority/getpriority
Message-ID: <20101227080238.2B4B11858356@rubyforge.org>

Bugs item #28801, was opened at 2010-12-25 11:02
You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=411&aid=28801&group_id=85

Category: win32-process
Group: Code
Status: Open
Resolution: None
Priority: 3
Submitted By: Rafa? Michalski (royaltm)
Assigned to: Nobody (None)
Summary: unclosed handle in Process.setpriority/getpriority 

Initial Comment:
Using Process.setpriority/getpriority was crashing my ruby program.

I think the problem was with unclosed handle from OpenProcess() function.

my patch for lib/win32/process.rb (0.6.4) follows, however the bug exists also in previous releases.


----------------------------------------------------------------------

>Comment By: Rafa? Michalski (royaltm)
Date: 2010-12-27 09:02

Message:
I'm applying this patch since version 0.6.2 in production
environment (though didn't report it until now) and it
servers me right.
I'm using Process.setpriority from ruby program to give more
cpu attention to one of (not ruby) child processes (spawned
with Process.create) which handles serial communication with
card reader device.
Didn't notice any trouble on WinXP and Vista64.


----------------------------------------------------------------------

Comment By: Daniel Berger (djberg96)
Date: 2010-12-26 18:59

Message:
Yeah, I need to clean up getpriority and setpriority in general to more strictly match the current spec.

I think I originally avoided CloseHandle because I seem to recall it causing issues because it would affect the current process somehow. Maybe I'm imagining things, though. I'll definitely take a look.

Thanks,

Dan

----------------------------------------------------------------------

You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=411&aid=28801&group_id=85

From noreply at rubyforge.org  Mon Dec 27 11:32:03 2010
From: noreply at rubyforge.org (noreply at rubyforge.org)
Date: Mon, 27 Dec 2010 11:32:03 -0500 (EST)
Subject: [Win32utils-devel] [ win32utils-Bugs-28801 ] unclosed handle in
	Process.setpriority/getpriority
Message-ID: <20101227163203.EF4CD1858354@rubyforge.org>

Bugs item #28801, was opened at 2010-12-25 03:02
You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=411&aid=28801&group_id=85

Category: win32-process
Group: Code
>Status: Closed
>Resolution: Accepted
Priority: 3
Submitted By: Rafa? Michalski (royaltm)
>Assigned to: Daniel Berger (djberg96)
Summary: unclosed handle in Process.setpriority/getpriority 

Initial Comment:
Using Process.setpriority/getpriority was crashing my ruby program.

I think the problem was with unclosed handle from OpenProcess() function.

my patch for lib/win32/process.rb (0.6.4) follows, however the bug exists also in previous releases.


----------------------------------------------------------------------

>Comment By: Daniel Berger (djberg96)
Date: 2010-12-27 09:32

Message:
Ok, I've fixed it in 0.6.5, which I just pushed.

Please note that I've made the API for getpriority and setpriority stricter, i.e. there are no longer any default arguments. This was so that it matched the MRI spec.

Thanks for the report.

Regards,

Dan

----------------------------------------------------------------------

Comment By: Rafa? Michalski (royaltm)
Date: 2010-12-27 01:02

Message:
I'm applying this patch since version 0.6.2 in production
environment (though didn't report it until now) and it
servers me right.
I'm using Process.setpriority from ruby program to give more
cpu attention to one of (not ruby) child processes (spawned
with Process.create) which handles serial communication with
card reader device.
Didn't notice any trouble on WinXP and Vista64.


----------------------------------------------------------------------

Comment By: Daniel Berger (djberg96)
Date: 2010-12-26 10:59

Message:
Yeah, I need to clean up getpriority and setpriority in general to more strictly match the current spec.

I think I originally avoided CloseHandle because I seem to recall it causing issues because it would affect the current process somehow. Maybe I'm imagining things, though. I'll definitely take a look.

Thanks,

Dan

----------------------------------------------------------------------

You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=411&aid=28801&group_id=85

From noreply at rubyforge.org  Mon Dec 27 11:55:28 2010
From: noreply at rubyforge.org (noreply at rubyforge.org)
Date: Mon, 27 Dec 2010 11:55:28 -0500 (EST)
Subject: [Win32utils-devel] [ win32utils-Bugs-28801 ] unclosed handle in
	Process.setpriority/getpriority
Message-ID: <20101227165528.730911858354@rubyforge.org>

Bugs item #28801, was opened at 2010-12-25 11:02
You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=411&aid=28801&group_id=85

Category: win32-process
Group: Code
Status: Closed
Resolution: Accepted
Priority: 3
Submitted By: Rafa? Michalski (royaltm)
Assigned to: Daniel Berger (djberg96)
Summary: unclosed handle in Process.setpriority/getpriority 

Initial Comment:
Using Process.setpriority/getpriority was crashing my ruby program.

I think the problem was with unclosed handle from OpenProcess() function.

my patch for lib/win32/process.rb (0.6.4) follows, however the bug exists also in previous releases.


----------------------------------------------------------------------

>Comment By: Rafa? Michalski (royaltm)
Date: 2010-12-27 17:55

Message:
Oh, i forgot to thank you for the great library.


----------------------------------------------------------------------

Comment By: Daniel Berger (djberg96)
Date: 2010-12-27 17:32

Message:
Ok, I've fixed it in 0.6.5, which I just pushed.

Please note that I've made the API for getpriority and setpriority stricter, i.e. there are no longer any default arguments. This was so that it matched the MRI spec.

Thanks for the report.

Regards,

Dan

----------------------------------------------------------------------

Comment By: Rafa? Michalski (royaltm)
Date: 2010-12-27 09:02

Message:
I'm applying this patch since version 0.6.2 in production
environment (though didn't report it until now) and it
servers me right.
I'm using Process.setpriority from ruby program to give more
cpu attention to one of (not ruby) child processes (spawned
with Process.create) which handles serial communication with
card reader device.
Didn't notice any trouble on WinXP and Vista64.


----------------------------------------------------------------------

Comment By: Daniel Berger (djberg96)
Date: 2010-12-26 18:59

Message:
Yeah, I need to clean up getpriority and setpriority in general to more strictly match the current spec.

I think I originally avoided CloseHandle because I seem to recall it causing issues because it would affect the current process somehow. Maybe I'm imagining things, though. I'll definitely take a look.

Thanks,

Dan

----------------------------------------------------------------------

You can respond by visiting: 
http://rubyforge.org/tracker/?func=detail&atid=411&aid=28801&group_id=85