[ruby-core:36406] 1.8.7 release next month

[ruby-core:36406] 1.8.7 release next month

[Urabe Shyouhei]

Hello core people,

As several patches are introduced to 1.8.7 I'd like to have a release.  I hope
that'd happen next month.  If you have any show-stoppers please let me know.
Thank you.

[ruby-core:36414] Re: 1.8.7 release next month

[Luis Lavena]

2011/5/23 Urabe Shyouhei <shyouhei@ruby-lang.org>:
> Hello core people,
>
> As several patches are introduced to 1.8.7 I'd like to have a release.  I hope
> that'd happen next month.  If you have any show-stoppers please let me know.

Hello,

At RubyInstaller project we are still attempting to get Tcl/Tk work
properly under MinGW, which hasn't been an easy task.

We have a ticket open for 1.9.x:

http://redmine.ruby-lang.org/issues/4141

We would like to be able to ship newer 1.8.7 with tcl/tk included, but
for that we will require first fix trunk and then backport
appropriately.

At this time I'm compiling and will run tests. Will let you know any
show-stopper beyond the above mentioned.

Thank you.

-- 
Luis Lavena
AREA 17
-
Perfection in design is achieved not when there is nothing more to add,
but rather when there is nothing more to take away.
Antoine de Saint-Exupéry

[ruby-core:36487] Re: 1.8.7 release next month

[Urabe Shyouhei]

Hi Luis,

(05/24/2011 07:27 AM), Luis Lavena wrote:
> 2011/5/23 Urabe Shyouhei <shyouhei@ruby-lang.org>:
>> Hello core people,
>>
>> As several patches are introduced to 1.8.7 I'd like to have a release.  I hope
>> that'd happen next month.  If you have any show-stoppers please let me know.
> 
> Hello,
> 
> At RubyInstaller project we are still attempting to get Tcl/Tk work
> properly under MinGW, which hasn't been an easy task.
> 
> We have a ticket open for 1.9.x:
> 
> http://redmine.ruby-lang.org/issues/4141
> 
> We would like to be able to ship newer 1.8.7 with tcl/tk included, but
> for that we will require first fix trunk and then backport
> appropriately.
> 
> At this time I'm compiling and will run tests. Will let you know any
> show-stopper beyond the above mentioned.
> 
> Thank you.

I'm now designing a release schedule.  How long do you think will it
take for issue #4141 to be solved?  Like I said I want a release in June
so if it takes more than a month, I'm afraid I have to leave it.

[ruby-core:36488] Re: 1.8.7 release next month

[Hidetoshi NAGAI]

From: Urabe Shyouhei <shyouhei@ruby-lang.org>
Subject: [ruby-core:36487] Re: 1.8.7 release next month
Date: Thu, 26 May 2011 17:30:10 +0900
Message-ID: <4DDE0F90.7070200@ruby-lang.org>
> I'm now designing a release schedule.  How long do you think will it
> take for issue #4141 to be solved?  Like I said I want a release in June
> so if it takes more than a month, I'm afraid I have to leave it.

Now, I'm working for the issue.
Please give me one or two days to release or tell how long it takes.
-- 
Hidetoshi NAGAI (nagai@ai.kyutech.ac.jp)
Department of Artificial Intelligence, Kyushu Institute of Technology

[ruby-core:36496] Re: 1.8.7 release next month

[Hidetoshi NAGAI]

From: Hidetoshi NAGAI <nagai@ai.kyutech.ac.jp>
Subject: [ruby-core:36488] Re: 1.8.7 release next month
Date: Thu, 26 May 2011 18:16:20 +0900
Message-ID: <20110526.181619.71088183.nagai@ai.kyutech.ac.jp>
> > I'm now designing a release schedule.  How long do you think will it
> > take for issue #4141 to be solved?  Like I said I want a release in June
> > so if it takes more than a month, I'm afraid I have to leave it.
> Now, I'm working for the issue.
> Please give me one or two days to release or tell how long it takes.

I've committed a patch to trunk.
Could you try it (just copy latest extconf.rb from trunk) ?
-- 
Hidetoshi NAGAI (nagai@ai.kyutech.ac.jp)
Department of Artificial Intelligence, Kyushu Institute of Technology

[ruby-core:36712] Re: 1.8.7 release next month

[Urabe Shyouhei]

Ping Luis, how's it going?

(05/27/2011 08:42 AM), Hidetoshi NAGAI wrote:
> From: Hidetoshi NAGAI <nagai@ai.kyutech.ac.jp>
> Subject: [ruby-core:36488] Re: 1.8.7 release next month
> Date: Thu, 26 May 2011 18:16:20 +0900
> Message-ID: <20110526.181619.71088183.nagai@ai.kyutech.ac.jp>
>>> I'm now designing a release schedule.  How long do you think will it
>>> take for issue #4141 to be solved?  Like I said I want a release in June
>>> so if it takes more than a month, I'm afraid I have to leave it.
>> Now, I'm working for the issue.
>> Please give me one or two days to release or tell how long it takes.
> 
> I've committed a patch to trunk.
> Could you try it (just copy latest extconf.rb from trunk) ?

[ruby-core:36717] Re: 1.8.7 release next month

[Luis Lavena]

Hello, sorry for the top posting. I have some results and news on this. Will email in a few hours (using a limited device right now). Thank you.

Sent from mobile.

El Jun 3, 2011, a las 5:18, Urabe Shyouhei <shyouhei@ruby-lang.org> escribió:

> Ping Luis, how's it going?
> 
> (05/27/2011 08:42 AM), Hidetoshi NAGAI wrote:
>> From: Hidetoshi NAGAI <nagai@ai.kyutech.ac.jp>
>> Subject: [ruby-core:36488] Re: 1.8.7 release next month
>> Date: Thu, 26 May 2011 18:16:20 +0900
>> Message-ID: <20110526.181619.71088183.nagai@ai.kyutech.ac.jp>
>>>> I'm now designing a release schedule.  How long do you think will it
>>>> take for issue #4141 to be solved?  Like I said I want a release in June
>>>> so if it takes more than a month, I'm afraid I have to leave it.
>>> Now, I'm working for the issue.
>>> Please give me one or two days to release or tell how long it takes.
>> 
>> I've committed a patch to trunk.
>> Could you try it (just copy latest extconf.rb from trunk) ?

[ruby-core:36748] Re: 1.8.7 release next month

[Luis Lavena]

On Fri, Jun 3, 2011 at 5:18 AM, Urabe Shyouhei <shyouhei@ruby-lang.org> wrote:
> Ping Luis, how's it going?
>

Hello,

r31904 compiles and runs, tests results:
https://gist.github.com/1008713

tcl/tk *did* compile and is working, without the need for manually
backport Hidetoshi NAGAI tk/extconf.rb modifications.

Backport of above mentioned patch (copy of trunk extconf.rb) does not work:
compiling tk
../../../../../../ruby/ext/tk/extconf.rb:41:in `foreach': No such file
or directory - ../../../../../../ruby/ext/tk/config_list.in
(Errno::ENOENT)
        from ../../../../../../ruby/ext/tk/extconf.rb:41
        from ../../../../ruby/ext/extmk.rb:167:in `load'
        from ../../../../ruby/ext/extmk.rb:167:in `extmake'
        from ../../../../ruby/ext/extmk.rb:444
        from ../../../../ruby/ext/extmk.rb:443:in `each'
        from ../../../../ruby/ext/extmk.rb:443
make: *** [all] Error 1

===

I believe we can leave it as is (without the full backport of it)

Comparing this revision against the one for patchlevel 334, there are
some differences in the tests errors:

334:
https://gist.github.com/834607

Versus ruby_1_8_7 branch:
https://gist.github.com/1008713

  9) Error:
test_truncate(TestFileExhaustive):
Errno::EACCES: Permission denied -
C:/Users/Luis/AppData/Local/Temp/rubytest-file20110605-788-ebwvs0/filetest_file_exhaustive.rb.788.test
    C:/Users/Luis/Projects/oss/ruby/lib/fileutils.rb:1304:in `unlink'
    C:/Users/Luis/Projects/oss/ruby/lib/fileutils.rb:1304:in `remove_file'
    C:/Users/Luis/Projects/oss/ruby/lib/fileutils.rb:1312:in `platform_support'
    C:/Users/Luis/Projects/oss/ruby/lib/fileutils.rb:1303:in `remove_file'
    C:/Users/Luis/Projects/oss/ruby/lib/fileutils.rb:1292:in `remove'
    C:/Users/Luis/Projects/oss/ruby/lib/fileutils.rb:762:in `remove_entry'
    C:/Users/Luis/Projects/oss/ruby/lib/fileutils.rb:1342:in
`postorder_traverse'
    C:/Users/Luis/Projects/oss/ruby/lib/fileutils.rb:1346:in
`postorder_traverse'
    C:/Users/Luis/Projects/oss/ruby/lib/fileutils.rb:1341:in
`postorder_traverse'
    C:/Users/Luis/Projects/oss/ruby/lib/fileutils.rb:1340:in `each'
    C:/Users/Luis/Projects/oss/ruby/lib/fileutils.rb:1340:in
`postorder_traverse'
    C:/Users/Luis/Projects/oss/ruby/lib/fileutils.rb:760:in `remove_entry'
    C:/Users/Luis/Projects/oss/ruby/lib/fileutils.rb:679:in
`remove_entry_secure'
    ../../../../ruby/test/ruby/test_file_exhaustive.rb:31:in `teardown'

 10) Failure:
test_utime(TestFileExhaustive)
[../../../../ruby/test/ruby/test_file_exhaustive.rb:306]:
<Sat Jan 01 00:00:01 -0500 2000> expected but was
<Fri Dec 31 23:00:01 -0500 1999>.

 12) Failure:
test_step_ruby_core_35753(TestRange)
[../../../../ruby/test/ruby/test_range.rb:23]:
<3> expected but was
<4>.

 13) Failure:
test_cgi(TestWEBrickCGI)
    [../../../../ruby/test/webrick/test_cgi.rb:27:in `test_cgi'
     C:/Users/Luis/Projects/oss/ruby/lib/net/http.rb:1054:in `request'
     C:/Users/Luis/Projects/oss/ruby/lib/net/http.rb:2144:in `reading_body'
     C:/Users/Luis/Projects/oss/ruby/lib/net/http.rb:1053:in `request'
     C:/Users/Luis/Projects/oss/ruby/lib/net/http.rb:1037:in `request'
     C:/Users/Luis/Projects/oss/ruby/lib/net/http.rb:543:in `start'
     C:/Users/Luis/Projects/oss/ruby/lib/net/http.rb:1035:in `request'
     ../../../../ruby/test/webrick/test_cgi.rb:26:in `test_cgi'
     ../../../../ruby/test/webrick/utils.rb:44:in `call'
     ../../../../ruby/test/webrick/utils.rb:44:in `start_server'
     ../../../../ruby/test/webrick/utils.rb:52:in `start_httpserver'
     ../../../../ruby/test/webrick/test_cgi.rb:17:in `test_cgi']:
<"/\244\333\244\262/\244\333\244\262"> expected but was
<"/\017U\017\375/\017U\017\375">.

From above failures, the only one that worries me is #12, TestRange failure.

Thoughts?
-- 
Luis Lavena
AREA 17
-
Perfection in design is achieved not when there is nothing more to add,
but rather when there is nothing more to take away.
Antoine de Saint-Exupéry

[ruby-core:36751] Re: 1.8.7 release next month

[Hidetoshi NAGAI]

Hi, 

From: Luis Lavena <luislavena@gmail.com>
Subject: [ruby-core:36748] Re: 1.8.7 release next month
Date: Sun, 5 Jun 2011 15:59:38 +0900
Message-ID: <BANLkTi=NoziHx6kBpJ5ezkiHBBrHsdaanw@mail.gmail.com>
> tcl/tk *did* compile and is working, without the need for manually
> backport Hidetoshi NAGAI tk/extconf.rb modifications.
> 
> Backport of above mentioned patch (copy of trunk extconf.rb) does not work:
> compiling tk
> ../../../../../../ruby/ext/tk/extconf.rb:41:in `foreach': No such file
> or directory - ../../../../../../ruby/ext/tk/config_list.in

Ah, sorry. Please copy config_list.in also.

# I updated it for new options at 2011-06-05 17:23:43 +0900 [ruby-cvs:39103].
-- 
Hidetoshi NAGAI (nagai@ai.kyutech.ac.jp)
Department of Artificial Intelligence, Kyushu Institute of Technology

[ruby-core:36777] Re: 1.8.7 release next month

[Luis Lavena]

On Sun, Jun 5, 2011 at 4:30 AM, Hidetoshi NAGAI <nagai@ai.kyutech.ac.jp> wrote:
> Hi,
>
> Ah, sorry. Please copy config_list.in also.
>
> # I updated it for new options at 2011-06-05 17:23:43 +0900 [ruby-cvs:39103].

Hello, I've copied "as-is" both extconf.rb and config_list.in to
ruby_1_8_7 branch and tcl/tk compiled and works perfectly.

Would you mind dealing with the backport process? I believe Yugui is
still trying to apply the massive change to ruby_1_9_2 branch so I
think Urabe Shyouhei will have the same problem.

Thank you.

-- 
Luis Lavena
AREA 17
-
Perfection in design is achieved not when there is nothing more to add,
but rather when there is nothing more to take away.
Antoine de Saint-Exupéry

[ruby-core:36778] Re: 1.8.7 release next month

[Urabe Shyouhei]

(06/06/2011 01:16 PM), Luis Lavena wrote:
> On Sun, Jun 5, 2011 at 4:30 AM, Hidetoshi NAGAI <nagai@ai.kyutech.ac.jp> wrote:
>> Hi,
>>
>> Ah, sorry. Please copy config_list.in also.
>>
>> # I updated it for new options at 2011-06-05 17:23:43 +0900 [ruby-cvs:39103].
> 
> Hello, I've copied "as-is" both extconf.rb and config_list.in to
> ruby_1_8_7 branch and tcl/tk compiled and works perfectly.

Good to hear that.

> Would you mind dealing with the backport process? I believe Yugui is
> still trying to apply the massive change to ruby_1_9_2 branch so I
> think Urabe Shyouhei will have the same problem.

I see.  I will release 1.8.7 with this issue fixed.  To do so I might
ask you to check 1.8.7 branch on occasions.  Thanks in advance.

[ruby-core:36795] Re: 1.8.7 release next month

[Luis Lavena]

On Mon, Jun 6, 2011 at 12:24 AM, Urabe Shyouhei <shyouhei@ruby-lang.org> wrote:
>
>> Would you mind dealing with the backport process? I believe Yugui is
>> still trying to apply the massive change to ruby_1_9_2 branch so I
>> think Urabe Shyouhei will have the same problem.
>
> I see.  I will release 1.8.7 with this issue fixed.  To do so I might
> ask you to check 1.8.7 branch on occasions.  Thanks in advance.
>

I have no problem on doing it, appreciate your time on doing the patching :-)

Thank you.
-- 
Luis Lavena
AREA 17
-
Perfection in design is achieved not when there is nothing more to add,
but rather when there is nothing more to take away.
Antoine de Saint-Exupéry

[ruby-core:37212] Re: 1.8.7 release next month

[Luis Lavena]

On Mon, Jun 6, 2011 at 1:24 AM, Urabe Shyouhei <shyouhei@ruby-lang.org> wrote:
>>
>> Hello, I've copied "as-is" both extconf.rb and config_list.in to
>> ruby_1_8_7 branch and tcl/tk compiled and works perfectly.
>
> Good to hear that.
>
>> Would you mind dealing with the backport process? I believe Yugui is
>> still trying to apply the massive change to ruby_1_9_2 branch so I
>> think Urabe Shyouhei will have the same problem.
>
> I see.  I will release 1.8.7 with this issue fixed.  To do so I might
> ask you to check 1.8.7 branch on occasions.  Thanks in advance.
>

I've tried latest ruby_1_8_7 branch (r32129) and while tk, tcltk and
friends compile successfully, we are having problems to make
TkVarAccess to read our pre-set TCL_LIBRARY variable.

In RubyInstaller, we preset that variable to
File.join(RbConfig::CONFIG['libdir'], 'tcltk', 'tcl8.5'), so we avoid
polluting ruby/lib with all tcl&tk files.

This works perfectly under 1.9.2 and trunk, but fails for 1.8.7:

$ ruby -e "require 'tk'"
C:/ruby-1.8.7-p348-i386-mingw32/lib/ruby/1.8/tk.rb:1133:in
`initialize': Can't find a usable init.tcl in the following
directories:  (RuntimeError)
    C:/ruby-1.8.7-p348-i386-mingw32/lib/tcl8.5
C:/ruby-1.8.7-p348-i386-mingw32/lib/tcl8.5 C:/lib/tcl8.5
C:/ruby-1.8.7-p348-i386-mingw32/library C:/library
C:/Users/Luis/Projects/oss/oneclick/rubyi
        from C:/ruby-1.8.7-p348-i386-mingw32/lib/ruby/1.8/tk.rb:1133:in `new'
        from C:/ruby-1.8.7-p348-i386-mingw32/lib/ruby/1.8/tk.rb:1133
        from -e:1:in `require'
        from -e:1

It seems that is ignoring TCL_LIBRARY environment variable?

-- 
Luis Lavena
AREA 17
-
Perfection in design is achieved not when there is nothing more to add,
but rather when there is nothing more to take away.
Antoine de Saint-Exupéry

[ruby-core:37229] Re: 1.8.7 release next month

[Hidetoshi NAGAI]

From: Luis Lavena <luislavena@gmail.com>
Subject: [ruby-core:37212] Re: 1.8.7 release next month
Date: Sat, 18 Jun 2011 09:04:07 +0900
Message-ID: <BANLkTinBM67Fagd8=ph+9zAgWSwqsFsN8w@mail.gmail.com>
> $ ruby -e "require 'tk'"
> C:/ruby-1.8.7-p348-i386-mingw32/lib/ruby/1.8/tk.rb:1133:in
> `initialize': Can't find a usable init.tcl in the following
> directories:  (RuntimeError)
>     C:/ruby-1.8.7-p348-i386-mingw32/lib/tcl8.5
> C:/ruby-1.8.7-p348-i386-mingw32/lib/tcl8.5 C:/lib/tcl8.5
> C:/ruby-1.8.7-p348-i386-mingw32/library C:/library
> C:/Users/Luis/Projects/oss/oneclick/rubyi
>         from C:/ruby-1.8.7-p348-i386-mingw32/lib/ruby/1.8/tk.rb:1133:in `new'
>         from C:/ruby-1.8.7-p348-i386-mingw32/lib/ruby/1.8/tk.rb:1133
>         from -e:1:in `require'
>         from -e:1
> 
> It seems that is ignoring TCL_LIBRARY environment variable?

I can't generate the problem.
"ruby -r tk -e 'p Tk.tk_call("info", "library")'" prints the value of
TCL_LIBRARY environment variable.
Is the env var really exported?
When not do "export TCL_LIBRARY" on the terminal, the value of the env
var has no effect.
-- 
Hidetoshi NAGAI (nagai@ai.kyutech.ac.jp)
Department of Artificial Intelligence, Kyushu Institute of Technology

[ruby-core:37262] Re: 1.8.7 release next month

[Luis Lavena]

On Sun, Jun 19, 2011 at 10:55 PM, Hidetoshi NAGAI
<nagai@ai.kyutech.ac.jp> wrote:
>
> I can't generate the problem.
> "ruby -r tk -e 'p Tk.tk_call("info", "library")'" prints the value of
> TCL_LIBRARY environment variable.
> Is the env var really exported?
> When not do "export TCL_LIBRARY" on the terminal, the value of the env
> var has no effect.

Hello,

I'll try to clarify my previous statement.

Tcl/Tk libs are not placed in the default directory ($ruby/lib) but
instead in a custom directory ($ruby/lib/tcltk)

Because of it, attempting to require 'tk' fails because is unable to
find init.tcl anywhere in the whole path options for lib.

What we did, as this patch shows:

https://github.com/oneclick/rubyinstaller/blob/master/resources/patches/tk/0002-Use-Tcl-Tk-from-different-directory.diff

Is set TCL_LIBRARY before loading the C extension.

This works perfectly against 1.9.3 (trunk), but fails on 1.8.7

TCL_LIBRARY is set inside Ruby and can't be set/exported from outside
ruby code because that will pollute the environment.

Your example works on trunk:

V:\sandbox>ruby -v
ruby 1.9.3dev (2011-06-21 trunk 32188) [i386-mingw32]

V:\sandbox>ruby -r tk -e 'p Tk.tk_call("info", "library")'"
"V:/sandbox/ruby19_mingw/lib/tcltk/tcl8.5"

But fails with ruby_1_8_7 branch:

V:\sandbox>ruby -v
ruby 1.8.7 (2011-06-16 patchlevel 348) [i386-mingw32]

V:\sandbox>ruby -r tk -e 'p Tk.tk_call("info", "library")'"
V:/sandbox/ruby18_mingw/lib/ruby/1.8/tk.rb:1132:in `initialize': Can't
find a usable init.tcl in the following directories:  (RuntimeError)
    V:/sandbox/ruby18_mingw/lib/tcl8.5
V:/sandbox/ruby18_mingw/lib/tcl8.5 V:/sandbox/lib/tcl8.5
V:/sandbox/ruby18_mingw/library V:/sandbox/library
V:/sandbox/tcl8.5.9/library V:/tcl8.5.9/library



This probably means that Tcl wasn't installed properly.
        from V:/sandbox/ruby18_mingw/lib/ruby/1.8/tk.rb:1132:in `new'
        from V:/sandbox/ruby18_mingw/lib/ruby/1.8/tk.rb:1132

===

As you can see, it is ignoring TCL_LIBRARY value set from inside Ruby itself.

Now, setting it from outside the process works:

V:\sandbox>set TCL_LIBRARY
Environment variable TCL_LIBRARY not defined

V:\sandbox>SET TCL_LIBRARY=V:\sandbox\ruby18_mingw\lib\tcltk\tcl8.5

V:\sandbox>ruby -r tk -e 'p Tk.tk_call("info", "library")'"
"V:\\sandbox\\ruby18_mingw\\lib\\tcltk\\tcl8.5"

===

Which could indicate another problem with TCL_LIBRARY?

-- 
Luis Lavena
AREA 17
-
Perfection in design is achieved not when there is nothing more to add,
but rather when there is nothing more to take away.
Antoine de Saint-Exupéry

[ruby-core:37556] Re: 1.8.7 release next month

[Urabe Shyouhei]

Hi Luis.  Sorry for this late reply but I finally transplanted tk's
extconf.rb from trunk to 1.8.7.  Can you see if it works?  I will
release it when it is OK.

[ruby-core:37706] Re: 1.8.7 release next month

[Luis Lavena]

On Sun, Jun 26, 2011 at 11:52 PM, Urabe Shyouhei <shyouhei@ruby-lang.org> wrote:
> Hi Luis.  Sorry for this late reply but I finally transplanted tk's
> extconf.rb from trunk to 1.8.7.  Can you see if it works?  I will
> release it when it is OK.
>

Hello Urabe-san,

As replied privately to you and Hidetoshi NAGAI, ruby_1_8_7 branch
seems to be working for us and we will handle the TCL_LIBRARY issue in
RubyInstaller itself.

Thank you again for your patience and work in the backports.

-- 
Luis Lavena
AREA 17
-
Perfection in design is achieved not when there is nothing more to add,
but rather when there is nothing more to take away.
Antoine de Saint-Exupéry

[ruby-core:37711] Re: 1.8.7 release next month

[Urabe Shyouhei]

(07/01/2011 09:42 AM), Luis Lavena wrote:
> On Sun, Jun 26, 2011 at 11:52 PM, Urabe Shyouhei <shyouhei@ruby-lang.org> wrote:
>> Hi Luis.  Sorry for this late reply but I finally transplanted tk's
>> extconf.rb from trunk to 1.8.7.  Can you see if it works?  I will
>> release it when it is OK.
>>
> 
> Hello Urabe-san,
> 
> As replied privately to you and Hidetoshi NAGAI, ruby_1_8_7 branch
> seems to be working for us and we will handle the TCL_LIBRARY issue in
> RubyInstaller itself.
> 
> Thank you again for your patience and work in the backports.


Thank you Luis!  Then I release it this weekend.