7775-8407 subjects 8175-12006

[ ruby-Bugs-4731 ] bug in newton.rb
7981 [noreply@ru y] Bugs item #4731, was opened at 2006-06-10 19:09

Validity of local VALUEs
7982 [hisham.hm@gm] I've been wondering about the lifetime of Ruby objects declared from C
7983 [aredridel@nb] They are indeed bound somewhere: the stack. The Garbage Collector scans

New GC for Ruby
7984 [cmills@fr es] I have been researching GC options for Ruby.  I would like to throw

[ ruby-Bugs-4740 ] Ruby 1.8.4 bug with File.stat( fileName )
7985 [noreply@ru y] Bugs item #4740, was opened at 2006-06-12 09:28
7986 [usa@ga ba ec] What platfrom do you use?

[ ruby-Bugs-4753 ] Bug Segmentation fault with MSScriptControl
7989 [noreply@ru y] Bugs item #4753, was opened at 2006-06-13 13:32

ENV['x'] = nil doesn't work in 1.9
7990 [drbrain@se m] $ ruby19 -ve 'ENV["x"] = nil'
7991 [drbrain@se m] Ah, I see, hash.c revision 1.150 references [ruby-list:40865] (which

[PATCH] SSL client certificate support for OpenURI
7992 [mvandegrift@] ...
8003 [akr@m1 n. rg] 1. Why :ssl_cert and :ssl_key are separated?
8006 [mvandegrift@] ...

Net-Ping Timeout Error
7993 [anoopcj@gm i] ...

Ruby Kaigi date confusion
7994 [headius@he d] ...
7996 [matz@ru y- a] (1) Yes, YARV will support only native threads, unless someone persuade
+ 7997 [headius@he d] ...
+ 7998 [steven@lu os] How would one begin to help with progress? Pick something from ToDo
  7999 [matz@ru y- a] That would help.  Discussion on the list would also push progress.

method :do do
7995 [joaopedrosa@] ...
+ 8000 [shiba@ma l2 ] (This may be my first post for this list.)
+ 8002 [nobu@ru y- a] Fixed now.  Thank you for the report.

[ ruby-Bugs-4782 ] Net::HTTP#post does not behave as expected since ruby 1.8.4
8001 [noreply@ru y] Bugs item #4782, was opened at 2006-06-16 12:46

[PATCH] Test::Unit::Assertions RDoc
8008 [drbrain@se m] ...
8009 [matz@ru y- a] It seems OK for me.  Could you check it in (for both HEAD and 1.8)?

small patch for drb/drb.rb. (fwd)
8012 [hgs@dm .a .u] People on Ruby-Doc tell me this sort of thing should go to
8021 [drbrain@se m] This patch was in 1_8 but not HEAD, so I committed it to HEAD.

[OT] Download page on ruby-lang has numeric URL
8013 [hgs@dm .a .u] Would it be possible to alias
8022 [drbrain@se m] Note that that page has been updated many times since 20020102...
8024 [hgs@dm .a .u] Then may I suggest that the name is misleading?  Also, on the web
+ 8026 [asbradbury@t] I'll agree that the name isn't ideal, but this is fixed in the new ruby-lang
+ 8027 [meta@po ox c] You have learned two falsehoods. See
  8028 [hgs@dm .a .u] A lot of people don't know that servers such as Apache give you a
  8030 [meta@po ox c] Yes, in practice all URLs change eventually.

[ ruby-Bugs-4805 ] ruby 1.8.4 does not compile on win32 using VC8
8015 [noreply@ru y] Bugs item #4805, was opened at 2006-06-19 15:57

Re: ruby 1.8.4 does not compile on win32 using VC8
8016 [usa@ga ba ec] This problem has already fixed on CVS.

[PATCH] rb_iterate_proc()
8017 [dbatml@gm .d] ...
8025 [nohmad@gm il] This is very cool. Didn't anyone confirm this patch yet?

Announcing Beta Release of GP Ruby.NET Compiler
8023 [w.kelly@qu .] We are pleased to announce the preliminary Beta release of the Gardens Point

[patch] code-cleanup: K&R-Style functions and const
8029 [stefan@hu hn] attached is another cleanup patch. This one replaces most of the
8031 [meta@po ox c] I thought the use of K&R rather than ANSI C was a deliberate policy of
8032 [matz@ru y- a] It was, but I abandoned it for 1.9.  And thank you for the patch,

[ ruby-Bugs-4820 ] XMLRPC's conv2value does not handle DateTime at all (it chops the time off)
8033 [noreply@ru y] Bugs item #4820, was opened at 2006-06-20 15:12
8034 [drbrain@se m] On Jun 20, 2006, at 3:02 PM, <noreply@rubyforge.org>

mkmf.rb issue (and patch) on FreeBSD 6-STABLE
8035 [alexb@ne se ] ...

[PATCH] File.mkfifo File.mknod
8036 [guillaume.pi] ...

Bug in net/http
8037 [arnaud@me re] Just to point out a small bug in net/http.rb, dating back from rev 1.68
8051 [aamine@lo er] Thank you very much.  It seems Matz already

bug in $. ?
8038 [wybo@se va y] def out; print "#$. "; end # print $. plus a space
+ 8039 [pit@ca it in] $. is the number of the last line read from the current input file. When you
| 8040 [wybo@se va y] 1. because DATA does not point to my input file `t', but to the DATA
| 8042 [pit@ca it in] DATA does indeed "point to" the input file, right after the __END__. You can
| + 8043 [wybo@se va y] Sure, but that is part of the bug, in my opinion. After DATA.rewind I
| | + 8045 [pit@ca it in] Makes sense. According to Matz's reply, he sees a bug somewhere, too. Maybe he
| | | 8048 [drbrain@se m] There is only one DATA, $0's DATA.  There is not a DATA per __FILE__.
| | | 8049 [pit@ca it in] Thanks Eric, I wasn't aware of that. So Wybo is right that in order to read the
| | + 8046 [james@gr yp ] Just FYI, Perl also allows you to reach the source via DATA.
| | + 8047 [sroberts@un ] You mentioned perl before, note that this is identical to perl, you can
| + 8044 [decoux@mo lo] Well try this
+ 8041 [matz@ru y- a] No.  It is very likely a bug.  I will examine it.

Thank-you to the Rails Core Team
8050 [devlists-rub] While we were listening to Dave Thomas' Keynote address today at
8052 [ mfp@ac .o g] Did you mean to send this to rails-core@lists.rubyonrails.org ?
8053 [devlists-rub] I had figured everyone on this list was at RailsConf :)
8054 [dblack@wo bl] Actually only two of the top ten posters are here (me and Dan Berger).

Segfault in rb_ary_push, ruby 1.8.4 (2005-12-24) [powerpc-openbsd3.8]
8056 [sstephenson@] ./script/../lib/cardinal/handler.rb:30: [BUG] Segmentation fault
8059 [drbrain@se m] Do you have a script to reproduce this?  Three lines of ruby aren't

Possible bug in $?
8057 [headius@he d] ...
8058 [headius@he d] ...
8060 [headius@he d] ...

Win32 Extension Issues Wanted!
8061 [halostatue@g] Everyone. I had a conversation with folks from Microsoft today about
+ 8062 [usa@ga ba ec] The problem of errno is not serious, I think.
| 8063 [shyouhei@ic ] * The main concern is that each DLL holds their own file descriptors
| 8064 [ncannasse@mo] I agree with all the statements here.
+ 8285 [nobu@ru y- a] Is this still open?

Core documentation patches
8065 [alex@bl ck e] I've got a little documentation patch for 1.8.5...  Should it go here?
8067 [matz@ru y- a] Yes, drbrain or myself would commit the patch.
8068 [alex@bl ck e] ...
8069 [drbrain@se m] Should this me mentioned?
8070 [lukfugl@gm i] To head of those that might ask (since I was about to before sitting
+ 8071 [alex@bl ck e] Yes - that's where I originally found the problem.  I was assuming it
+ 8072 [drbrain@se m] o = Object.new

and darwin
4804 [laurent.sans] Apparently the change introduced by [rubycore:04804] does not work on
8093 [matz@ru y- a] Yes.  More precisely, it should work for the systems that
8095 [laurent.sans] Strange that it does not work here.  Almost (all?) BSD-based systems

1.8.5p1 build failure on Solaris 10
8073 [Daniel.Berge] Solaris 10
8074 [matz@ru y- a] In that case, configure should have used replacement implementation
+ 8075 [andrew.c.mor] Dan -
| + 8081 [Daniel.Berge] Thanks Andrew, it seemed to work for 3.4.5 as well.
| + 8085 [matz@ru y- a] Ah, I almost forgot it, sorry.  Would someone using Solaris review
|   8086 [Daniel.Berge] ...
+ 8077 [Daniel.Berge] ...
+ 8078 [Daniel.Berge] config.status: WARNING:  Makefile.in seems to ignore the --datarootdir setting
  + 8079 [decoux@mo lo] Well I don't know if it's important but in configure.in there is
  | + 8083 [matz@ru y- a] It is to fix [ruby-core:7138], that is
  | + 8096 [ville.mattil] This is needed so that the sun's forte compiler works correctly. We should
  |   8097 [matz@ru y- a] Can you write a patch for us?
  |   8098 [ville.mattil] should
  |   8099 [matz@ru y- a] Thank you for your contribution.  The preview2 is scheduled for July 14th.
  |   8100 [ville.mattil] vacation
  |   8101 [Daniel.Berge] Actually, shouldn't we test for the feature instead of the platform?  I don't
  |   + 8104 [ville.mattil] don't
  |   | 8106 [Daniel.Berge] I don't totally understand how the build process works, which is part of my
  |   | 8108 [ville.mattil] my
  |   + 8105 [matz@ru y- a] Actually this hack is added to resolve usual test failure for isinf
  |     8107 [ville.mattil] This does not restore the test if gcc is not used. I tested this patch
  |     8109 [matz@ru y- a] I didn't mean your patch.  I was saying about removing previous modify
  + 8080 [nobu@ru y- a] No, just ignore.  And it's fixed in CVS already.

Extensions and symbols
8076 [vincent.four] ...

Two test failures with 1.8.5p1, Solaris 10
8082 [Daniel.Berge] Solaris 10
8084 [matz@ru y- a] This one is fixed on CVS today.

[ ruby-Patches-4879 ] optparse.rb to RDoc documentation patch
8087 [noreply@ru y] Patches item #4879, was opened at 2006-06-28 20:50
8088 [drbrain@se m] On Jun 28, 2006, at 11:50 AM, <noreply@rubyforge.org>
8089 [robin@ni or ] The reason I chose '_' is because in core documentation it is done that
+ 8090 [robin@ni or ] Done.
| 8092 [drbrain@se m] Committed.
+ 8091 [drbrain@se m] I believe the core docs are older.  Maybe one day I'll get really

[ ruby-Bugs-4885 ] Ambiguous error in OpenSSL with 1.8.4 on Windows XP
8094 [noreply@ru y] Bugs item #4885, was opened at 2006-06-29 12:33

Reorganizing configure.in by platform?
8102 [Daniel.Berge] After looking at the docs for autoconf (ick), would it help to have separate
+ 8103 [sroberts@un ] A primary design goal of the autoconf tools is that they do not have to
+ 8115 [steven@lu os] But... it's sort of the point of autoconf to not do that.  And also
  8116 [Daniel.Berge] Ah, ok.  Disregard.

Solaris patches for configure.in work!
8110 [Daniel.Berge] 7791) and both worked fine against the nightly snapshot (I messed up Andrew's
8111 [ville.mattil] with

[ ruby-Bugs-4898 ] Subclasses of IO do not ensure that IO#close is called in block form
8112 [noreply@ru y] Bugs item #4898, was opened at 2006-06-30 08:18

Ruby 1.8.5 - warnings when compiled with Sun compiler
8113 [Daniel.Berge] ...
8117 [nobu@ru y- a] It would be GLOB_VERBOSE.
8121 [matz@ru y- a] Can you commit?
8122 [ville.mattil] This is is correct fix. I included it in my 64 bit warning clean up

Solaris, OpenSSL - patch #1 - _FILE_OFFSET_BITS
8114 [Daniel.Berge] This is a simple patch for ossl.h where I just moved around the ruby.h include

[ ruby-Bugs-4912 ] ri crash after running rdoc on ActiveSupport
8118 [noreply@ru y] Bugs item #4912, was opened at 2006-07-02 05:08
8119 [drbrain@se m] On Jul 1, 2006, at 10:08 PM, <noreply@rubyforge.org>

[PATCH] ruby 1.9 HEAD 64 bit warnings clean up
8120 [ville.mattil] This patch fixes the follwing  warnings sun studio 11 cc warnings.

Unit/Regression tests for Ruby builtin classes and modules
8123 [w.kelly@qu .] As part of our Gardens Point Ruby.NET project we have spawned off a
+ 8124 [djberg96@ya ] I actually started something like that already.  See
| 8125 [headius@he d] ...
+ 8126 [drbrain@se m] We have a ton of tests for the core libraries already done for BFTS

Build target for core lib only?
8127 [Daniel.Berge] Is there a make target that would let me build and install only the core Ruby
8162 [nobu@ru y- a] No for only objects, but there are lib and dll targets.  They

Patch for time.c
8128 [Daniel.Berge] This patch eliminates the "integer overflow detected" warnings in time.c when

[ ruby-Bugs-4949 ] segmentation fault while evaluating printf:Kernel
8129 [noreply@ru y] Bugs item #4949, was opened at 2006-07-05 18:03
+ 8130 [Daniel.Berge] This must be against 1.9, because in 1.8.4 and the 1.8.5 nightly build I get
+ 8186 [matz@ru y- a] It's fixed in the recent version of Ruby.

thread mystery
8131 [ara.t.howard] any idea why this script slows downs drastically as it runs and seems to leak
+ 8135 [hgs@dm .a .u] I'm not familiar with enough with sync to answer this I suspect, esp. the
+ 8143 [ser@ge ma e-] Slow compared to what?  Number of calls per second?
  8144 [ara.t.howard] slow compared to how fast it starts - eventually it actually will halt.

[PATCH] rdoc, C extensions, stop and start
8132 [Daniel.Berge] Here's a patch that lets you use '--' and '++' within C extension comments to
8158 [matz@ru y- a] I'm not sure whether '--' and '++' appropriate for switch the
8160 [halostatue@g] # documented
+ 8166 [Daniel.Berge] As Austin mentioned, this is exactly how it works now for pure Ruby
| 8168 [hgs@dm .a .u] which is about the sytax of the markup?
+ 8173 [matz@ru y- a] Then, it's OK, if Ryan agreed.

[BUG] sync.rb memory corruption
8133 [ara.t.howard] ...
8147 [nobu@ru y- a] I couldn't reproduce it with recent version.  Can't you try

Re: [ ruby-Bugs-4949 ] segmentation fault while evaluating
8134 [tom@in oe he] Done!

[Bug?] Confused exception handling in Continuation Context
8136 [robert.dober] ...
+ 8137 [nobu@ru y- a] @next isn't set more after @enum.each ended.  Therefore i.next
| 8139 [robert.dober] ...
+ 8138 [pit@ca it in] Robert, just change your test code to
  8140 [robert.dober] ...
  + 8148 [matz@ru y- a] No.  It's about the continuation.
  | 8155 [robert.dober] ...
  | 8156 [matz@ru y- a] This is the point.  A continuation is like a time machine.  By calling
  | 8157 [robert.dober] ...
  | 8159 [matz@ru y- a] This is what continuation is.  This behavior is defined elsewhere, so
  + 8150 [pit@ca it in] I think it has been explained here or on ruby-talk, but I don't have the
    8154 [robert.dober] ...
    8163 [pit@ca it in] Oh, no need to apologize, Robert! Continuations certainly aren't easy.

Results of a build on Solaris9
8141 [hgs@dm .a .u] ...
8145 [hgs@dm .a .u] ...

[BUG] thread/sync.rb memory corruption
8142 [ara.t.howard] ...
8146 [shyouhei@ic ] Could someone please confirm this can be reproduced on 1.8.5 pre1?
8149 [hgs@dm .a .u] brains hgs 123 %> ./!$
8151 [ara.t.howard] if you don't have electric fence you'll need to run for quite a long time to
+ 8152 [ara.t.howard] ...
| 8187 [hgs@dm .a .u] ...
+ 8153 [hgs@dm .a .u] Oh!  I see. I was thinking that the gets was to keep the console open

[PATCH] --fqname option to test/unit/autorunner.rb
8161 [dev@sa iz at] ...
8165 [Daniel.Berge] It's not a bad idea, though in practice I would say it's unusual to have
8189 [dev@sa iz at] Daniel --
8202 [dev@sa iz at] Daniel --
8203 [Daniel.Berge] Not unless Matz or Nathaniel Talbott approve it.
8206 [dev@sa iz at] Thanks again for getting back to me. Perhaps my original focus on the
+ 8207 [meta@po ox c] Could you expand on why? It never occurred to me that there might be a
| 8209 [Daniel.Berge] Since we're going OT, I've moved my reply over to ruby-talk.
+ 8210 [sroberts@un ] For what its worth, I think your point makes sense.
  + 8211 [Daniel.Berge] Ok, just to clarify.  You've got a file that does something like this?
  + 8213 [Daniel.Berge] ...
    8215 [dev@sa iz at] Sam and Daniel --

Solaris9 and nightly snapshot.
8164 [hgs@dm .a .u] ...

[ ruby-Bugs-4970 ] bug in printf
8167 [noreply@ru y] Bugs item #4970, was opened at 2006-07-07 14:18
8401 [nobu@ru y- a] I think the latter would be a bug.
+ 8404 [matz@ru y- a] Can you commit the patch please?
+ 8408 [lukfugl@gm i] I'm curious, is there a reason why the former (number is converted,
  8410 [matz@ru y- a] Unlike C, Ruby integers does not have artificial limit of 32bits nor
  8411 [lukfugl@gm i] Ok, that makes sense. Should we maybe patch the documentation on
  8415 [matz@ru y- a] Perhaps.  Would anyone volunteer to make a patch?
  8418 [lukfugl@gm i] How's this look? It's probably too verbose, and I'm not even sure
  8420 [matz@ru y- a] Thank you.

[BUG] next in ensure body (Ruby 1.8.x)
8169 [dbatml@gm .d] => nil
8170 [decoux@mo lo] Not really,
+ 8171 [dbatml@gm .d] That's interesting, it did work in May (I didn't install a current version
+ 8172 [nobu@ru y- a] Hmmm, makes sense.
  8174 [matz@ru y- a] Can you commit?