4555-4797 subjects 4919-5394

^ snapshots broken?
4748 [ntalbott gma] -rw-r--r--   1 snapshot nogroup        45 Apr 20 19:00 snapshot.tar.gz
4749 [matz ruby-la] Because all CVS service is stopped now for CVS vulnerability.

^ Illegal regexp causes segfault
4751 [andrew walro] irb(main):058:0> a = /\[([^]]*)\]/
+ 4752 [mailing-list] irb
| 4758 [ocean m2.ccs] Andrew, can you see us a backtrace?
| + 4761 [decoux moulo] Well, I've no free time to look at it but perhaps it's an alignement
| + 4763 [andrew walro] (gdb) set args -d -e 'a = /\[([^]]*)\]/'
|   4764 [decoux moulo] Another example
|   4767 [decoux moulo] An example to see why I dont think that the problem is regex.c
+ 4765 [akr m17n.org] Try following patch.
  4766 [decoux moulo] uln% grep VERSION *.h
  4768 [akr m17n.org] Oops.
  4769 [decoux moulo] yes, it fix it
  4770 [ocean m2.ccs] (2005/04/24 20:03)
  4771 [decoux moulo] Yes, I think.

^ [ ruby-Bugs-1823 ] NUM2UINT does not rase RangError for negative Fixnum
4753 [noreply ruby] Bugs item #1823, was opened at 2005-04-22 11:34

^ Re: Welcome to our (ruby-core ML)         You are added automatically
4754 [odm aitop.co] ...

^ date/holiday.rb
4756 [ntalbott gma] Is there a reason holiday.rb from the date2 library has not been
4759 [kjana dm4lab] On my own understanding, it's just for Japanese holidays + some (such
4773 [ntalbott gma] While the Japanese holidays are not necessarily generally applicable,

^ BUG: rexml/encodings/Shift-JIS.rb
4772 [speakillof y] It seems that rexml/encodings/SHIFT-JIS.rb is broken.
4780 [ocean m2.ccs] (2005/04/24 22:13)

^ [PATCH] enhanced $0 modification
4774 [evanwebb gma] The attached patch allows for ruby to use more of the available stack
4796 [nobu.nokada ] It resulted SEGV with bccwin.  And envp doesn't seem necessary.
+ 4799 [matz ruby-la] Can you commit if you feel it appropriate?
| 4805 [nobu.nokada ] I'd just commited [ruby-core:04804] to trunk in the meantime.
+ 4800 [akr m17n.org] ProFTPD is GPLed.
  4801 [evanwebb gma] The update patch shares no code with original ProFTPD, and thus would
  4804 [nobu.nokada ] Yes, to HEAD of trunk.  And it didn't work after modifying ENV.

^ profiler.rb Schroedinbug
4775 [erlercw gmai] A ruby program with the single instruction "require 'profile'"
+ 4776 [ocean m2.ccs] Maybe, did you require "profile.rb" on irb?
| 4777 [erlercw gmai] Neither irb nor ruby -rprofile cause the bug for me.  It happens when
| 4778 [ocean m2.ccs] Strange. I works for me ...
| 4779 [erlercw gmai] ruby 1.8.2 (2004-12-25) [i386-mswin32]
| 4781 [ocean m2.ccs] (2005/04/25 14:11)
| 4782 [ocean m2.ccs] I don't know if this comes from same issue but...
+ 4785 [florgro gmai] This is AFAIk caused by rubygems which sets the RUBYOPT environment
  4793 [ocean m2.ccs] Like this? Seems to solve both rubygems issue and [ruby-core:4782].

^ Extending a SimpleDelegator
4790 [transfire gm] I noticed that Delegator (delegate.rb) does not work as one might
5199 [matz ruby-la] SimpleDelegator creates an object that delegates every method of the
5203 [transfire gm] Wow! I had all but forgotten about this. Thanks for getting back to

^ Switching to Subversion [Was: Re: Win32: Ruby & APR; build problems for Ruby Subversion SWIG bindings]
4792 [drbrain segm] As soon as the debian packages are updated.
4798 [ehuels gmail] Heh, yes ofcourse: doesn't ruby get fixes?

^ [ANN] Anonymous CVS Service Restart
4802 [shugo ruby-l] Anonymous CVS Service was restarted.
4803 [daniel.amela] No, thank YOU!

^ -Wall
4806 [k.shutemov s] Why does ruby build without -Wall in CFLAGS by default? -Wall can help to
4807 [vincent.isam] I think it's because there may still be warnings on not well tested
+ 4808 [k.shutemov s] to
| 4809 [Daniel.Berge] There "may still be warnings"?  No, there *are* warnings.  Quite a few,
+ 4810 [mailing-list] When are warnings a bad thing?  It's not like they stop things from
  4811 [vincent.isam] I never said that warnings are a bad thing. It's a good tool for developers.
  + 4812 [perlguy gmai] And by not showing us the warnings, you foster a false sense of
  | 4824 [jlsysinc all] -Wall on GNU gcc complains about perfectly legal C code that doesn't need
  | 4829 [peterreilly ] All warnings are legal C code - some may need fixing. For example, I
  + 4813 [mailing-list] (I wasn't confroning you personally, if that's how you interpreted it.)
    4814 [vincent.isam] No problem :)
    4815 [nobu.nokada ] I confirmed many "parentheses around assignment" warnings and
    + 4816 [k.shutemov s] to
    + 4817 [nobu.nokada ] I know, and they're all OK.
    | 4818 [mailing-list] Well, why not initialize them to values that make sense?
    | 4823 [nobu.nokada ] Suppressing warnings itself is not the goal.
    + 4831 [matz ruby-la] Can you commit these fixes, if you haven't committed yet?

^ Enhancing backtraces
4819 [emiel il.fon] I've just read a post on Peter Williams' blog about his annoyance
4822 [transfire gm] I would go even further and propose that a more OOP approach be taken.

^ [PATCH] for [ #1823 ] NUM2UINT and rb_num2ull do not rase RangeError for negative Fixnum
4820 [mschonberg p] This is a multi-part message in MIME format.
4821 [tom infoethe] When someone applies this, can they pls also close the bug here?

^ mysql_adapter question
4825 [skaes gmx.ne] puzzled by it, I'll post the question here. In mysql_adapter.rb, the
4826 [skaes gmx.ne] Sorry, wrong mailing list. My apologies.

^ Re: [ ruby-Bugs-1859 ] URI.escape/unescape does the wrong thing with spaces
4828 [s-urabe par.] No. uri.rb is based on RFC 2396 and this RFC does not define '+' as

^ Trouble compiling ruby 1.8.2 on Intel ia64
4830 [sam.overdorf] charset="us-ascii"
4832 [Daniel.Berge] Redhat 4?!
4833 [k.shutemov s] I think, it's Red Hat Entrprise Linux 4
4921 [sam.overdorf] charset="us-ascii"
4923 [djberge qwes] Got a config.log file?

^ Fwd: Re: Whats so different about a Hash?
4834 [andrew walro] ri Object.hash says
4835 [decoux moulo] What is your version of ruby
+ 4836 [andrew walro] $ ruby --version
| 4838 [decoux moulo] Can you update to the latest stable
| 4839 [andrew walro] See other message - same result as you.
| 4840 [decoux moulo] svg% ./irb
| + 4841 [andrew walro] So 1.9 does what I would consider to be the _right thing_ :)
| + 4842 [andrew walro] I assume {1=>2}.hash == {1=>2}.hash on 1.9 ?
|   4843 [decoux moulo] yes,
+ 4837 [andrew walro] irb(main):001:0> a={1=>2}

^ Getting rid of Object#equal?()?
4844 [florgro gmai] Moin.
+ 4845 [matju arteng] This message is in MIME format.  The first part should be readable text,
| 4847 [florgro gmai] I think backwards-compatibility is not too much of a problem -- we don't
+ 4846 [dooby d10.ka] I thought Object#id was deprecated.
+ 4850 [matz ruby-la] I'm not motivated too much to remove "equal?".  I assume Ruby users
| + 4851 [ruby-ml magi] I suppose it is not a huge source of bugs; however, the term
| | 4852 [matz ruby-la] If it's not a source of bugs, why bother?  Lisp have even more (=, eq,
| | 4854 [ruby-ml magi] Not a *great* source of bugs, no, although I am sure that it is not
| + 4853 [florgro gmai] While they are documented it can be irritating to have to look up what
| + 4876 [gsinclair so] I'm not so smart; I've had to consult the documentation several times
|   4878 [ruby-ml magi] I do not think any methods should be removed if it can be avoided.
+ 4877 [dblack wobbl] This message is in MIME format.  The first part should be readable text,
+ 4880 [lists bertra] If there is a source of weird bugs it's that the method

^ [ ruby-Bugs-1878 ] No way to change socket timeout on opened URI
4848 [noreply ruby] Bugs item #1878, was opened at 2005-05-05 17:55
4867 [akr m17n.org] Does adding :read_timeout option solve the problem?
5698 [ocean m2.ccs] This is old entry, but Nicholas was waiting for this on RubyForge without notice

^ [BUG?] Array#slice!() adds elements to Array
4849 [florgro gmai] Moin.
4873 [transfire gm] Thanks Florian. I'll look it over. I'm just getting back to work now.

^ Method hooks in singleton classes
4855 [pit capitain] currently the hooks method_added, method_removed, method_undefined are
+ 4859 [nobu.nokada ] What will you need if method_added is redefined too?
| 4860 [florgro gmai] Doesn't that trigger an event in the old method_added or
+ 4861 [matz ruby-la] In that case, we have to decide to make "mataclass" in the language
  4870 [pit capitain] thank you all for your replies.

^ Segfault in regular expression parsing on x86_64
4856 [ser germane-] charset="utf-8"
4857 [decoux moulo] it was corrected

^ [ ruby-Bugs-1883 ] Build fails on OSX Tiger 10.4
4858 [noreply ruby] Bugs item #1883, was opened at 2005-05-06 14:55
4862 [matz ruby-la] This means libreadline on Mac OSX Tiger has no rl_attempted_completion_over
+ 4863 [ruby-lists l] its not in tiger at all. see his -I's
| + 4864 [nobu.nokada ] Then the extension shouldn't be compiled.  The -I flags contain
| + 4866 [s-urabe par.] so why "-DREADLINE_21_OR_LATER"?
+ 4865 [ryand-ruby z] $   */
  4868 [nobu.nokada ] Hmmm, it looks like the file contained in libedit.
  5053 [shugo ruby-l] libedit on my machine (Debian sid) seems to lack
  5056 [discordantus] It doesn't work on my tiger install. libedit's header file seems to be
  5060 [shugo ruby-l] The original libedit provides a compatibility layer for the GNU Readline
  + 5064 [discordantus] For some reason, my /usr/include/readline dir is empty. Since I just
  + 5118 [ryand-ruby z] I've finally had time to dig deeper into this. I just did a make
  | 5128 [shugo ruby-l] It looks like GNU Readline, not libedit.
  | 5129 [ryand-ruby z] /usr/include/readline
  | + 5130 [nobu.nokada ] Do those headers contain FSF copyright notice and
  | + 5131 [discordantus] Actually, it doesn't. They have stated publicly that they can't ship
  | | 5132 [ryand-ruby z] DOH.
  | + 5133 [shugo ruby-l] Thank you.
  |   5134 [shugo ruby-l] I got an account on his machine. Thank you, Ryan.
  |   5135 [ryand-ruby z] *sigh*
  + 5119 [ryand-ruby z] % rm -f .ext/powerpc-darwin8.1.0/readline.bundle ext/readline/

^ [BUG] Infinite loop on YAML.dump (Re: ruby-list:40801)
4869 [ocean m2.ccs] Kouhei Yanagita reported infinite loop bug on YAML.dump.
4956 [ruby-core wh] This is fixed in Syck 0.5x.  Can you try with Syck from CVS?
4959 [ocean m2.ccs] I tried Syck 0.54, and seems working.
4960 [ocean m2.ccs] But be care, I've only tested that attached file.

^ [ ruby-Bugs-1890 ] URI.parse fails to encode/decode IPv6 addresses
4871 [noreply ruby] Bugs item #1890, was opened at 2005-05-09 22:03

^ [ ruby-Patches-1893 ] Aliases for sec and min in date.rb
4872 [noreply ruby] Patches item #1893, was opened at 2005-05-10 08:54
4875 [matz ruby-la] This is merged to the CVS HEAD (1.9).

^ [THIN] - Need to reduce Ruby Sources to the Minimal
4874 [ilias lazari] After becoming familar with the Object Model, I'm ready to dive into the
+ 4879 [pit capitain] When building Ruby, one of the first things created is a "miniruby"
| 4883 [ilias lazari] very complicated those makefiles.
| + 4884 [ryand-ruby z] This really isn't a ruby-core issue.
| | 4888 [ilias lazari] I like to dive into the ruby source-code base.
| | 4889 [ruby-ml magi] The Ruby Makefile, while not an artistic achievement in itself,
| | 4890 [ilias lazari] Feel free to open a new thread to discuss processing-models.
| | 4891 [ruby-lists l] if illias isn't blocked from this mailing list soon
| | + 4893 [ilias lazari] Feel free to open a new thread to ask for my 'execution'.
| | + 4899 [matju arteng] This message is in MIME format.  The first part should be readable text,
| |   4903 [ilias lazari] This is a development list.
| |   4905 [fg siamecomm] <smile> Readers of development list should know about 'make'</smile>
| |   4910 [ilias lazari] if the project encapsulates the "make" domain-knowledge correctly,
| |   4934 [nobu.nokada ] What will you propose instead?
| |   4937 [ilias lazari] see below
| |   4938 [nobu.nokada ] Where?  I meant alternatives to "make".
| |   4941 [ilias lazari] I do not propose an alternatives.
| |   4950 [nobuyoshi.na] 1.8.3 preview1 has been released, and you need to use latest version
| |   4965 [ilias lazari] a) I don't act as a developer at this point.
| + 4892 [nobu.nokada ] What do you expect by "thin"?
| | 4895 [ilias lazari] I will try.
| | + 4897 [mailing-list] That”Ēs all the files found in the root of the Ruby source.
| | | 4902 [ilias lazari] ok
| | | 4912 [mailing-list] Wow, I actually agree with you on something :-),
| | | 4917 [ilias lazari] that's nice.
| | + 4900 [hgs dmu.ac.u] Do you mean core libraries, not standard libraries -- exclude things
| |   4906 [ilias lazari] would be a good starting point - and a code-unit to isolate.
| + 4896 [mailing-list] Actually, this is a rather simple Makefile and it”Ēs simple to
|   + 4898 [pit capitain] Don't worry, Nikolai, none taken.
|   + 4904 [ilias lazari] This tool poduces terrible make-files, thus it's not of intrest to.
+ 4988 [ilias lazari] lists all files which are linked to a preliminary "miniruby" executable.

^ Encoding Pragma
4881 [lists bertra] I asked for that on ruby-talk but got no satisfying answer.
4882 [nobu.nokada ] Although ruby doesn't have pragma right now, I'm using this
4885 [matz ruby-la] Does this patch switch encodings in the middle of the file?
4894 [nobu.nokada ] Yes.
4901 [matz ruby-la] Encoding pragma in the middle of the file should be just ignored, I
4908 [nobu.nokada ] I'm wondering how "at the top" can be detect.  ruby_sourceline
+ 4909 [matz ruby-la] Aoki-san will take care of ripper.  ;-)
| 4935 [nobu.nokada ] OK, this patch seems to work.
+ 4920 [aamine lover] They are defined differently, but they work in the same manner.

^ ruby 1.8.3 preview1
4886 [matz ruby-la] I finnally published the first preview to the 1.8.3.
+ 4907 [tom infoethe] Yours,
+ 4928 [tilman code-] What about [ruby-core:4296]?
| 4948 [ocean m2.ccs] I noticed patches under [ruby-core:4296] were already commited by matz.
| 4952 [tilman code-] Sweet! I also noticed some other RDoc patches I have submitted have been
| 4954 [tom infoethe] that have been fixed or applied, please let me know and I'll close them.
| 4994 [ocean m2.ccs] I want to close http://rubyforge.org/tracker/index.php?func=detail&aid=1900&group_id=426&atid=1700
| 5019 [tom infoethe] Could you pls try to close the bug again?  I think it should work now,
+ 4973 [djberge qwes] Solaris 10

^ [ ruby-Bugs-1898 ] FXMemoryBuffer::data gives FXMemoryBuffer::to_s
4887 [noreply ruby] Bugs item #1898, was opened at 2005-05-12 09:37

^ [ ruby-Patches-1900 ] Pointless argc check in Array#select
4911 [noreply ruby] Patches item #1900, was opened at 2005-05-12 09:33
+ 4913 [djberge qwes] Dang it, it seems that removing that causes a segfault after all.
| 4914 [decoux moulo] You have modified rb_define_method() but not modified static VALUE
| 4915 [djberge qwes] Yep, I caught that and posted a followup.  Thanks.
| 4916 [decoux moulo] Why do you want that it raise an error ?
| 4918 [halostatue g] Daniel is adding to the test suite for Ruby to verify that things do
| 4922 [djberge qwes] Ah, here's the issue.
| 4926 [dblack wobbl] Since the block would never be executed with an empty array anyway,
| 4927 [dblack wobbl] [].select
| 4929 [matju arteng] This message is in MIME format.  The first part should be readable text,
| 4930 [Daniel.Berge] Ok, I can accept that. :)
| 4946 [ocean m2.ccs] I've commited your patch. Thank you.
| 4947 [djberg96 yah] assert_raises(LocalJumpError){ @a.find_all }
| 4949 [ocean m2.ccs] OK, I've commited the test after excluded assert_raises.
+ 4924 [ryand-ruby z] Can you clean up this and any other pure whitespace diffs?