6473-6830 subjects 6966-7294

[BUG] String#index does not work correctly on SuSE10.0 x86_64
6721 [Kanis@CO CA ] ...
6730 [decoux@mo lo] Can you test ?
+ 6744 [djberg96@ya ] Looks good here (Suse Linux 9.3).
+ 6746 [kanis@co ca ] charset="utf-8"
  6747 [matz@ru y- a] I've committed in.

6739 [wink@sa il e] With the post concerning Neko generating some interest, I thought this

[ ruby-Bugs-2872 ] TCPServer should not use SO_REUSEADDR in Cygwin port
6765 [noreply@ru y] Bugs item #2872, was opened at 2005-11-23 11:36

Re: semantics of if/unless/while statement modifiers
6777 [skaes@gm .n ] I revoke this statment. Your use of RHS is correct. Please excuse me.

[PATCH] rdoc: don't generate duplicate diagrams
6780 [lucas@lu as ] Diagrams generated by rdoc often use a lot of disk space. One of the

Ruby pipes
6781 [ser@ge ma e-] charset="utf-8"
6782 [hgs@dm .a .u] irb(main):001:0> a,b = IO.pipe
6784 [ser@ge ma e-] charset="iso-8859-1"

Documentation for shell.rb
6787 [djberg96@ya ] Any chance of adding some (all?) of this documentation
6788 [gsinclair@gm] Great timing; I just finished my exams, and was thinking this morning

Updating shell.rb and shell/*
6789 [tsumerubycor] Before I submit a patches for documentation, I wanted to ask about file.

[PATCH] ri displaying lists in columns
6792 [g_ogata@op u] While ri is a really nice quick reference, I found it hard to read the
6793 [vjoel@pa h. ] Really lovely! Thank you.

in-block local variables
6794 [w.kelly@qu .] ...
6795 [tsumerubycor] Please use text mode when sending emails. Most of us like our security, and I

[ ruby-Bugs-2892 ] [PATCH] DRbObservable in the CVS-HEAD is not work
6796 [noreply@ru y] Bugs item #2892, was opened at 2005-11-28 21:38

[ ruby-Bugs-2895 ] [PATCH] rdoc generates duplicate diagrams
6797 [noreply@ru y] Bugs item #2895, was opened at 2005-11-28 23:19

ruby 1.8.4 preview2
6798 [matz@ru y- a] Try this, if you please.
+ 6799 [Daniel.Berge] djberge@~/local/src/ruby/ruby-1.8.4-527>make test-all
| 6806 [ocean@m2 cc ] Can you test this with verbose mode? (-v)
| 6807 [Daniel.Berge] test_bracket(TestEnv): ./ruby/test_env.rb:19:in `[]=': failed to allocate
| 6809 [ocean@m2 cc ] Sorry I cannot help you... I've never seen this kind of error.
| 6839 [Daniel.Berge] I'm not going to worry too much about this one.  I'm more worried about the
| 6840 [decoux@mo lo] moulon% tar ztvf ruby-1.8.4-preview2.tar.gz ruby-1.8.4/install-sh
| 6841 [Daniel.Berge] Like I said before, the install target generated in the makefile for extensions
+ 6800 [james@gr yp ] I really want to try this out, just to help you guys find issues, but
| 6805 [ocean@m2 cc ] I fixed this on CVS, so it should be working.
+ 6801 [curt.hibbs@g] ...
| 6810 [nobuyoshi.na] It's due to a bug of Bison 2.1.
| 6811 [usa@ga ba ec] We should deal even if it's a bug of Bison 2.1.
| 6813 [nobu@ru y- a] I don't think so.  Just re-packaging with Bison 1.x, 2.0 or
| + 6815 [matju@ar en ] Wow!
| | 6816 [nobuyoshi.na] Just a wrong way.
| + 6817 [usa@ga ba ec] I agree that ruby distribution should be re-packaged with
|   6818 [nobuyoshi.na] $ bison-2.1 -o parse.c parse.y && sed -i '/^    ;$/d' parse.c
|   6819 [dooby@d1 .k ] +1
|   6820 [nobuyoshi.na] For ripper.
+ 6803 [aaron@sc ra ] YAML loading of quoted Symbols is still broken.
| 6804 [ocean@m2 cc ] yaml/syck is developed on other CVS. maybe this bug is fixed there.
| 6808 [ocean@m2 cc ] This is not word for _why himself, sorry if this looks aggressive.
+ 6821 [ara.t.howard] <nil> expected but was
+ 6822 [kero@ch ll .] $ uname -a
+ 6823 [kero@ch ll .] $ uname -a
| 6831 [nobu@ru y- a] The patch in [ruby-core:06521] would help you?
| 6843 [kero@ch ll .] 1224 tests, 13455 assertions, 0 failures, 0 errors
+ 6826 [drbrain@se m] I am experiencing problems with Ruby and GCC 4 on OS X 10.4.3.  See
  6827 [decoux@mo lo] I can't reproduce
  6832 [drbrain@se m] I have gcc 4.0.0.  Could it be a PPC-only bug?
  6835 [decoux@mo lo] Probably, I've had a bad experience with "la chose" :-)
  6842 [ryand-ruby@z] Guy, I'm gonna set you up an account on one of my servers so you can
  6854 [decoux@mo lo] I can reproduce
  + 6855 [ryand-ruby@z] That would be my fault. The server is very new and has a stock ruby
  + 6861 [decoux@mo lo] There is something strange, that I've not yet understood, and which seems

[ ruby-Bugs-2905 ] Windows XP installation of Ruby causes PATH corruption
6802 [noreply@ru y] Bugs item #2905, was opened at 2005-11-30 19:54

RŮ«.: ruby 1.8.4 preview2
6812 [tad.bochan@b] I have been trying to get 1.8.3 installed on HP-UX, with various C

Pathname implementation using Array
6814 [transfire@gm] Any interest in this reimplementation of Pathname that uses an Array

Ruby + GCC4 on OS X (Was: Strange error messages using DRb/TupleSpace{
6824 [drbrain@se m] Compiling Ruby with GCC 3.3 on OS X 10.4.x causes these problems to
6825 [drbrain@se m] I get the same behavior with 1.8.4p2 and GCC4.0 vs GCC3.3.  GCC4
6983 [steven@lu os] Is it possible that this is related to [ruby-talk:153006]?  I'm

bug in mailread.rb, and: proposal for Mail#to_s
6828 [wybo@se va y] ...
+ 6833 [dooby@d1 .k ] I agree.  It has puzzled me why it's there at all.
| 6834 [wybo@se va y] According to the mbox(5) manpage, it's the so-called "postmark line". That
| + 6836 [dooby@d1 .k ] Ah, thanks.
| + 6838 [meta@po ox c] Just to throw in some information here...
+ 6837 [matz@ru y- a] Interesting.  But I have seen wide range of variety of From line format.
  + 6844 [wybo@se va y] ...
  | + 6845 [matz@ru y- a] Yes, but I'm not sure yet if we don't need to support "software
  | + 6852 [dooby@d1 .k ] def to_print
  |   6858 [wybo@se va y] this suggests that it's meant for printing. Actually, it is the reverse of
  + 6846 [kevin.jackso] sumarize/summarize!
    6851 [zdennis@mk e] If this just for the message body of an email, or is it to display headers to? If it is just for

Re: Proposed patch for optparse to fix multi line argument handling
6847 [hobe@gm il c] ...
6848 [nobuyoshi.na] Thank you for the patch.  That bug has been fixed at Nov 22.
6849 [hobe@gm il c] Doh. I didn't think to check against the development branch.
6850 [nobuyoshi.na] Try it with 1.8.4 preview2.

1.8.4 and respond_to
6853 [caleb@ae -t ] It seems that since 1.8.3 the behavior of rb_respond_to is now defaulting to
6856 [nobuyoshi.na] How broken?
6859 [caleb@ae -t ] if (rb_respond_to(result, rb_intern("initialize")) != 0) {
6860 [matz@ru y- a] It is a bug anyway, since every object has its own initialize method,
6862 [caleb@ae -t ] Agreed - we removed the test.  I think the original (years ago) intention was

profiling class methods
6857 [skaes@gm .n ] During profiling some code using ruby 1.8.2 and standard profiler, I

[PATCH] class document with prototype
6863 [nobuyoshi.na] I found documents of classes don't get generated with 1.9, which uses

ruby 1.8.4 rc breaks alias_method/rails in bad ways
6864 [ara.t.howard] i'm on a box with latest (today's) ruby and rails.  i started seeing bizzare
+ 6865 [ara.t.howard] quick follow up.  tweaking activerecord/base.rb to
+ 6868 [nobuyoshi.na] Don't you have simpler script which can reproduce it?
| 6869 [ara.t.howard] believe me - i tried for about an hour and gave up...  i may have a bit of
+ 6870 [florgro@gm i] Are you certain the file containing the alias_method() call isn't
  + 6871 [ara.t.howard] that's brilliant florian... i'll investigate.  although it does work with
  + 6872 [ara.t.howard] i added
  | 6873 [james@gr yp ] I haven't followed this discussion much, but I've seen similar issues
  | 6874 [ara.t.howard] o.k.  that's pure evil.  any suggestions?
  | + 6875 [ mfp@ac .o g] Look at $" to infer where things are getting loaded from. The last time
  | + 6876 [james@gr yp ] Hmm, I just went to look into this a little with the aforementioned
  |   6877 [james@gr yp ] As usual, I was wrong.  The /vendor/rails/ is the same version.  My
  |   + 6878 [jim@we ri hh] Yes.  I believe it does it to autoload missing model classes.
  |   + 6879 [ara.t.howard] it does.  i haven't read where, however... somthing to do with dependancies
  |   | 6880 [jeremy@bi sw] Rails segregates 'framework' and 'app' files so the latter may be
  |   | 6881 [ara.t.howard] why not?  a combination of expand_path, mtime, and an md5 of some/all of the
  |   | + 6882 [james@gr yp ] James Edward Gray II
  |   | + 6883 [jeremy@bi sw] Perhaps Rails should implement these heuristics since Ruby has them in
  |   + 6884 [nobuyoshi.na] Similar workaround could work to wrap methods.
  + 6887 [ara.t.howard] this was indeed the issue florain.  thanks much!

Wilderness: Version 2.0 IDBG Debugger - Example eval.c usage
6866 [ruby-core@ha] ...

Wilderness: IDBG 2.0 Manual
6867 [ruby-core@ha] ...

question about superclass of singleton class
6885 [dblack@wo bl] I see there's a change from 1.8.2 to 1.8.3, involving the superclass
6886 [nobuyoshi.na] It has been changed in 1.9.

Iconv library - differences between Ruby 1.8.2 and 1.8.4?
6888 [dave@bu t. d] I'm new to ruby-core, so firstly, hello to all the awesome Ruby-bosses here.
6889 [nobuyoshi.na] 2 changes.
6890 [dave@bu t. d] Thanks. That probably means no incompatible changes.
6897 [nobuyoshi.na] And one more change after preview1.
6900 [dave@bu t. d] It seems to me that that solution could be used to potentially turn this

Time.utc! and Time.localtime!
6891 [hobe@gm il c] Writing a script yesterday I found out, much to my surprise, that the
+ 6892 [drbrain@se m] I consider it not quite a ! method because it only changes the
| 6893 [hobe@gm il c] I disagree that they are not really ! methods.  It changes the offset
| + 6895 [drbrain@se m] Typically a method that modifies the receiver has a non-! method that
| + 6898 [dblack@wo bl] The thing is, ! doesn't mean that the method changes the receiver
|   6899 [ara.t.howard] but surely
+ 6901 [matz@ru y- a] Hmm, usually I don't respond to mails saying "un-rubyish" or "POLS",
  + 6903 [kevin.jackso] Could you not keep the old methods for bwc and add new more 'ruby-ish'
  | 6904 [matz@ru y- a] Adding two new methods is OK.  Even removing methods is OK too.  But
  | + 6908 [ara.t.howard] class Time
  | | + 6921 [hobe@gm il c] Wouldn't it be okay to change it in 1.9 so that it will eventually be
  | | + 6931 [matz@ru y- a] I don't think this is good for transition, just because new code would
  | |   6932 [ara.t.howard] well, it seems either that or a suprising change must be in order... maybe not
  | + 6933 [florgro@gm i] But is breaking compatibility in 2.0 really this much of an issue?
  |   6935 [dblack@wo bl] ...
  |   6944 [wilsonb@gm i] What about #to_utc for the non-destructive version, and possibly a
  + 6923 [meta@po ox c] I think so too. I would expect atime.utc to return the time in UTC

[PATCH] Dir.tmpdir RDoc
6894 [drbrain@se m] ...
6896 [djberg96@ya ] Speaking of tmpdir, I'm curious why the tmpdir source
+ 6902 [matz@ru y- a] Mostly because it's easier to implement tmpdir in Ruby than in C.
+ 6907 [nobuyoshi.na] TEMP direcotry is defaulted to "%USERPROFILE%/Local
  6910 [zdennis@mk e] #include "windows.h"
  + 6913 [usa@ga ba ec] CSIDL_PERSONAL means "My Documents".
  | 6915 [zdennis@mk e] You are correct, that should be CSIDL_PROFILE.
  + 6917 [halostatue@g] It's actually SHGetSpecialFolderPathW, and the distinction will be made

Problem to embed Ruby in a threaded Application (Mingw)
6905 [frederic.coi] I have a problem with my application which embeds the Ruby engine.

[PATCH] Add Missing HTTP Headers and Status Codes to Ruby CGI
6906 [pabs@pa lo r] Attached is a patch against Ruby 1.8.4 which adds missing HTTP headers,
6986 [pabs@pa lo r] Matz, anyone?
6991 [matz@ru y- a] It will be applied to 1.9 soon (not 1.8.4).
6992 [pabs@pa lo r] CGI"

refurbishing ipaddr.rb - Class IPAddr?
6909 [hadmut@da is] I was working a little bit with the IPAddr class defined in
6977 [hadmut@da is] nobody maintaining ipaddr.rb at the moment?

IO.open not calling close in block form?
6911 [Daniel.Berge] What happened to the block form of IO.open after 1.8.2?  It's supposed to

[PATCH] Re: IO.open not calling close in block form?
6912 [ mfp@ac .o g] FWIW
6914 [nobuyoshi.na] It won't work fine if closed? method is overridden.
6916 [ mfp@ac .o g] $ ./ruby -ve 'open("/dev/null") {|f| def f.close; $stdout.puts "closed"; end; def f.closed?; false end; raise ArgumentError }'
6972 [Daniel.Berge] Perhaps an explicit return within an ensure clause should raise a

change to yaml in 1.8.4
6918 [ara.t.howard] intentional?
+ 6919 [james@gr yp ] [ruby-talk:157302]
+ 6920 [jeremy@bi sw] I believe this was changed in 1.8.2 -> 1.8.3 and is what
  6922 [ara.t.howard] so 1.8.4 will break gems too?  this isn't good...
  6924 [tom@in oe he] That's what I'm hearing - folks are building Gems with 1.8.4 preview1
  6925 [drbrain@se m] I believe the gem index builder thingy will work with 1.8.3+ gems
  + 6926 [tom@in oe he] Hm, interesting!  Any idea about installing Gems built with 1.8.2 on
  | 6927 [drbrain@se m] I have had no problems since 1.8.3+ do not need the '---' document
  | 6928 [tom@in oe he] Cool.  Hm.  Maybe I'll upgrade RubyForge to 1.8.4p2 and see if that
  | + 6929 [djberg96@ya ] Perhaps it's time to consider XML over YAML.
  | + 6945 [jim@we ri hh] This only pushes the problems out to users who have not yet upgraded.  That
  |   + 6946 [tom@in oe he] Ah, good thinking... hm.  I'll hold off on that upgrade.  Once again,
  |   + 6948 [james@gr yp ] Good call.
  + 6955 [halostatue@g] It's a problem mostly on Windows because there are enough problems

WIlderness: What is meangin on 'noex = -1'
6930 [ruby-core@ha] While it seems clear the 'noex' 0-4 indicates the number of parameters

1.8.x, YAML, and release management
6934 [ryand-ruby@z] I'm concerned that 1.8.3's acceptance of non-backwards-compatible
+ 6936 [james@gr yp ] I agree 100%.  I too am disappointed with the change.
+ 6937 [caleb@ae -t ] I also agree.
| 6938 [shyouhei@ic ] 1.10.x never come because ruby's version string must be sortable using
| + 6940 [florgro@gm i] Let's just call it 1.A.x then? :)
| + 6941 [matju@ar en ] Call the next stable series 1.A.x
| + 6942 [caleb@ae -t ] "1.A.1" then perhaps?
| + 6952 [hgs@dm .a .u] Where is this constraint set?  Modifying the code that handles
|   6953 [shyouhei@ic ] In matz himself.  He announced that in, for instance, [ruby-dev:9366].
|   6954 [hgs@dm .a .u] I see. I wonder what would change his mind...
+ 6939 [shyouhei@ic ] Acceptable or not matz releases 1.8.4 this Saturday.  It's hard to
| 6957 [ryand-ruby@z] I understand your disagreement, but at the same time, I'm afraid this
+ 6943 [meta@po ox c] I agree. I was pretty shocked when I upgraded to 1.8.4 and Rake and
+ 6951 [ruby-core@wh] Sorry about the silence, it was not intentional.  My time online has
| + 6958 [ryand-ruby@z] I'd support this fully.
| | + 6959 [james@gr yp ] Absolutely.  Let's not ship another version of Ruby that people
| | + 6963 [ruby-core@wh] It's a change of emitter->headless = 0 rather than emitter->headless =
| + 6960 [matz@ru y- a] Hm, if it would be a benefit for many people, go ahead and apply the
|   6961 [kevin.jackso] I'd suggest another round of release candidates afterwards since
|   6962 [matz@ru y- a] I didn't mention but I am planning another preview on 21st (on JST).
+ 7034 [caleb@ae -t ] patch)
  7047 [ocean@m2 cc ] (2005/12/31 03:57)
  7048 [ocean@m2 cc ] Sorry, patch was broken. Try again...
  + 7049 [ocean@m2 cc ] Umm, my patch doesn't work with BigDecimal. probably more work is needed...
  | 7053 [ocean@m2 cc ] This is because BigDecimal is written with C. So I think this is another story.
  | 7054 [meta@po ox c] Ideally YAML would have the equivalent of both _dump & _load from
  + 7089 [ocean@m2 cc ] I have committed these patches.

Problem building 1.8.4p2 on Windows
6947 [djberg96@ya ] Windows XP Pro
6949 [nobuyoshi.na] See [ruby-core:06801].
6950 [djberg96@ya ] Whoops, sorry about the noise.  Thanks.

Re: [OT-version comparisons] 1.8.x, YAML, and release management
6956 [ryand-ruby@z] This is waaay tangental to the point I'd like to keep on topic.

Array Documentation Issues
6964 [james@gr yp ] $ ri -T Array.indices
+ 6965 [ mfp@ac .o g] This is due to
+ 6967 [matz@ru y- a] This is a bug in a document.  I will fix it soon.
  + 6968 [drbrain@se m] array.initialize_copy(other_array)  -> array
  + 6969 [james@gr yp ] Thanks much.
  + 6970 [ryand-ruby@z] RDoc is stumping you too??? YAY! Now I don't feel so bad. :P