13491-13985 subjects 13904-14484

^ Still exists: Problems with \M-x in utf-8 encoded strings
13718 [ed.odanow wo] Unfortunately not completely, but let me show details and
13721 [david davidf] The \M escape is a weird one, but it is really uncommon.  I'd be more
+ 13724 [ed.odanow wo] I don't know if it is really used somewhere.
+ 13729 [duerst it.ao] One thing to consider is backwards compatibility. In 1.8, the only

^ Encoding Confusion...
13722 [ed.odanow wo] This is a multi-part message in MIME format.
13725 [david davidf] \C-\M-a is an 7-bit clean ASCII character.  The encoding of that string
13727 [ed.odanow wo] Unfortunately not...
13730 [matz ruby-la] We consider it a bug.  We try to fix it soon.

^ Re: Iconv#iconv returning wrong object (ri broken?)
13723 [sroberts uni] Are you sure you understand how iconv works, particularly with

^ YAML buffer overrun
13735 [akr fsij.org] core dump as follows.
13794 [why ruby-lan] Oh, thankyou, akr.  I'll take care of this.

^ What is nil_plus?
13737 [Daniel.Berge] #ifdef NIL_PLUS
13738 [matz ruby-la] It's something like appendix.  We can now discard it.

^ 1.8.6: gc.c and __builtin_frame_pointer
13739 [lists-ruby s] I'm playing with porting Ruby to Stratus VOS, which is an unusual but
13740 [matz ruby-la] For the workaround, you can try specifying --disable-frame-address to

^ retry semantics changed
13741 [dave pragpro] def do_until(cond)
13742 [binary42 gma] I hope so... see ruby-core[13259] for a thread on this. I originally
13743 [dave pragpro] Removing it because of the potential side effects seems somewhat
+ 13744 [binary42 gma] You probably right but I am betting the biggest reason it was removed
| 13745 [binary42 gma] Quick note... that proxy call should forward a block as well (obvious
+ 13746 [matz ruby-la] That's what Evan Phoenix said.  But can you elaborate?
  13747 [dave pragpro] I'm thinking that allowing retry in a block still has the effect of
  13748 [matz ruby-la] I understand, but does side effect issue matter?  The primary reason
  + 13749 [dave pragpro] OK, I didn't understand the performance issues driving the decision.
  | + 13751 [charles.nutt] I find retry-outside-rescue to be extremely dangerous. See the example I
  | | + 13753 [charles.nutt] Sorry, this needs a block for retry to work (which is pretty goofy in
  | | | 13760 [matz ruby-la] But how about retry being inside of the block?
  | | | 13768 [charles.nutt] In the simple case, it's ok as far as I can tell, since it's being
  | | | 13770 [matz ruby-la] Hmm, it's implicit and weird, I admit.  But is it dangerous?
  | | | 13772 [dave pragpro] But is there a benefit to having it there at all?
  | | | 13774 [matz ruby-la] No, that's not my point.  I wanted to know if it's danger or not.
  | | | 13775 [dave pragpro] I guess I'm saying that perhaps the danger discussion is academic. If
  | | | 13776 [matz ruby-la] Perhaps.  I just wanted to know what he meant in the word "dangerous".
  | | | 13777 [charles.nutt] By dangerous I mean "undetectable and unexpected side effects". The
  | | | + 13778 [charles.nutt] - Charlie
  | | | + 13789 [ed.odanow wo] I was plying a little bit with both usages of "retry" (block and
  | | + 13754 [ed.odanow wo] If the wish is a way to evaluate lazy parameters, but not to evaluate the
  | |   13755 [charles.nutt] Quite right, and in this case the caller is explicitly accepting that
  | + 13758 [dave pragpro] But.. having said that...
  | | 13762 [mneumann nte] No! And those few pieces of code (I only ever used it once in one place
  | | 13765 [ed.odanow wo] I didn't recognize the possibility to use 'retry' in a method body when an block
  | | + 13769 [charles.nutt] The reason this happens is because retry triggers both the receiver and
  | | | 13771 [binary42 gma] T24gTm92IDI0LCAyMDA3IDg6MTMgQU0sIENoYXJsZXMgT2xpdmVyIE51dHRlciA8Y2hhcmxlcy5u
  | | | 13773 [ed.odanow wo] I would like to leave it open for discussion for a while. When I said, that my
  | | + 13779 [ed.odanow wo] I want to inform you about two interesting reactions
  | + 13761 [matz ruby-la] I don't want to irritate you, but how this change make retry closer in
  + 13759 [akr fsij.org] Remember [ruby-dev:31425], [ruby-dev:31426], [ruby-dev:31433] and
  + 13763 [evan falling] Chiming in again on this...
    + 13766 [akr fsij.org] Does rubinius evaluates arguments in right-to-left order?
    | 13798 [evan falling] Yes, it does. Almost every language says that the order that the
    | + 13799 [matz ruby-la] Unfortunately, I expect the evaluation order to be left-to-right.
    | | 13810 [evan falling] Then I shall put it on my todo list to change the order in rubinius.
    | + 13800 [charles.nutt] But it does change execution behavior...
    | | 13801 [dave pragpro] It also plays into calculated argument evaluation
    | | 13809 [evan falling] Very true, but rubinius handles default argument code in left to right
    | + 13804 [akr fsij.org] Several exceptions: Ruby, Python, Java.
    |   13808 [evan falling] Well, in my own defense, I have no clue if it's specified in ruby or a
    |   13811 [jeremy asynk] As far as I know File.exists? has only been deprecated and replaced by
    |   13812 [charles.nutt] ~/NetBeansProjects/jruby $ ../ruby1.9/ruby -e "File.exists? 'build.xml'"
    |   13813 [Daniel.Berge] For crying out loud, of all the methods to get rid of. Just make it an =
    + 13795 [ara.t.howard] this sounds reasonable.  i actually do a lot of networking code and

^ Weightless Threads?
13750 [djberg96 gma] Also known as "microthreads".
+ 13752 [charles.nutt] I've done experiments with microthreading in JRuby, as part of exploring
+ 13756 [garbagecat10] The article you quoted is five years old, but the basic ideas have
+ 13757 [garbagecat10] Ah, sorry, Dan, I just noticed this was addressed to Ruby-core rather

^ Access modifier : private
13764 [dharmarth gm] I'm missing some core ruby concept here that I just can't figure out;
13767 [murphy rubyc] access modifiers are only for methods; constants and variables are

^ threads+sockets probs
13780 [rogerpack200] linux to a lesser extent) when you mix threads + sockets sometimes
14968 [rogerpack200] It appears that whenever you have a lot of open sockets, then you open
15017 [pbrannan atd] Do you have any short examples that can reproduce the problem?

^ C-Core-Questions
13781 [saladin.mund] This is a multi-part message in MIME format.
+ 13783 [ed.odanow wo] The "Ruby C Language API" is described in details in chapter 21 of the PickAxe.
| 13784 [saladin.mund] Yes, I know that some information about C is written in the PickAxe.
| 13785 [decoux moulo] look at the file variable.c
| 13786 [saladin.mund] I think I have explained it too confusing
+ 13791 [ryand-ruby z] On Nov 26, 2007, at 02:22 , <saladin.mundi@gmx.de>

^ gem_prelude.rb raises exception
13782 [mame tsg.ne.] I cannot invoke ruby-thunk because of gem_prelude.rb.
13793 [rich infoeth] I indeed added that check to skip unless dash and also wrapped the
13826 [mame tsg.ne.] Thank you.

^ Syntax error when using comment between two lines in new method chain syntax
13787 [ed.odanow wo] It dosn't give any problems to write comments between two
13788 [ed.odanow wo] [[1], [2], [3]]
13790 [murphy rubyc] I'm not sure if this is even supposed to work...but I agree that

^ Anyone tried -r debug on OSX?
13792 [dave pragpro] It hangs for me here. I have to kill -9 to stop it.
+ 13828 [murphy rubyc] same here. the process does nothing. But ^C was enough for me.
| 13831 [murphy rubyc] I'm stupid. skip that. it works fine with 1.8.6p111 on Mac OS X 10.5.
| 13832 [dave pragpro] ruby 1.9.0 (2007-11-27 patchlevel 0) [i686-darwin9.1.0]
| + 13836 [rick.denatal] 1.8 Installed with macports on Tiger hangs if you don't give it
| + 13837 [ryand-ruby z] % ./miniruby ./runruby.rb --extout=.ext -- -rdebug -e 0
+ 13829 [ryand-ruby z] export DISPLAY=/tmp/launch-qEVszF/:0
+ 13830 [laurent.sans] Which version of OSX and Ruby do you run?

^ autoload reminder
13796 [transfire gm] Just want to remind core team that #autoload does not use the normal
+ 13797 [fgeller gmai] ...
+ 13827 [transfire gm] Is this just considered a non-issue?

^ Hash#eql? and Hash#== in trunk
13803 [gwtmp01 mac.] {1 => 2} == {1 => 2.0}          # true
13823 [matz ruby-la] It should.  It's a bug.  Thank you for the report.

^ Socket.gethostbyname and Reverse Lookups: A Strange and Terrible Saga
13805 [billk cts.co] (with apologies to Hunter Thompson ;)
13806 [gwtmp01 mac.] gethostbyname, gethostbyaddr, getaddrinfo, and getnameinfo
13807 [billk cts.co] It does set AI_CANONNAME.  Basically my code was derived directly
14193 [rogerpack200] The thing that got me was ipv6 versus ipv4.

^ -v affects parse tree, causes tests to fail in ParseTree
13814 [ryand-ruby z] The problem is that the void_expr macro is used in void_stmts and
+ 13815 [ryand-ruby z] % svn diff parse.y
+ 13816 [charles.nutt] I'm not really clear why any of these command-line parameter modify the
| 13817 [ryand-ruby z] just looks to me like an unintended side effect from the macro.
+ 13822 [nobu ruby-la] Thank you, now it is always removed regardless of -v.

^ Dear Matz,
13818 [steven lumos] I just stumbled on this really great quote[1] and was inspired to
+ 13819 [aktxyz gmail] <!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
+ 13820 [charles.nutt] I would say "practicalists" rather than "minimalists". A language that

^ Inhering from Proc in 1.9?
13824 [flori nixe.p] I get the following behaviour from the current trunk version ruby 1.9.0
13825 [ko1 atdot.ne] class Foo

^ Bug in Signal Handling
13833 [james graypr] I'm pretty sure we have found a bug in Ruby while working on TextMate.
13841 [vjoel path.b] 1. system("")  has the same effect as   %x{true}
13929 [james graypr] Any thoughts on this from the core team?

^ warn() vs. puts()
13834 [james graypr] Why isn't warn() as smart as puts()?
13838 [ryand-ruby z] I gotta say... I really dislike puts' "smarts". Let me handle the

^ "initialize" method in 1.9 Thread gone?
13835 [btoj frisurf] I just experimented with some thread code in Ruby 1.9 HEAD and discovere=
13839 [flori nixe.p] This seems to be the same problem I reported in ruby-core:13824. ko1
13840 [nobu ruby-la] Accurately, they are different.

^ Better introspection for Frames, Thread, and enhancing binding.
13842 [rocky.bernst] The below is a little long. So here's a summary.
+ 13845 [ryand-ruby z] You might want to look at rubinius.
+ 13847 [charles.nutt] Be careful what you ask for.
  + 13848 [sylvain.joye] That would be the equivalent of the -O0 flag on gcc-compiled languages.
  + 13849 [rocky.bernst] I don't think we necessarily disagree. But a couple of comments and
  + 13850 [ryand-ruby z] That's simply not true. Look at self, smalltalk, and many lisps for
    13857 [charles.nutt] Perhaps you could provide links to a concrete example showing those
    13858 [ryand-ruby z] I'm pretty sure you can google with the best of them.
    13860 [charles.nutt] The point is, if user code can expect to grab a frame object at any
    13869 [rocky.bernst] On Dec 3, 2007 1:49 AM, Charles Oliver Nutter <charles.nutter@sun.com>
    13872 [charles.nutt] It is, because many implementations may choose to manage frame

^ Regexp regression
13843 [evan cloudbu] => nil
+ 13844 [charles.nutt] Marcin fixed this in Joni already, so at least JRuby does not have
+ 13883 [mame tsg.ne.] This is working, though M17N-related bug might still remain.
  13884 [mame tsg.ne.] Sorry it had a small mistake.  Here is a fixed version.

^ Rant: Ruby's development process ...
13846 [sylvain.joye] A while ago, I posted bugreports and patches about a bug in the dl/gc
13853 [shyouhei rub] I know your frustration... A show-stopper here is that we are waiting DL

^ Array#flatten works quadratic time on length of resulting array. It could be linear
13851 [artem.vorozt] I encountered problem with Array#flatten slowness (it can be much
+ 13852 [artem.vorozt] Well, on ruby 1.9.0
+ 13859 [ryand-ruby z] With that not being recursive, how do you deal with sub-sub-arrays?
| + 13862 [neleai sezna] I guess quadratic behaviour is caused by appending array over and over
| + 13866 [artem.vorozt] Here the corrected version
+ 13863 [charles.nutt] I get similar results for JRuby (JRuby results are repeated to allow for
  13867 [artem.vorozt] I don't now. Only two assumptions
  13868 [artem.vorozt] class Array
  13870 [mame tsg.ne.] It seems to not be able to handle a non-recursive array that
  + 13871 [artem.vorozt] OK, i see.
  + 13876 [nobu ruby-la] I've tried same approach last evening too, and it was very
    13877 [mame tsg.ne.] You're right.  I have referred to the implementations of diff and
    13878 [nobu ruby-la] Using st_table, you don't need to call rb_obj_id().  It's OK to
    + 13879 [mame tsg.ne.] Hmmm, I see.
    | 13882 [nobu ruby-la] Matz and Shugo.
    + 13888 [matz ruby-la] Can you commit?
      13898 [artem.vorozt] Method Array#join has the same drawback.
      13900 [nobu ruby-la] HTH.
      13902 [matz ruby-la] Go ahead and commit, please.

^ A little concern about m17n
13854 [dan-ml dan42] I know it's kinda late to make any changes to the 1.9 roadmap, so I'm
13855 [halostatue g] No. #index always returns character position; it just so happens that
13856 [dan-ml dan42] That's just a different way of wording my concern :-P
13861 [charles.nutt] But that would be a bug in your code; if you don't encode a string, you
13864 [dan-ml dan42] Well, yes, that would be a bug in my code. Just as if I did str.size (in

^ Continuatations w/multiple Threads cause segfaults in 1.8.6 & 1.6.8
13865 [brent mbari.] This is a multi-part message in MIME format.

^ Kernel.Integer - possible error
13873 [ed.odanow wo] Good evening all together!
13875 [eric sempere] I have suspected issues in Ruby 1.8.6 after seeing calls to a controller me=
13874 [eric sempere] To clarify, I'm talking about in Rails.

^ Idea: always yield current thread to Thread.new
13885 [Daniel.Berge] Thread.new{ Thread.current['foo'] =3D 'bar' }
+ 13886 [brent mbari.] Dan,
+ 13979 [matz ruby-la] This is interesting, and mostly compatible.  Any opinion from others?
  + 13980 [mental rydia] What about remembering a thread's parent (which is what's really meant
  | 13982 [mental rydia] Agh, I'm sorry.  No, it's not.
  + 13981 [dave pragpro] I'm not a big fan of magic arguments.
    13983 [Daniel.Berge] Hm, I suppose being explicit here is probably for the best then, as I

^ Can we foresee any big changes for Ruby in year 2008?
13887 [songmash gma] I understand this question is a little bit "earlier" before Ruby 1.9.x debut
13894 [rick.denatal] I don't know that it exists as Ruby is pretty much designed iteratively.
13901 [songmash gma] Thank you very much Rick.
13963 [lists-ruby s] Those look really cool.  For the ADD among us, does anyone happen to have
13964 [murphy rubyc] great talk matz, thank you! Ruby 1.9 has been on my chrismas wish list

^ [BUG] Continuations on dead threads crash 1.8.6 and 1.6.8
13889 [brent mbari.] This script crashes my builds of Ruby versions 1.8.6 and 1.6.8

^ Re: [Spam] [BUG] Continuations on dead threads crash 1.8.6 and	1.6.8
13891 [ggarra advan] Ruby1.9 does not (currently) support continuations.
+ 13892 [james graypr] People are saying this almost daily now, but that still does make it =20
+ 13895 [brent mbari.] I fairly certain recent builds of 1.9 will run this script
  13896 [hramrach cen] well, it does crash for me quite fast :)

^ How to have fine grain control over GL in C extensions
13893 [gethemant gm] I am writing a smallish C extension for enabling parallel execution of
13909 [gethemant gm] #include <ruby.h>
13910 [ko1 atdot.ne] OK.  I'll add a C API such as rb_thread_interrupted(void).  But you
+ 13911 [gethemant gm] But lots of C extension use CHECK_INTS i suppose? Not needed in my
| 13912 [gethemant gm] because on my dual core machine..i saw printf's from second
+ 13947 [gethemant gm] void check_for_interrupts() {

^ issue with Date class freeze
13897 [gilltots gma] Hopefully this is the right list..if not please tell me.
13899 [matz ruby-la] Date objects are modifies itself to cache calculated result.  Since
+ 13918 [murphy rubyc] I would still consider it a bug.
| + 13919 [ed.odanow wo] I think that under this circumstances the error should occur when
| + 13923 [yemi weldfas] On 1.8.2, I only see one value as an array, the other as a Fixnum.
+ 13927 [gwtmp01 mac.] How about making Date#freeze compute the cached to_s string and

^ Clarification of retry change
13903 [charles.nutt] Matz confirmed that retry-outside-rescue will no longer work, but I
13905 [ko1 atdot.ne] Should I remove from retry feature from YARV?
+ 13906 [charles.nutt] I would think they should be "rescue retries" in both cases, and bubble
| 13925 [binary42 gma] In these cases I could see it being confusing if retry didn't work in
| 13930 [charles.nutt] You make a number of great points. I don't like retry in a block in a
| + 13931 [hramrach cen] When talking about the places where retry is allowed .. what does this
| | + 13932 [binary42 gma] I'm not the authority in this case, but those should be working fine
| | + 13938 [charles.nutt] The scary thing is that what you're saying sounds totally sane, but
| |   13940 [gwtmp01 mac.] You lost me there.  retry is certainly legal and unambiguous
| |   13943 [charles.nutt] Michal didn't say the if/case/while would be inside a rescue, so I
| |   13944 [hramrach cen] No, I *was* talking about placing the branching constructs inside retry.
| + 13933 [gwtmp01 mac.] If you get rid of the iterator-retry semantics then it seems like the
|   13937 [charles.nutt] I won't try to make an implementation case, though there are complications.
+ 13907 [matz ruby-la] Yes, please.