9500-10206

9208-9739 subjects 9725-9948

^ [ ruby-Bugs-6583 ] self-referential constants
9500 [noreply ruby] Bugs item #6583, was opened at 2006-11-09 06:43
9505 [nobu ruby-la] A bug in the code which makes such references.

^ What does rb_memsearch do?
9512 [murphy rubyc] I've been wondering what the function rb_memsearch does. It is
9513 [matz ruby-la] It is multibyte aware string search using Karp Rabin algorithm.

^ External entropy pool for random number generator
9515 [k.shutemov g] In the attachment patch which allow to use external entropy pool for
+ 9522 [nobu ruby-la] It uses /dev/urandom already.
| 9525 [k.shutemov g] It uses only for initial random seed and when you call srand(0);
| 9526 [matz ruby-la] I am not sure I understand.  By your patch, if you specify something
| 9527 [k.shutemov g] It allow you specify entropy pool once(in platform specific way) and
| 9528 [matz ruby-la] Can you elaborate how much it is useful in general enough to make it
| + 9529 [k.shutemov g] To make rand() not deterministic. The other way is modification
| | + 9531 [nobu ruby-la] I don't think it enough general.  Couldn't OpenSSL::Random be
| | | 9532 [k.shutemov g] Yes, it's an alternative, but not every platform have openssl. :(
| | + 9533 [matz ruby-la] I still don't get it.  If you want real non-deterministic random
| |   9534 [k.shutemov g] If I rewrite genrand_int32() to use external pool(if specifyed) it
| + 9551 [gus progress] /dev/random and /dev/urandom are interfaces to a kernel-resident random
+ 9535 [khaines enig] This feels like something that is best left as an extension instead of a

^ Different return values for setter methods
9540 [marcel verni] => nil
9541 [ mfp acm.org] It seems deliberate (as you say, there's no gain in making foo.send(:bar=, 1)
9542 [marcel verni] Thanks for taking the time to point it out. I suspected it might be that but

^ adding rb_warn gives a segmentation fault
9543 [fxn hashref.] I want to add some traces to rb_define_class and put a few rb_warn()
9544 [nobu ruby-la] It is before the initialization in rb_define_class().  It can't
9545 [fxn hashref.] Thank you, I had to be missing something. Which initialization do you

^ [PATCH] Net::FTP should check the control connection on EPIPE
9547 [simon.willia] This patch applies to the Net::FTP class that comes with Ruby 1.8.5.
10196 [shugo ruby-l] I merged your patch to SVN trunk.
10204 [simon.willia] No problem.  Thanks for applying it.  However, it appears that there
10205 [shugo ruby-l] Oh, sorry. I applied your patch.  Thank you.
10206 [simon.willia] Thanks.

^ 64 bit pack/unpack issues
9549 [dave frascon] I've found a bug in the 64 bit implementation of ruby -- I'm kind of new
9550 [dave frascon] It's fixed in 1.8.5

^ [ANN] sandbox 0.3
9552 [why ruby-lan] The 0.3 version of the sandboxing extension is out.  This extension allows you

^ Ruby 1.[89].\d+ and beyond.
9554 [hgs dmu.ac.u] I've been thinking about how version numbers are restricting what we can do.
+ 9555 [znmeb cesmai] For some reason, I've never seen "Spinal Tap". I suppose that's going to
+ 9556 [tjw omnigrou] Version numbers are often (even typically?) treated as a series of
| 9557 [hgs dmu.ac.u] which is a string.  Thus string comparisons are done in most cases where
| 9560 [tjw omnigrou] class Version
+ 9561 [drbrain segm] So what if we run out of numbers?
| 9563 [hgs dmu.ac.u] The only problem I have with that is that .a and .b look like alpha
| 9564 [drbrain segm] Letters in versions is not without precedent.
| + 9565 [hgs dmu.ac.u] True, there is /1.1[abcd]/ in the ruby ftp site.
| | 9567 [drbrain segm] Oh, ok.
| + 9571 [robert.dober] <SNIP>
|   + 9573 [james graypr] I'm pretty sure all of this is a non-issue.  Hasn't Matz said we wont
|   | 9576 [drbrain segm] What if we need to exceed 1.8.9?
|   | 9578 [murphy rubyc] I guess the XP people would say: Let's see if we really run into that
|   | 9580 [drbrain segm] Not backwards compatible.
|   | 9585 [hgs dmu.ac.u] That's why I was proposing that the feature be backported to 1.6 etc.
|   | 9587 [djberg96 gma] So, let me get this straight people.  We're avoiding 1.8.10 because of
|   | + 9588 [j.vimal gmai] What would be wrong with
|   | | 9589 [jan.svitok g] Don Knuth managed around similar problem with TeX versions converging
|   | | + 9595 [robert.dober] "Sorry I cannot release the next version of TeX because I do not know the
|   | | + 9596 [charles.nutt] Hmm...I seem to remember a numerical constant that started with 1.85,
|   | |   9597 [j.vimal gmai] But still it would come before 1.8.9 :)
|   | + 9593 [hgs dmu.ac.u] Earlier versions of ruby that rely on this behaviour.  But now see
|   + 9575 [drbrain segm] I can't parse anything you've just written.
|   | 9579 [robert.dober] Sorry 'bout that  but  I only made a minor point of confusion between  "2.a"
|   + 9586 [ceo hawthorn] What?  Microsoft only works with Base 17!
+ 9590 [shyouhei ice] Please.  Don't forget RUBY_VERSION_CODE.  Browse the Google Code Search
  9594 [hgs dmu.ac.u] Ouch.  The only solution to that that I can think of would be
  9601 [nobu ruby-la] It doesn't work in preprocessor.  Expression in C is not matter.
  9605 [hgs dmu.ac.u] Grrrrrrr! I'll have to admit defeat on this, then.  This probably

^ optimized mutex progress
9566 [mental rydia] Just a quick update:  I've released a 'fastthread' library which

^ YAML bug.
9568 [escii hotmai] This is a multi-part message in MIME format.

^ [ ruby-Bugs-6764 ] tempfile.rb should let you preserve a file
9569 [noreply ruby] Bugs item #6764, was opened at 2006-11-18 19:09
9598 [ben bleythin] This would be handy to me, and if there's interest I'd like to attempt

^ Patch for HalfOrder, help please!
9570 [robert.dober] I do not think that this shall make it into the core, maybe into the
9619 [nobu ruby-la] lib/tsort.rb may help you?
9620 [robert.dober] <cut>

^ io_write (io.c) bug (and its fix) under MS Windows for GUI apps (rubyw)
9572 [idrassi gmai] I recently encountered a problem under MS Windows where a GUI
+ 9574 [dominic.coon] Is it good style to #ifdef the if statement guard?
| 9577 [idrassi gmai] For the #ifdef of the if statement, it was the simplest way to achieve
+ 9640 [nobu ruby-la] It doesn't feel appropriate to modify io.c.
  9671 [matz ruby-la] Could you commit this to 1.9?

^ type information
9581 [denigeorge c] how does the interpreter (ruby1.8.5) store the type information of a
+ 9582 [murphy rubyc] what do you mean? obj.class?
+ 9583 [j.vimal gmai] You seem to be from IITB! Great! I am from IIT Madras.
| 9584 [murphy rubyc] Sorry to be the smartass ^^ but because Ruby is such a great language,
+ 9602 [nobu ruby-la] In Ruby, variables have no type information at all.
  9606 [denigeorge c] type information
  9607 [michael.seli] This is only for LOCAL variables.
  + 9608 [denigeorge c] thanx Michael
  | + 9609 [dblack wobbl] s/type/class/g :-)
  | | 9610 [murphy rubyc] I heard the Arachno Ruby IDE tries to do something like this. never
  | + 9633 [drbrain segm] You may want to look at ParseTree then.  It will make your life much
  + 9614 [denigeorge c] what for the global variables ??
    9615 [michael.seli] Have a look at rb_eval() in eval.c

^ [ ruby-Bugs-6788 ] In a c++ program, when I call "fclose()", or "ofstream::close()" in a file that includes ruby.h, the program crashes
9591 [noreply ruby] Bugs item #6788, was opened at 2006-11-20 16:35
9592 [jan.svitok g] I've had a similar problem recently, when compiling a C++ extension

^ [ ruby-Bugs-6802 ] net/ftp hangs
9599 [noreply ruby] Bugs item #6802, was opened at 2006-11-21 12:31
9600 [drbrain segm] On Nov 21, 2006, at 3:31 AM, <noreply@rubyforge.org>

^ [ ruby-Bugs-6819 ] YAML loses data
9603 [noreply ruby] Bugs item #6819, was opened at 2006-11-22 17:19

^ [ ruby-Bugs-6820 ] #ancestors never includes the singleton class (inconsistent)
9604 [noreply ruby] Bugs item #6820, was opened at 2006-11-22 08:49
9618 [matz ruby-la] It is supposed to.  Singleton classes (or eigenclasses, if you want to
9629 [sylvain.joye] And, out of curiosity, why do you want that ? The singleton classes look
+ 9630 [binary42 gma] I can't speak for matz, but I have heard a number of reasons this is
| + 9631 [dblack wobbl] I've only ever heard the wonderful tool opinion :-)  Mostly I think it
| + 9638 [sylvain.joye] this is not a problem since my problem is with #ancestors being called *on*
|   9639 [dblack wobbl] I tend to agree.  Anonymous classes are visible in ancestors, and
|   9745 [matz ruby-la] I just remembered I made decision that singleton classes (or
|   + 9746 [calamitates ] Then would you consider making the following changes as well?
|   + 9747 [wilsonb gmai] Is this an implicit vote for the 'eigenclass' name, or just a useful
|   + 9748 [pit capitain] Yeah, finally! Thanks Matz for this Christmas present.
+ 9632 [ceo hawthorn] So dit I (my_ancestors) for various purposes ;-)  cet

^ Wilderness: How are Class actually built
9611 [ceo hawthorn] I have been trying to fully understand all of the steps in defining
9612 [murphy rubyc] [murphy]
9613 [ceo hawthorn] Thanks!

^ [ ruby-Bugs-6866 ] Create Dates from Times
9616 [noreply ruby] Bugs item #6866, was opened at 2006-11-24 23:43
9617 [murphy rubyc] maybe Time#to_date would be more intuitive.

^ Typo in string.c
9621 [murphy rubyc] rb_raise(rb_eTypeError,
9622 [matz ruby-la] Thank you!

^ Truncating string set on $0
9624 [joaopedrosa ] I have a "pre-ruby-launcher", which loads some environment
+ 9625 [eegreg gmail] Using a launcher will also break RDoc::Usage() in usage.rb because it
+ 9626 [nagai ai.kyu] I got same problem on Ruby/TkORCA.
  9627 [joaopedrosa ] Thanks for the idea and code. I might use it as well.

^ Sync/Sync_m bug?
9628 [mental rydia] I've been working on an optimized version of Sync_m from the stdlib, and

^ (assunto em branco)
9634 [malp_pontes ] Unsubscribe

^ Formatting patch to benchmark.rb
9635 [gavin refine] The Problem

^ [ ruby-Bugs-6933 ] Problem with Benchmark::Tms#add! - expects &blk but takes no parameters
9636 [noreply ruby] Bugs item #6933, was opened at 2006-11-28 09:50

^ [ ruby-Bugs-6934 ] Benchmark::Tms#add! missing block argument
9637 [noreply ruby] Bugs item #6934, was opened at 2006-11-28 04:52

^ [ ruby-Bugs-6987 ] Double-free problem in the interpreter
9641 [noreply ruby] Bugs item #6987, was opened at 2006-11-30 17:19
9642 [matz ruby-la] Thank you for the report.  Although this indicates double free
9643 [sylvain.joye] Well, I can show you the *application* which does that. The problem is that

^ [ ruby-Bugs-7001 ] Determinant Problem in Matrix.rb
9644 [noreply ruby] Bugs item #7001, was opened at 2006-12-01 03:50
9645 [shyouhei ice] This is 100% intended.  You have to learn integer arithmetics when you
9646 [nex342 gmail] This is a multi-part message in MIME format.
+ 9647 [shyouhei ice] Sorry not defining determinants in a typical way.  It does divide.
+ 9649 [steven.jenki] Just out of curiosity, are you trying to actually calculate something
| 9650 [nex342 gmail] Actually, it first came up when I was checking my Linear Algebra
| 9652 [steven.jenki] Ah. You're probably OK, then; homework problems are usually
+ 9651 [nex342 gmail] Why is this? It seems very strange to have a class that doesn't work

^ Float numbers comparison
9654 [pcsoeiro gma] a=1.9
+ 9655 [wilsonb gmai] 0.1 is a repeating series in binary, like 1/3 in decimal. (0.3333333333~)
| + 9656 [zdennis mkte] java suffers from this too, i have seen this first hand, but for an
| | 9657 [murphy rubyc] The Test::Unit library provides a special method to deal with that
| + 9735 [pcsoeiro gma] Thank you all for your awnsers, i haven't read them all. But i don't see
|   9736 [meta pobox.c] A Ruby float is a 64 bit float using your system's native floating
|   9737 [pcsoeiro gma] Thank you very much.
+ 9673 [gwtmp01 mac.] This seems to be a common source of confusion for many programmers in

^ What's up with dates?
9658 [wyldeone gma] I have a question about why dates are like they are in ruby. In ruby,
9659 [drbrain segm] Time is a time_t wrapper, and is limited in range by your OS (as

^ Dir.exist?
9661 [murphy rubyc] Wouldn't it be nice to have
+ 9663 [nwiger scea.] I like it personally - why not consider filing a feature ticket with
+ 9667 [nobu ruby-la] Also Socket.exist?, Pipe.exist?, Symlink.exist? and so on?
  + 9668 [murphy rubyc] Yes! If you work with those. Never used them before, actually.
  + 9669 [murphy rubyc] Yeah, you got me with that. There are no Pipe or Symlink classes.
    9674 [nobu ruby-la] I've heard that Windows Vista will have symlinks.  What do you
    9699 [murphy rubyc] thanks!

^ [ ruby-Bugs-7036 ] Cannot parse one-way WSDL - binding style="document", no wsdl:output tag
9662 [noreply ruby] Bugs item #7036, was opened at 2006-12-04 23:25

^ [ ruby-Bugs-7045 ] Bigdecimal isn't comparable
9664 [noreply ruby] Bugs item #7045, was opened at 2006-12-04 19:20
9665 [abergeron gm] The class itself is not Comparable, this is expected.  But istances of
9709 [neleai sezna] No its unexpected.
9710 [matz ruby-la] Yes, BigDecimal is Numeric.  And
9711 [neleai sezna] Problem was at code
9712 [matz ruby-la] class BigDecimal < Numeric

^ CSV Issue
9670 [james graypr] Greg Brown located an issue with the CSV library last night.  I won't
9672 [nakahiro sar] I don't think it's a lie.  It's the == of Array.

^ [ ruby-Bugs-7150 ] Exceptions raised during 'at_exit' hooks produce inaccurate messages
9675 [noreply ruby] Bugs item #7150, was opened at 2006-12-07 11:42
9678 [nobu ruby-la] Indeed.
9679 [matz ruby-la] Can you commit please?

^ [ ruby-Bugs-7153 ] WeakRefs not "weak" on FreeBSD/Mac OS X
9676 [noreply ruby] Bugs item #7153, was opened at 2006-12-07 13:51
9677 [drbrain segm] On Dec 7, 2006, at 10:52 , <noreply@rubyforge.org>

^ ri cannot display documents about the attributes of classes
9680 [usa garbagec] I found that ri cannot display documents about the attributes
9681 [drbrain segm] How much slower is your patch?  I supposed loading all the extra
9682 [usa garbagec] Yes, you are right.

^ rdoc aborting?
9683 [hadmut danis] I just had a strange behavior of rdoc. I tried to run rdoc on a ruby
9684 [hgs dmu.ac.u] That looks exactly like the beginning of a "here document" that doesn't

^ Segmentation fault - bignorm()
9685 [ico pruts.nl] (If this is not the right place to ask this question, please redirect me
+ 9686 [matz ruby-la] This is a direct reason for segmentation fault.  Both len and ds have
| 9689 [shiba mail2.] I may make void noise, but I still doubt; is it the duplicate of
| 9690 [ico pruts.nl] I can now reproduce this behaviour on ruby 1.8.5 (2006-08-25)
| 9691 [matz ruby-la] The fix was done on 2006-11-24.  If it's possible, try newer snapshot.
| 9692 [ico pruts.nl] I will, thank you. Is it easy to explain why/when this problem is
| 9694 [matz ruby-la] If it's the case, you can work around this by avoiding floats in your
+ 9687 [Daniel.Berge] What version of Ruby was this with?
  9688 [ico pruts.nl] Ruby 1.8.something - I am very sorry, but I do not know the exact

^ [ ruby-Bugs-7191 ] Error msg: /tmp/mysql.sock not found
9693 [noreply ruby] Bugs item #7191, was opened at 2006-12-09 21:38

^ [ ruby-Bugs-7202 ] String#[]= decimal bug
9695 [noreply ruby] Bugs item #7202, was opened at 2006-12-09 23:21
9760 [ming.lei ora] See the source code of the rb_str_aset function in string.c

^ using OptionParser.on_tail leads to missed AmbiguousOption
9696 [ruby meyerin] From a quick look at ruby1.9's optparse.rb, it seems options registered
9698 [nobu ruby-la] Intended behavior.

^ [ ruby-Bugs-7205 ] ERb throws SystemStackError on Solaris 10
9697 [noreply ruby] Bugs item #7205, was opened at 2006-12-10 19:35

^ [ ruby-Bugs-7218 ] OpenSSL library not updated with new PKCS #12 features
9700 [noreply ruby] Bugs item #7218, was opened at 2006-12-11 17:55

^ [PATCH] File.syscopy: close from if open on to fails
9701 [matled gmx.n] File.syscopy did not close from if open on to failed.  The garbage
9705 [nobu ruby-la] ftools.rb is obsolete.

^ another unpatched CGI vulnerability from 2004?
9702 [murphy rubyc] has this been fixed yet? I'm not sure...
9703 [matz ruby-la] It's fixed way back in 1.8.1.
9704 [murphy rubyc] Ah, there. I actually checked the Changelog, but only after that

^ [ ruby-Bugs-7224 ] net/ftp login appears to be broken
9706 [noreply ruby] Bugs item #7224, was opened at 2006-12-12 13:34
9707 [djberg96 gma] As I commented on RubyForge, FTP.open takes a block, FTP.new does not.

^ [ ruby-Bugs-7233 ] Net::IMAP seems unable to parse NOMODSEQ line from Cyrus IMAP server
9708 [noreply ruby] Bugs item #7233, was opened at 2006-12-12 11:36

^ [ ruby-Bugs-7253 ] Sets and String subclasses
9713 [noreply ruby] Bugs item #7253, was opened at 2006-12-13 12:26
9714 [murphy rubyc] require 'set'
9715 [jan.svitok g] hsh[key] = value => value
9716 [jan.svitok g] The real problem is in string.c:str_new4() that copies only string

^ [ ruby-Bugs-7262 ] Bug in string literal '\c'
9717 [noreply ruby] Bugs item #7262, was opened at 2006-12-14 00:19

^ [ ruby-Bugs-7263 ] bug with string literal '\c'
9718 [noreply ruby] Bugs item #7263, was opened at 2006-12-14 00:20
9719 [shyouhei rub] Fixed in 1.8.5.

^ [ ruby-Bugs-7274 ] Inconsistant parsing within a block
9720 [noreply ruby] Bugs item #7274, was opened at 2006-12-14 13:08
9721 [drbrain segm] On Dec 14, 2006, at 13:08, <noreply@rubyforge.org>

^ [ ruby-Bugs-7278 ] Kernel#system broken inside Dir.chdir(&block) if system command doesn't have shell characters
9722 [noreply ruby] Bugs item #7278, was opened at 2006-12-14 13:59
+ 9723 [michael.seli] I bet your script /usr/local/bin/p4 is a shell script which doesn't have a
| 9724 [drbrain segm] Nope.
| 9728 [perlguy gmai] Sounds like the difference betwen exec and /bin/sh running p4 could still be
| 9732 [ryand-ruby z] I think you're both barking up the wrong tree.
+ 9733 [nobu ruby-la] Could you elaborate the problem?
  9767 [drbrain segm] There's a minimal test case in [ruby-talk:230639].
  9768 [nobu ruby-la] Thanks for the pointer.
threads.html
top