11576-13175 subjects 11945-18230

^ Inheritable mixin
11673 [transfire gm] Concept for Ruby 2.0...
+ 11684 [matz ruby-la] By separating modules and classes, we get class hierarchy design
| + 11686 [cdcarter gma] Matz,
| | 11687 [matz ruby-la] Perhaps I read him wrong.  But it doesn't still make sense.  Modules
| | + 11688 [transfire gm] That's exactly what I am proposing. But, Module and Class do share
| | + 11723 [transfire gm] This part I don't understand. They both share functionality like
| + 11689 [transfire gm] Okay, mind you this is in the context of Facets, where I'm inclinedto
+ 11685 [warlickt ope] I agree.  Class extends Module is incorrect by definition.  For Class to

^ ri File Broken in 1.8.6
11675 [james graypr] We need to figure out a way not to end up with the data from a
11678 [ryand-ruby z] hrm?
11680 [james graypr] $ ri -T File

^ rb_cstr_to_inum use of strtoul as an optimization has unfortunate side effects
11691 [florgro gmai] rb_cstr_to_inum which is used for methods like String#hex,
11693 [florgro gmai] On another note, String#oct allows the base to be changed by a base
11695 [nobu ruby-la] Yes, it's intentional.
11696 [florgro gmai] OK, any chance we can adjust the RI documentation?
11698 [florgro gmai] Returns 0 if the conversion fails.

^ String#rindex(other) doesn't try to convert other via to_str
11692 [florgro gmai] obj = Object.new
11694 [nobu ruby-la] Though rindex accepts also Regexp and Fixnum, but String should
+ 11697 [florgro gmai] Hm, is there a conversion method for Regexp?
| 11702 [nobu ruby-la] to_regexp for Regexp, to_int for Integer, to_ary for Array, and
| 11704 [florgro gmai] I can't find it documented anywhere (and it doesn't seem to be defined
+ 11706 [matz ruby-la] Can you commit?

^ What are the semantics of to_str?
11699 [jflam micros] Should to_str return an instance of String?
+ 11721 [charles.nutt] That's where you went wrong; <=> does not call to_str.
+ 11722 [florgro gmai] This is quite odd actually. MRI just does a respond_to?() on to_str
  11725 [charles.nutt] Such are the joys of implementing Ruby. I think we've got it mostly

^ [PATCH] XMLRPC::ModRubyServer
11720 [cwailes2 uiu] Made some changes to the XMLRPC::ModRubyServer class that allows it to be loaded as a RubyHandler by ModRuby and prevent it from being reloaded every time a request is made.  This can greatly improve performance when the server takes a while to load, such as when it is establishing database connections or loading lots of libraries.

^ ruby extension help
11724 [python152 gm] I am working on a ruby extension for an existing C library, and ran
11726 [nobu ruby-la] How would char **args be supplied?
11729 [python152 gm] thanks for reply.
11732 [ryand-ruby z] Sorry Oliver, but this isn't the place for these types of questions.

^ indentation / emacs woes
11754 [ryand-ruby z] OK. Can someone give me the tweaks I need to make ruby source read
11755 [charles.nutt] Dunno about emacs, but if I set tabs to eight characters things line up
11757 [ceo hawthorn] (setq indent-tabs-mode nil)          ##This will convert tabs to the
11759 [ryand-ruby z] Um. no. It isn't up to us to modify the source to suit our whitespace
11762 [ceo hawthorn] You may be right -- However converting TABS does not change to
11766 [steven lumos] You missed the point.  If everybody sets their editor to mangle tabs
+ 11767 [david davidf] M-x set-variable
+ 11769 [ceo hawthorn] Is ignoring differences "of spacing" NOT an OPTION?
| 11833 [steven lumos] I'm no svn wizard, but if you could configure it to ignore whitespace
+ 11770 [nobu ruby-la] $ grep -B4 tab misc/ruby-style.el

^ [BUG] threads and heavy io on osx and linux
11756 [Ara.T.Howard] i've tested on several platforms (*nixes) all all will hang if run a
+ 11758 [vjoel path.b] Not with an arg of 10 (tried that about 10 times), but it does sometimes
+ 11760 [ville.mattil] times
+ 11761 [vincent.isam] I tried running it on Mac OS X 10.4, and all versions of Ruby I tried
+ 11763 [brabuhr gmai] Red Hat Enterprise Linux Client release 5 (Tikanga)
| 11765 [brabuhr gmai] Only seems to hang if the et or ot threads are joined in the popen3 block.
+ 11764 [benschumache] FreeBSD 4.11-RELEASE-p11 #0 (P3-850Mhz)
+ 11768 [ml synthetic] Hung on the 13th run.
| 11771 [ara.t.howard] how about this?
| + 11772 [ml synthetic] Success for me, same ruby/OS as before.
| + 11773 [brabuhr gmai] Red Hat Enterprise Linux Client release 5 (Tikanga)
|   11774 [ara.t.howard] !!! wow.  guess that's good!
|   11775 [brabuhr gmai] I've tried running the program under ptrace.  It appears that when the
|   11776 [ara.t.howard] right - ruby uses select under the hood for thread scheduling but
+ 11789 [ara.t.howard] charset=US-ASCII;

^ Bug? array.push(*array) segfault
11777 [rm_rails che] Is this a good place to report bugs?
11778 [nobu ruby-la] Yes.

^ What libraries to be unbundled?
11795 [nakahiro sar] Note that this topic is based on the premise bundling RubyGems with
+ 11796 [ggarra advan] +1 to unbundle - too specific and Drb, for most things, is better
+ 11797 [david davidf] I don't think that json should be unbundled.  It is the interchange
| + 11798 [james graypr] Ruby is a general purpose programming language.  One use it gets put
| | 11799 [david davidf] Obviously, I'm only stating my opinion about the relative importance of
| | 11800 [blowmage gma] Precisely!
| | 11801 [tjw omnigrou] Indeed.  The smaller the standard library is, the more likely it is
| | 11832 [evan falling] There is also a middle ground that has not been mentioned that Chad
| | + 11834 [zimbatm oree] Personally, I like ruby with "all strings attached". I remember that when I
| | + 11835 [ruby dougle.] This seems like a logical approach. Give the user what they expect,
| | + 11836 [meta pobox.c] I would certainly like to see RubyGems as a standard part of the
| + 11802 [ryand-ruby z] I'm not going to repeat Jame's concerns, but I agree with him.
|   11804 [david davidf] NaHi's original message included a small list of candidates for
|   11805 [james graypr] I'm a little fuzzy on why it's written in C for similar reasons.
|   11822 [charles.nutt] As well as being portable and predictable across implementations, since
+ 11825 [shyouhei rub] I'd push mailread.  It's nothing but a namespace pollution (holds "::Mail")
+ 11848 [sheepman she] I oppose the unbundle of webrick because webrick provides webrick/cgi,
+ 12324 [nakahiro sar] Here the list of bundled libraries of 1.8, and statuses on 1.9,
  12327 [james graypr] Is this offer open to those who don't currently have commit access?
  + 12349 [nakahiro sar] I think it's open but Matz has the final say.  Matz?
  | + 12364 [james graypr] Sure.  Just let me know what I'm expected to cover.
  | | 12535 [nakahiro sar] I haven't get a response for a week.  List updated.
  | | 12536 [james graypr] Thanks for letting me know.
  | + 12538 [mneumann nte] James, go for it! Please, feel free to apply patches!
  |   12541 [james graypr] Thanks Michael.  I'm glad to see you haven't left us completely.
  |   12545 [mneumann nte] I'm working on other Ruby-related projects ;-)
  + 12350 [matz ruby-la] Sure, I don't think there's any problem granting commit account for

^ Status of installed.files
11813 [ser germane-] charset="utf-8"
11814 [ser germane-] charset="iso-8859-1"
11816 [nobu ruby-la] You're right.  It'd been forgotten to update too often, so now

^ Re: Parsing bug with 'return not' (was, Syntax bug)
11815 [rick.denatal] Except that return isn't an operator.  As far as I can tell, looking

^ Columnize, irb, and ruby-debug
11821 [rocky.bernst] I've been working on/with Kent SIbilev's ruby-debug. The current sources in
11823 [cdcarter gma] Perhaps as an extension to, or bundled with PrettyPrint/PP ?
11824 [rocky.bernst] Yes, PrettyPrint/PP seems like a good place. I'll work up a small patch for

^ Rdoc allowing arbitrary HTML
11826 [robert.dober] I just really needed this for footnotes and some fun in my rdoc.
11827 [ryand-ruby z] please file this as a patch on the ruby project on rubyforge and
11828 [vjoel path.b] But the OP should be aware that there is a bit of a backlog. There are
+ 11829 [robert.dober] OP is not aware of anything, OP just wanted to share the code, that's
| 11830 [vjoel path.b] Sorry, Robert (I'll stop calling you "OP"--that's a bad habit!). I
| 11831 [robert.dober] No offense taken, but it was extremely funny to talk of myself as OP. :)
+ 11837 [ryand-ruby z] No, patches should not come to the ML first anymore. I've taken over

^ Bullet points on Ruby 1.8 vs. 1.9 language??
11840 [znmeb cesmai] Is there some list of "bullet points" on the major differences between
+ 11841 [james graypr] James Edward Gray II
+ 11842 [david davidf] Inspired by this query, I've written up a bulleted list of the important

^ [PATCH] Curses - add default_color() routines
11843 [pat polycrys] I added conditionally compiled functions to ext/curses of Ruby 1.9 to

^ Mac OSX singleton
11849 [fygrave gmai] This might be an actual bug in ruby itself. any comments?
+ 11850 [hgs dmu.ac.u] In so far as your example shows no rails code....
+ 11851 [sylvain.joye] This is a common misconception about singletons. Most singleton

^ continuations in Ruby 1.9?
11852 [david davidf] In a comment on my recent blog post
11853 [binary42 gma] Considering the availability in a default build I think it is the
11855 [ko1 atdot.ne] Do you have any concrete data?
+ 11857 [chad chadfow] Can you elaborate on this, Koichi?
| 11858 [znmeb cesmai] Can someone post test cases or benchmarks for call/cc? I'm cranking up
+ 11859 [binary42 gma] $ cat callcc.rb

^ [PATCH] ext/Win32API - Manual FreeLibrary support
11854 [abhisek ruby] I think it will be great if Win32API allows manual free'ing of
11856 [nobu ruby-la] The use of Win32API is no longer encouraged, use DL instead.

^ Is this really what we want?
11860 [james graypr] I'm investigating some recent breakage in FasterCSV and have tracking
+ 11862 [Daniel.Berge] Something definitely broke between 1.8.5 and 1.8.6 with Date.parse, as
+ 11864 [matz ruby-la] I asked the author, and he told us it's a side effect of parse
  + 11865 [james graypr] We definitely need to update the documentation then.  It says it
  | 11869 [matz ruby-la] It says "An ArgumentError will be raised if +str+ cannot be parsed."
  + 11866 [hgs dmu.ac.u] Can we take that last sentence as an invitation for someone else to
    + 11867 [james graypr] Boy, I sure hope so.  I mean, it did use to validate dates, so I'll
    | 11870 [Daniel.Berge] Perhaps we need a port of this, or something like it.
    + 11868 [matz ruby-la] If he's willing to accept.  We have to ask him first.

^ REXML changes
11863 [nobu ruby-la] These changes won't be in 1.8.7, is it right?

^ ruby-openssl: == incorrect for X509-Subjects
11871 [hadmut danis] When comparing two objects of class
11872 [ruby dougle.] Test case?
11875 [hadmut danis] Call with a file path to any certificate .pem file
11878 [nakahiro sar] I'm not the original author but I think it's intentional.  DN equality

^ Thread.abort_on_exception= in Ruby 1.9
11873 [david davidf] Thread.abort_on_exception=true is not working the way I would expect it
+ 11874 [david davidf] I've just checked out fresh Ruby 1.9 sources from svn and have confirmed
+ 11877 [matz ruby-la] The 1.9 does not handle abort_on_exception yet.  It should be before
  11879 [nobu ruby-la] You meant [ruby-core:11874]?  Those stackframes seem to be
  11880 [matz ruby-la] Yes.

^ priorities of newly-created threads
11876 [david davidf] In Ruby 1.8 a newly created thread has the priority of the thread that
11881 [nobu ruby-la] The current VM uses native threads.  On linux, the scheduling
+ 11882 [david davidf] Nobu,
| 11883 [david davidf] More specifically, this is the patch to thread_create_core I propose.
| 11885 [sylvain.joye] I tend to disagree on that. I was very surprised when I saw that priority
+ 11884 [sylvain.joye] Would it be possible to have access to the scheduling class of the

^ Core dump with simple web scraper when run via cron
11886 [Daniel.Berge] Solaris 10
+ 11887 [Daniel.Berge] /usr/local/lib/ruby/1.8/timeout.rb:52: [BUG] Segmentation fault
+ 11888 [nobu ruby-la] Didn't you see [BUG] message?  If it wasn't, maybe stack
  11889 [Daniel.Berge] I'm afraid I don't understand your point. What is the significance of

^ Ruby and Solaris door library
11890 [asari.ruby g] the guidelines, but if I made mistakes, please kindly point them out.
+ 11892 [djberg96 gma] I think this is best done as a 3rd party library given that it's
| 11896 [asari.ruby g] Now that Solaris is open source, it is conceivable that door library
| + 11897 [shyouhei rub] +1 to merge into core.  Making an extension is generally preferred, but
| + 11911 [Daniel.Berge] When it actually *is* ported, and people *actually* use it, then we
|   11914 [asari.ruby g] I wonder if anyone other than Dan and Urabe-san has opinions.
|   + 11915 [hgs dmu.ac.u] I see the wisdom of the YAGNI approach, but solaris is out there,
|   | 11929 [asari.ruby g] Well, I actually want/need it now.  As I stated before, I have the C
|   | 11938 [Daniel.Berge] class File::Stat
|   + 11916 [sroberts uni] There is a linux version of Doors, though it is beta, and I think its
|   + 11936 [nobu ruby-la] I second Dan.  It's easy to add File.door? and File::Stat#door?
|     11937 [Daniel.Berge] I could add this to my solaris-file library at
|     11939 [asari.ruby g] Indeed, it might be....  I'll take a look at this library.
|     + 11940 [ayyappank ma] This is a multi-part message in MIME format.
|     + 11943 [djberg96 gma] I've added it to 0.3.3 and released it today. I recommend we take any
+ 11898 [steven lumos] true?
  11901 [asari.ruby g] Indeed.  :-)
  + 11902 [ville.mattil] SunOS new-beat 5.8 Generic_117350-26 sun4u sparc SUNW,Ultra-60
  + 11908 [hgs dmu.ac.u] Seems to have existed on Solaris 8.  When we had solaris 8 I managed
  + 11917 [steven lumos] SunOS sam 5.8 Generic_117350-39 sun4u sparc SUNW,Sun-Blade-2500

^ Behavior of local_variables in Ruby 1.8 vs 1.9
11891 [charles.nutt] $ ruby -e "p local_variables; a = 1; p local_variables; b = 1; p
11942 [charles.nutt] For what it's worth, I've committed the change to JRuby to support the

^ UDP sockets raise exception on MIPS platform
11893 [B.Candler po] I am running ruby-1.8.6 under OpenWrt (*), which is a small MIPS platform
12016 [B.Candler po] #  define READ_DATA_PENDING(fp) (!feof(fp))
12020 [B.Candler po] OK, I think I have the correct solution now :-)

^ IO#seek and whence problem
11894 [lists bertra] on exactly 1 of 5 machines here I have a problem
+ 11895 [lists bertra] The inner ones of course. Sorry!
+ 11905 [lists bertra] I completely turned off optimization and distcc.
  11906 [B.Candler po] Two arguments, one format specifier?
  11913 [lists bertra] Arrrgh! Not more than a typo.

^ pack/unpack 64bit Integers
11899 [hadmut danis] do the pack and unpack procedures support 64 bit signed/unsigned
11903 [B.Candler po] The documentation is at http://www.ruby-doc.org/core/
+ 11904 [B.Candler po] irb(main):001:0> str = [0x8877665544332211].pack("Q")
+ 11919 [hadmut danis] Thanks for hinting me.
  11920 [B.Candler po] Correct. There's no "network byte order 64-bit" conversion.

^ missing bison, gperf not detected, do I need ruby to build ruby?
11900 [szabgab gmai] It seems ./configure did not detect the fact that bison was missing from
11907 [hgs dmu.ac.u] IIRC it would try to use yacc otherwise, but your output makes me think
11909 [vincent.isam] No, Ruby 1.9 requires bison, yacc is not enough.
11910 [binary42 gma] I think trunk requires ruby to generate the respective vm files from
11912 [szabgab gmai] Thanks for the comments.

^ Latest benchmark results
11918 [znmeb cesmai] As I mentioned, I'm starting to do some profiling on Ruby. I just

^ Kernel#proc and Kernel#lambda
11921 [jflam micros] Dumb question of the day: are Kernel#proc and Kernel#lambda identical?
+ 11922 [vincent.isam] - in 1.8 yes
| + 11924 [jflam micros] Thanks! That's what the behavior seemed to indicate.
| | 11927 [dblack rubyp] I think that in 1.9, proc and Proc.new will be identical, and lambda
| + 11925 [Daniel.Berge] Thanks Vincent.
| + 11926 [B.Candler po] ... but not the same as Proc.new
+ 11923 [vincent.isam] ...
+ 11928 [ryand-ruby z] Please don't hijack threads. Start a new one separately.

^ Bug in select?
11930 [robert.dober] I was just testing Labrador with JRuby when JRuby failed a test MR
+ 11931 [matz ruby-la] Hash#select invokes
| + 11932 [shyouhei rub] Would you explain a bit more about it?  It's rather confusing than useful to me.
| + 11933 [robert.dober] Good to know, especially because I believe JRuby wants to be 100%
+ 11934 [florgro gmai] all of this behaviour (and more) is part of Rubinius' executable Ruby

^ Lambda return value
11935 [Tomas.Matous] In the following sequence of calls/yields all return values are 1 except fo=

^ Re: Hijacked thread
11941 [B.Candler po] Have you tried Google?

^ Freenode#ruby-core
11944 [ko1 atdot.ne] We made an IRC channel Freenode#ruby-core to discuss about Ruby