13289-14004

^ Proposal of a new operator for Method and Proc
13289 [mumismo gmai] I think I found a powerful new operator for ruby that can make the
13290 [ggarra advan] You probably want to learn more about ruby blocks, procs and variable
+ 13294 [mumismo gmai] On Nov 8, 2007 7:03 PM, Gonzalo Garramu=F1o <ggarra@advancedsl.com.ar> wrot=
| 13295 [halostatue g] T24gMTEvOC8wNywgSm9yZGkgPG11bWlzbW9AZ21haWwuY29tPiB3cm90ZToKPiA9PSBUaGUgc2Vh
| 13300 [mumismo gmai] T24gTm92IDgsIDIwMDcgMTA6NDUgUE0sIEF1c3RpbiBaaWVnbGVyIDxoYWxvc3RhdHVlQGdtYWls
+ 13297 [transfire gm] Your proposal is similar to Rake tasks, and pre and post wrappers.
  + 13299 [mumismo gmai] I wasn't aware of Rake's tasks. In fact, I have never used Rake :P.
  + 13311 [transfire gm] s/I'd think/I don't think/

^ [ ruby-Bugs-15425 ] Leak with regexp in method with no local vars.
13292 [zimbatm oree] Let me know if you need any help. I think that it doesn't affect only regexpes.
+ 13317 [rogerpack200] Also reproducible with
| 13330 [ryand-ruby z] Not on the ruby that ships with osx 10.5.
| 13334 [zimbatm oree] Do you know if they published the applied patches ?
| + 13335 [rogerpack200] try it without RUBYOPT= anything?
| + 13336 [rogerpack200] It appears that when you have
| + 13339 [snakagawa in] Satoshi Nakagawa
|   13340 [zimbatm oree] Thanks, I've seen them, but they don't seem to be for osx 10.5
+ 13323 [tom infoethe] The problem is that the messages from the RubyForge bug tracker come

^ The significance of aliases vs synonyms
13301 [Daniel.Berge] The behavior below is the result of not using rb_define_alias() in the

^ unsubscribe
13304 [mmcconnell r] unsubscribe

^ The document of random algorithm?
13305 [yedingding g] It's said that we can get a copy from
+ 13307 [hgs dmu.ac.u] Found the citation for this: you are quoting random.c from the ruby source.
| 13308 [yedingding g] Yeah. I'll try it. Thanks.
+ 13313 [dooby d10.ka] Try this Wikipedia entry which has the current* MT home page link.
  13314 [dooby d10.ka] No, it's worse - I forgot to give the link - sorry.
  13316 [yedingding g] Great. That's it. Thanks very much.

^ Re: JRuby/Ruby top-level classes? (Fwd: Metaclasses)
13306 [hgs dmu.ac.u] [Quoted that much to make comparison easier.]
13312 [hgs dmu.ac.u] Oh, sorry.
13322 [charles.nutt] We don't intentionally target 1.9, but this and a few other cases (e.g.

^ primary encoding and source encoding
13315 [david davidf] I've got a couple of questions about the handling of primary encoding.
+ 13324 [nobu ruby-la] It's planned, but we have no mappings from locale name to
| + 13326 [david davidf] Did you consider nl_langinfo(CODESET)?  Your code looks good to me,
| | 13329 [nobu ruby-la] It needs setlocale() to be called before it, which sets global
| + 13327 [david davidf] Maybe you could put your locale_encoding() code in the rb_enc_primary in
| + 13328 [matz ruby-la] Hack is far better than nothing.  Can you commit?
|   13343 [david davidf] Now that primary encoding can be explicitly set, I wonder if the -v
|   13355 [matz ruby-la] I am not sure if we find it useful to know the primary encoding via
|   + 13357 [david davidf] Note that I meant to say "now that the primary encoding can be
|   | 13456 [david davidf] In a related matter, I see that the encoding method of string literals
|   + 13381 [hramrach cen] I have written a script for that in 1.8 already so I guess it is useful ;-)
+ 13331 [matz ruby-la] ASCII-8BIT, yes.
  + 13344 [david davidf] So -K should *not* set the source encoding if the script does not have a
  | 13356 [matz ruby-la] Compatibility problem.  I didn't think about that.  So at least -K
  | + 13358 [david davidf] It seems to me that -E and --encoding should work the same way.  And
  | | 13360 [david davidf] (Though I suppose it could introduce performance
  | + 13359 [nobu ruby-la] Then, separate it from -E option?
  |   13361 [david davidf] This seems confusing, and I don't understand why it is necessary.  I
  |   13373 [nobu ruby-la] I think -K to be deprecated in the future but now existing only
  |   13374 [ed.odanow wo] This should be valid for scripts, that are used by "require" and "load" too.
  + 13375 [akr fsij.org] I think "primary encoding" doesn't tell your intent well.
    13428 [nobu ruby-la] Move to `IO.default_encoding'?
    + 13429 [akr fsij.org] I feel it is applied to sockets but actually not.
    | + 13431 [david davidf] So sockets are always binary by default, correct?
    | | 13443 [akr fsij.org] Sockets are always binary by default, even on 1.8.
    | + 13581 [akr fsij.org] I made a patch for this.
    + 13433 [now bitwi.se] Why not simply Encoding.default?
      13445 [akr fsij.org] "The role for the primary encoding is the default for file
      13446 [ed.odanow wo] Am I right, that this behaviour will be implemented and available in parallel

^ [PATCH] http compression, zlib agnostic, for 1.9
13347 [hgs dmu.ac.u] This message is in MIME format.  The first part should be readable text,
13967 [aaron tender] My system has zlib, but I made it break on purpose to test on-zlib
+ 13974 [ryand-ruby z] you install ruby on a system without libz (the actual z library--not
+ 13975 [hgs dmu.ac.u] Thank you for doing that work.
+ 13986 [hgs dmu.ac.u] with a note to me that he wants people to review it.  Thank you
  14004 [aaron tender] Great!  Thank you so much Matz!

^ Keyword Arguments
13351 [transfire gm] Peter Vanbroekhoven mentioned this to me and I have to agree. I'd
+ 13352 [david davidf] Given that Matz has been planning on a Christmas release for a year, and
| 13367 [transfire gm] OK. If this works in 1.9, then I think it's good enough.
| 13684 [transfire gm] I just realized that this isn't good enough b/c it wouldn't properly
+ 13353 [binary42 gma] I would say that 1.9 should be released as planned with possible
  13460 [sroberts uni] Me, too. I didn't used to see the point, given that it was easy to pass

^ RubyGems imported into 1.9 trunk
13362 [drbrain segm] There are a few tests breaking due to rbconfig.rb not matching what ./
+ 13366 [mumismo gmai] I think it was pointed out in this same list but, getting gems to the
| 13376 [drbrain segm] Due to changes in 1.9 from 1.8, it makes no difference.
| 13379 [david davidf] Eric,
| 13427 [drbrain segm] Yes.
| 13432 [david davidf] But we're still waiting for a gem_prelude.rb to be checked in to make
| 13439 [drbrain segm] Yes, Rich is working on that.
| 13521 [tmorgan99 gm] Perhaps this is more specific to rubygems itself, but is there a reason why
| + 13522 [drbrain segm] Why shouldn't it?  I work with multiple platforms.
| + 13524 [dudley misno] There's not a way to do that at the moment, but the next release of
+ 13430 [drbrain segm] For the record, I've added two tasks to RubyGems' Rakefile to help

^ IO.read, IO#read (and similar methods) - Length Parameter Usage for Non One-Byte Encodings
13363 [ed.odanow wo] Good morning dear Ruby folks!
13364 [david davidf] Matz has said (off-list) that any methods that take length parameters
13365 [ed.odanow wo] Only to clarify and to avoid misunderstandings: This does mean, that the rad

^ method names in 1.9
13368 [dblack rubyp] I know that a 1.9 feature-freeze is coming soon, but I just wanted to
+ 13369 [matz ruby-la] - send, invoke_method
| + 13370 [dblack rubyp] I agree (no bang for safer version). I guess my favorite was send
| | 13371 [matz ruby-la] I don't think it's a good idea to inverse it for bang's sake.  It
| | + 13383 [mumismo gmai] I know that I'll be at best ignored by this comment but when I
| | + 13384 [rogerpack200] To this day I don't know what splat means (meaning--it's not too intuitive).
| |   13387 [charles.nutt] Groovy names it "spread". I don't have a preference.
| |   13408 [now bitwi.se] I like this name a lot.
| + 13388 [charles.nutt] What's the use-case/reason for providing a visibility-honoring send?
| | 13403 [halostatue g] Good question. I know that there have been a lot of posts from people
| | 13410 [david davidf] There was a discussion on ruby-core.  You can probably find it if you
| | + 13413 [charles.nutt] I thought it might be something like that. For that purpose,
| | | + 13415 [dblack rubyp] But "calling it" still really means "sending a message to the object";
| | | | 13458 [sroberts uni] call and send seem to be words from different OO languages for the same
| | | | 13469 [dblack rubyp] They're actually very different. You send the message "call" to
| | | | 13472 [pbrannan atd] I think you're confusing the issue by appealing to details in Ruby's
| | | | 13474 [dblack rubyp] I might say either -- but if I say the former, it's just a kind of
| | | | 13480 [zimbatm oree] IMHO the private/public "problem" comes from Java people who sees
| | | | + 13483 [vjoel path.b] I like that.
| | | | | 13494 [mumismo gmai] I like it also.  I guess dispatch! would be better if it is supposed
| | | | | 13548 [mumismo gmai] May I summary and translate this thread to japanese and send it to ruby-dev?
| | | | + 13551 [transfire gm] Eh... not too fond of dispatch. I mean it's ok, but it just doesn't
| | | + 13422 [halostatue g] Why not #call_method then?
| | |   + 13424 [mumismo gmai] method().call() sends the message call, call_method seems to call a
| | |   + 13438 [dblack rubyp] a_method.bind(obj).call
| | + 13423 [mumismo gmai] - Surprising or not, sending messages is not the same that calling
| |   + 13426 [mboeh desper] I'm not sure I follow you here. It seems to me that the 'dot' is #3; in fac=
| |   + 13437 [dblack rubyp] I don't see how 1/2 differ from 3/4, except in the terminology. What
| |     + 13440 [mumismo gmai] As mathew pointed out in his message, the difference appears when
| |     | 13441 [dblack rubyp] message gets sent to object (via '.' or send)
| |     | + 13442 [mumismo gmai] Absolutelly agree. What I want to know if it everyone agrees.
| |     | | 13447 [mboeh desper] That'd be a pretty radical change, considering that method_missing works by=
| |     | + 13444 [transfire gm] I have of course argued for object_send/instance_send (vs your send/
| |     | + 13467 [dblack rubyp] thing :-) But I think the above does actually make sense to me. It's
| |     |   13473 [hramrach cen] I wonder why this long winded thread did not name the options we want
| |     + 13465 [hramrach cen] As far as I can understand it there are two features we might (not)
| |       13466 [mumismo gmai] I can't understand your last sentence.
| |       13468 [dblack rubyp] I think it's extremely unlikely that send's ability to see private
| |       + 13471 [pbrannan atd] I think that most people expect send to be able to call
| |       + 13477 [halostatue g] I usually use #__send__, not #send.
| + 13389 [transfire gm] Well, there's just #invoke, though that doesn't techincally address
|   + 13390 [mumismo gmai] respond really sounds much better than send. As you said is the
|   + 13393 [mboeh desper] [snip]=20
|     13399 [transfire gm] (*splatting forehead*) Why didn't I think of that? That makes a lot
+ 13386 [transfire gm] But is this distiction usable in nay way? Whe do they not coccur
  + 13391 [mboeh desper] The first thing that comes to mind is method_missing.
  | + 13392 [mumismo gmai] IIRC, in Ruby everything is message sending to objects even in the
  | | 13394 [mboeh desper] to
  | | 13404 [rick.denatal] Oh I hope not.
  | | + 13406 [rick.denatal] Another thought which occurred to me after posting.
  | | + 13414 [mboeh desper] ctively
  | |   13417 [rick.denatal] You are correct. Sorry.
  | + 13396 [dblack rubyp] It's not a question of fitting Ruby to the task of pedagogy. I use my
  | + 13405 [rick.denatal] Understanding this distinction is crucial to understanding why Ruby
  |   + 13407 [dblack rubyp] Well, it's a *bit* like Ruby... :-)
  |   | 13416 [rick.denatal] Oops, I meant "...like C++ or Java"
  |   + 13418 [znmeb cesmai] Interesting distinction ... but some questions ....
  |     13419 [rick.denatal] As a designer, and student of object oriented languages over 25 or so
  |     13420 [znmeb cesmai] Ah, yes, the Byte magazine special language issues. I think I still have
  |     13421 [rick.denatal] Yes, the third, the "Scandinavian view"  doesn't stand in sharp
  + 13398 [dblack rubyp] It depends how you define actual Ruby practice. If you use

^ Link errors for trunk on Mac OS X
13377 [fxrubyguy gm] Apologies in advance if this is a FAQ, but I'm trying to build the
+ 13385 [duerst it.ao] I have very much the same problems on Cygwin, with revision 13875.
+ 13409 [ryand-ruby z] I just built trunk from clean yesterday on leopard and it seems to
  + 13411 [laurent.sans] I also reproduce on my Leopard box, even after a clean.
  + 13412 [vincent.isam] Yes, I checked a bit and revision 13864 (from yesterday) was working
    13425 [fxrubyguy gm] It is fixed as of revision 13888. Thanks, akr, whoever you are. ;)

^ [PATCH] mkmf.rb and create_makefile - important for rubygems
13378 [djberg96 gma] The following patch explains the significance of the second argument to

^ Have you contributed doucmentation to Ruby's std-lib?
13382 [james.britt ] I'm trying to get the names of people who have been contributing rdoc

^ What does "Encoding.compatible?" do?
13395 [ed.odanow wo] While writing Ruby 1.9 informational parts for the german Ruby wiki I saw (ruby
13397 [murphy rubyc] I'd also expect ?-methods to return either true or false...

^ \xDE instead of \336 (was: Re: \u escapes in string literals: proof of concept implementation)
13436 [duerst it.ao] Thanks. Encouraged by this, I worked on a tiny patch, and a small

^ Time#== bug?
13448 [Daniel.Berge] Looks like Time#=3D=3D always returns nil when compared against an =
13449 [jim.weirich ] I don't think it should ever be an error to ask if an object is equal
+ 13450 [Daniel.Berge] =20
+ 13452 [murphy rubyc] not all Comparable operators: the funny thing is that Time.now <=> 0
  13454 [Daniel.Berge] =20

^ patches, and imminent freeze?
13451 [hgs dmu.ac.u] Given talk of the imminent freeze of 1.9 could someone help me grease

^ [PATCH] Limit Ruby 1.8.6 process size
13453 [brent mbari.] This is a multi-part message in MIME format.

^ Help tracking down nasty SegFault bug during GC
13455 [brent mbari.] [Please excuse the possible duplicate posting...]
+ 13461 [znmeb cesmai] Yeah ... Mauricio Fernandez has a way to do exactly that. Try
+ 13462 [ggarra advan] Yes.  Failure to do so will result in random crashes.
  13463 [brent mbari.] Help tracking down nasty SegFault bug during GC
  13464 [brent mbari.] Help tracking down nasty SegFault bug during GC

^ mingw rename
13457 [rogerpack200] Currently for different windows' builds, the names for RUBY_PLATFORM
13459 [luislavena g] if RUBY_PLATFORM =~ /djgpp|(cyg|ms|bcc)win|mingw/
+ 13475 [rogerpack200] Right--this regex reminds me of perl somehow...
| 13476 [luislavena g] Yeah, too much regex for my taste :-P
+ 13481 [drbrain segm] Also, wince.
  13489 [ggarra advan] For what is worth, I've always advocated creating a simplified OS module
  + 13490 [lists-ruby s] Sorry I'm coming in late, but didn't Eric announce a month or two ago that
  | + 13499 [halostatue g] Does Ruby need the architecture triple as a single string? I think the
  | | + 13504 [luislavena g] You're correct. Most information in the GNU triple or RUBY_PLATFORM is
  | | + 13509 [lists-ruby s] Well, I agree that even what autoconf calls the "canonical system type"
  | |   13510 [luislavena g] cygwin
  | + 13503 [luislavena g] T24gTm92IDE0LCAyMDA3IDM6NTcgQU0sIEpheSBMZXZpdHQgPGxpc3RzLXJ1YnlAc2hvcHdhdGNo
  + 13512 [rogerpack200] That would be nice--

^ trunk's parse.c fails to compile
13470 [laurent.sans] I wasn't able to compile trunk using the latest revision, on my
13478 [charles.nutt] bison  -o y.tab.c parse.y
+ 13479 [laurent.sans] I got the same problem on 10.4 Tiger, which includes bison 1.28.
+ 13523 [murphy rubyc] same here on Tiger. after installing Bison 2.3, it worked perfectly -
| 13537 [vincent.isam] Well, if I'm not mistaken bison is only required to build from SVN
| 13538 [murphy rubyc] that makes sense. I didn't know that, thank you.
| 13695 [mytxyang215 ] ../yacc-1.9.1/yacc parse.y
| 13696 [nobu ruby-la] Since 3 years ago.
+ 13525 [david davidf] gcc -g -O2  -I. -I.ext/include/i686-linux -I./include -I.  -DRUBY_EXPORT
  13531 [nobu ruby-la] Bison 2.0 and 2.1 have other problems too, whitch makes broken

^ To the devs : Why do new threads copy the stack ?
13482 [zimbatm oree] Recently, I got a really slow stack overflow because I had a thread
+ 13484 [shyouhei rub] ^ this variable x (out of the thread)
+ 13502 [pbrannan atd] Do you see this same behavior in 1.9?

^ Proposal: Array#walker
13485 [ed.odanow wo] Good morning all together!
+ 13486 [ed.odanow wo] A nicer version may be...
| 13488 [transfire gm] On Nov 13, 8:11 pm, Wolfgang N=E1dasi-Donner <ed.oda...@wonado.de>
| 13495 [transfire gm] Actually, can anyone recommend a better name than #crack? --or a way
| + 13496 [hgs dmu.ac.u] partition (already taken I think), split (taken), divide, cleave, rend,
| | 13527 [murphy rubyc] no, it's not. however, String.split works in a different way, so maybe
| | 13530 [ed.odanow wo] There is one big difference between the actual proposals and my original
| | + 13532 [transfire gm] On Nov 14, 7:47 pm, Wolfgang N=E1dasi-Donner <ed.oda...@wonado.de>
| | | 13540 [ed.odanow wo] Oops - it is not easy to express in English what I really mean. I don't
| | | 13544 [now bitwi.se] On Nov 15, 2007 11:32 AM, Wolfgang N=E1dasi-Donner <ed.odanow@wonado.de> wr=
| | | 13546 [ed.odanow wo] I think it is more an example for extending standard classes in a Ruby book.
| | + 13539 [murphy rubyc] I know :) I wanted to propose a more general method that would still
| |   13553 [transfire gm] This is nice. Add in an extra length option, and IMO it's a worthy of
| + 13501 [pat polycrys] Maybe Array#partition_at to match Enumarable#partition ?
+ 13487 [murphy rubyc] Am 14.11.2007 um 01:37 schrieb Wolfgang N=C3=A1dasi-Donner =
| 13492 [dblack rubyp] This message is in MIME format.  The first part should be readable text,
| 13500 [ed.odanow wo] The only reason I put it here is, that I don't see a way to express it without
| 13508 [transfire gm] On Nov 14, 8:32 am, Wolfgang N=E1dasi-Donner <ed.oda...@wonado.de>
+ 13511 [transfire gm] On Nov 13, 7:37 pm, Wolfgang N=E1dasi-Donner <ed.oda...@wonado.de>