15182-15607

14918-15184 subjects 15360-15979

Empty bin and lib install directories with Rev 15170
15182 [rubys@in er ] ./lib
15183 [rubys@in er ] 15173 is much better.  Thanks!

[PATCH] Friendlier methods to compare two Time objects
15185 [jim.cropcho@] ...
+ 15187 [jim.cropcho@] ...
+ 15189 [duerst@it ao] There is even some cultural dependency here. In Japanese, for example,
  + 15190 [jim.cropcho@] ...
  + 15191 [mblondel@ru ] Ф = next week, lit. below week
    15193 [jim.cropcho@] ...
    15207 [jim.cropcho@] ...
    15208 [jim.cropcho@] ...
    15209 [duerst@it ao] I don't think so. Ruby doesn't localize it's method names.
    15224 [jim.cropcho@] ...

Build failure in 15174
15188 [dave@pr gp o] installing binary commands

build failure, rev 15195-15201
15192 [rubys@in er ] I don't know if others are seeing this, but since

Can large scale projects be successful implemented around a dynamic programming language?
15194 [mumismo@gm i] ...
+ 15195 [phlip2005@gm] A dynamic language + test driven development is considered, in some circles, to
+ 15196 [james@gr yp ] Well, I didn't find it to be a good article at all.  I felt it was
  15205 [lists-ruby@s] I find it telling that the article begins by citing Eclipse as a paragon of
  15219 [ks@ku ts ep ] What does any of this have to do with Ruby core?
  15220 [phlip2005@gm] Way to answer your own question!
  15229 [mumismo@gm i] ...
  15230 [alexey.verkh] Sorry if this comes across as rude (it's definitely not the

broken rexml
15197 [jan.dvorak@k] require 'rexml/document'
15198 [rubys@in er ] I'm having trouble building Ruby at all right now, example from 15221

Two build issues
15199 [dave@pr gp o] ...
15201 [decoux@mo lo] It find a \n at the end of its internal buffer
15202 [rubys@in er ] ./miniruby -I./lib  -I. -rrbconfig ./tool/compile_prelude.rb
15203 [decoux@mo lo] I don't have this error, but rather with
15204 [rubys@in er ] Much better!

Troubles building 1.8.6 on Solaris 10
15206 [jrich013@gm ] ...
15210 [haisenko@co ] It's libgcc_s.so.1 it can't find.

String#include? documentation patch
15211 [dblack@ru yp] I noticed that String#include? is still documented in 1.9 as taking a
15222 [matz@ru y- a] Could you check this in, please?

case when syntax changed ???
15212 [yvon.thorava] ...
+ 15213 [rubys@in er ] Colon (":") is no longer allowed as a delimiter.  Use semicolon (";") or
+ 15214 [decoux@mo lo] Yes,
| 15215 [yvon.thorava] ...
+ 15216 [mental@ry ia] Only inasmuch as you need to use ; instead of : for one-liner whens.

ruby-1.9.0-0 and Tk
15217 [yvon.thorava] ...
+ 15218 [christian@su] The bugreport contains a patch.
| 15226 [nagai@ai ky ] For #16634, the following is better.
+ 15225 [nagai@ai ky ] I'm very sorry. That's a known bug.
  15228 [yvon.thorava] ...

Ruby Class Object Serialization -- HELP
15221 [jackxxu@gm i] ...
15227 [phlip2005@gm] In theory, because the class fully exists at runtime as a big stack of nodes
15232 [jackxxu@gm i] ...

[Fwd: SimpleDelegator.class changed?]
15223 [rubys@in er ] Anybody have an answer?  html5lib makes use of this feature.

line event weirdness on if statements. Program to show trace event lines
15231 [rocky.bernst] ...

Regular expressions, string interpolation, and encoding
15233 [rubys@in er ] ~$ irb
15287 [rubys@in er ] No longer an issue with r15311.  Thanks!

Gem install error on head
15234 [dave@pr gp o] I built HEAD into a totally clean destination (so there were no
15326 [muhtaris@cs ] In revision 15359 currently, this still happens...
15327 [ks@ku ts ep ] This has been happening for me for a *long* time.  I can also recreate
15328 [luislavena@g] gem install --debug gem_name
15329 [ks@ku ts ep ] ...
15330 [ks@ku ts ep ] ...

Problem with ri on fresh install
15235 [dave@pr gp o] Built and installed HEAD into a clean tree.

Encoding of unicode strings is now ASCII-8BIT?
15236 [rubys@in er ] $ irb
+ 15237 [dave@pr gp o] Try it outside of irb
| 15239 [rubys@in er ] Dang.  I had hoped that the presence of a \u would unambiguously
+ 15241 [matz@ru y- a] It must be a bug.

CI for Ruby core?
15238 [charles.nutt] Perhaps it's time to set up a continuous integration server for ruby
+ 15240 [rubys@in er ] I'm not sure if you are talking about 1.9 or 1.8, but I have been doing
+ 15242 [znmeb@ce ma ] Well ... it's a great idea. I suppose you have something like that
| 15243 [alexey.verkh] From my (mixed) experience with running a CI build for Rails [1] for
| + 15244 [rubys@in er ] So... the trick is to only include packages for which there are
| | 15246 [alexey.verkh] Exactly. In other words, committers should say "yes, we want a
| | 15247 [luislavena@g] I was running something similar for 1.8 and 1.9 build of Ruby
| + 15251 [charles.nutt] I agree that this is key to making a CI server work. It seems that so
+ 15245 [akr@fs j. rg] I run a small continuous integration system since 2005.
  15249 [rubys@in er ] Does anybody try to fix the failures?

Symbol#empty? ?
15248 [dblack@ru yp] I see that symbols now have an #empty? method, but an empty symbol
15250 [matz@ru y- a] Which do you prefer?
+ 15252 [alexey.verkh] +1 for allowing :""
+ 15253 [dblack@ru yp] Removing Symbol#empty would still leave "".intern without :"", which
| 15255 [matz@ru y- a] OK, let me rephrase:  which do you prefer?
| 15272 [dblack@ru yp] I prefer the latter. But that's partly because I'm skeptical about the
+ 15254 [gwtmp01@ma .] It might be nice if nil.to_sym produced the empty symbol also.  This
+ 15256 [dave@pr gp o] For me, the answer depends on what you (Matz) feel that symbols are.
  + 15258 [gwtmp01@ma .] I think that any sequence of bytes, including the empty sequence
  | 15259 [dave@pr gp o] I don't necessarily disagree, but I'd like to explore.
  | + 15261 [rick.denatal] I'm not troubled by  :+ and :"+"  any more than I am by ["a" "b"]  and %w{a b}
  | | + 15262 [dave@pr gp o] irb(main):003:0> :"+(binary)".object_id
  | | + 15264 [jim.weirich@] $ irb
  | |   15265 [rick.denatal] Ahhh, now that IS disturbing.
  | + 15266 [gwtmp01@ma .] In the most abstract sense, symbols map from string equality to
  |   15267 [dave@pr gp o] OK, but why do we need this?
  |   + 15271 [dblack@ru yp] It's hard to specify what a valid symbol is, though -- not just
  |   + 15273 [gwtmp01@ma .] By inconsistencies, I assume you mean that s.to_sym works even if the
  |     15274 [dave@pr gp o] Actually, I was referring to the leaking of the "sometimes a string,
  |     15275 [gwtmp01@ma .] You were talking about 'the works of Shakespere' as a symbol, which I
  |     15276 [dave@pr gp o] Let's assume that I know what a Ruby Hash is... :)
  |     15278 [gwtmp01@ma .] Of course, which is why your comment 'a Hash object not to_hash' was
  + 15260 [rick.denatal] Ermmmm, I don't think so, []= is (invariably?) a mutating method.
    15263 [dave@pr gp o] Yeah, bad example. Sorry

a new kind of assertion
15257 [phlip2005@gm] I don't like the simple assertions - assert_equal, assert_match,
+ 15269 [meinrad.rech] cool !!!
| 15270 [phlip2005@gm] Thanks! but I forgot to extremely manage expectations these are "unit" test
| 15277 [jim.cropcho@] ...
| 15282 [alex@bl ck e] def runtime_assert_(&block)
| + 15283 [phlip2005@gm] That doesn't magically and efficiently disappear, all the way back to its call
| | 15285 [vjoel@pa h. ] It can be made a _little_ more efficient (swap the && terms; use yield
| + 15284 [jim.cropcho@] ...
|   15294 [alex@bl ck e] Yes.  I guess my point is that it's much easier to reason about these
|   15295 [phlip2005@gm] Just a progress report - I went ahead and chained to assert(). If you call it
+ 15306 [transfire@gm] Hmm... What about extraneous asset block content. Eg.
  15307 [phlip2005@gm] If the last statement passed, the assertion would do nothing. If it failed, the

unsubscribe
15268 [mlist4manu@g] unsubscribe

Is "Regexp#ascii_only?" needed?
15279 [rubys@in er ] $ irb
15281 [rubys@in er ] As of revision 15304, this no longer is an issue as the warning is no

Personal CI tests passing again -- with workarounds applied
15280 [rubys@in er ] I mention this as I believe that the REXML that is present in the Ruby

Circular dependency: revision 15317
15288 [rubys@in er ] - Sam Ruby
15289 [naruse@ai em] sorry, i forgot fix configure.in.
15290 [rubys@in er ] - Sam Ruby
15291 [usa@ga ba ec] Maybe we need some missing files.
15292 [naruse@ai em] Oh, sorry, commited at r15320.
15293 [rubys@in er ] Much better.  Builds clean now.  Thanks!

Deletion of element sequence in an Array
15297 [ed.odanow@wo] In Ruby 1.9 "array[<range>]=nil" is no longer supported to delete
15298 [usa@ga ba ec] array[<range>] = [] seems to be still supported.
15300 [ed.odanow@wo] Thank you, it works.
+ 15301 [muhtaris@cs ] I think I would agree with Wolfgang.
+ 15302 [nobu@ru y- a] It's not an exception.  The setter with a range *always*
+ 15304 [gavin.kistne] I don't think this acts as an exception at all.
  15305 [ed.odanow@wo] This is not what I would expect. The example...

Core team, I need your help
15303 [zimbatm@or e] two weeks ago, I sent two emails containing open issues about ruby
15314 [decoux@mo lo] Well, you have found the bug : it's completely described on rubyforge.
15324 [zimbatm@or e] ...
15325 [decoux@mo lo] Well, with this patch you reverse  [ruby-cvs:17330]

IRHG - TNODE Documentation?
15308 [ceo@ha th rn] Is there any documentation on the TNODE
+ 15315 [pbrannan@at ] You mean the T_NODE types?
| + 15316 [rocky.bernst] ...
| | 15318 [pbrannan@at ] I could put the latest documentation there.  I usually update that url
| + 15317 [ceo@ha th rn] Exactly what URL do you find "nodewrap/ext/node_help.txt"
|   15319 [pbrannan@at ] You will need to check nodewrap out of cvs.  It's on rubyforge.
|   15322 [ceo@ha th rn] Thanks Paul,
+ 15607 [pbrannan@at ] I have completed the documentation for the nodes wrapped by nodewrap.

1.9 ri problem
15309 [usa@ga ba ec] When using trunk, now we cannot look up a method by ri if we
15320 [drbrain@se m] Thank you.  Can you commit it?

FileUtils#cp_r options
15311 [martindemell] Why does FileUtils#cp_r have remove_destination as an option and not

RDoc would cause NameError
15312 [sheepman@sh ] RDoc at trunk would cause NameError.
15323 [ryand-ruby@z] sheepman! you exist! I've been googling for you!

rev 15349-15353 #274 test_syntax.rb:527
15321 [rubys@in er ] #274 test_syntax.rb:527:in `assert_syntax_error': begin eval(%q{0..:""},

How to submit patches for Ruby core?
15331 [ks@ku ts ep ] Is there a process for submitting patches for Ruby 1.9 and 1.8?  I've
15334 [luislavena@g] Suggest you to use RubyForge tracker [1] to submit the patch and then

Incorrect Time#<=> specification
15332 [vsizikov@gm ] ...
+ 15338 [jim.cropcho@] ...
+ 15342 [naruse@ai em] I commited the patch at r15367, thanks.

Incorrect Time#<=> specification (Corrected encoding)
15333 [vsizikov@gm ] ...

assert{ 2.0 } available with assert_xpath
15335 [phlip2005@gm] My assert{ 2.0 } project will soon have its own package. It's available now as a
15340 [ryand-ruby@z] please stop emailing ruby-core about this package. ruby-talk is much

Is there a Windows port of "narray"?
15336 [znmeb@ce ma ] Subject says most of it, except for the "why do I want this?" I want

[PATCH] make Benchmark#realtime a bit faster
15337 [dymo@uk po t] ...
15343 [matz@ru y- a] Just checked in.  Thank you.

Re: Welcome to our (ruby-core ML)         You are added automatically
15339 [identitypriv] guide

[PATCH] Append srcprefix to $INCFLAGS in create_makefile
15341 [Daniel.Berge] The following patch automatically appends the srcprefix to the $INCFLAGS

(none)
15344 [alexandra.u.] ...

Ruby 1.9.0-0 RubyGems and environment variables ???
15345 [yvon.thorava] ...

open-uri problem in 1.9 stops gems from working
15346 [dave@pr gp o] For at least the last week, 1.9 HEAD has been unable to install gems.
+ 15347 [transfire@gm] Looks like the same problem I reported a couple months ago -- and was
+ 15356 [ks@ku ts ep ] I reported the same stack trace in "Re: Gem install error on head", but

Expanding arrays in method calls - why the restriction?
15348 [meta@po ox c] I can do
+ 15349 [phlip2005@gm] system('du', *[opts + '/tmp'])
| + 15350 [phlip2005@gm] system('du', *[opts + ['/tmp']])
| + 15352 [dblack@ru yp] That's changed in 1.9. The starred one will get whatever's in the
+ 15351 [vincent.isam] You can in 1.9

extconf
15353 [tmorgan99@gm] ...
15355 [nobu@ru y- a] Add them to $INSTALLFILES before create_makefile.
15363 [tmorgan99@gm] ...
15367 [nobu@ru y- a] The former.  If you want to install generated files, use

[PATCH] net/http EOF error
15354 [aaron@te de ] I've encountered an error in net/http where if the server closes the

IRHG - OBJECT Types (T_MASK ITEMS)
15357 [ceo@ha th rn] The following are unclear -- Can anyone give
15358 [matz@ru y- a] ...

Timeout::Error
15359 [jerith@je it] This is a cryptographically signed message in MIME format.
15366 [drbrain@se m] No.  The exception would be rescued inappropriately then.
15370 [jerith@je it] This is a cryptographically signed message in MIME format.
+ 15371 [billk@ct .c ] As an aside, you'll likely want to take great care with which
| 15372 [jerith@je it] This is a cryptographically signed message in MIME format.
+ 15373 [nobu@ru y- a] Because Timeout::Error has changed since 1.8 but irb hasn't
| 15374 [jerith@je it] This is a cryptographically signed message in MIME format.
| 15412 [nobu@ru y- a] TimeoutError was a subclass of StandardError till 1.6, but now
| 15413 [jerith@je it] This is a cryptographically signed message in MIME format.
| 15414 [nobu@ru y- a] Sorry, I can't remember it correctly now.  Perhaps, not to
| + 15421 [vjoel@pa h. ] require 'timeout'
| | 15430 [jerith@je it] This is a cryptographically signed message in MIME format.
| | 15435 [vjoel@pa h. ] It not a question of _what_ is recoverable, but _where_ it should be
| | 15444 [jerith@je it] This is a cryptographically signed message in MIME format.
| | 15458 [vjoel@pa h. ] Sorry, I wasn't clear. I'm not disagreeing with you, just suggesting
| | + 15461 [nobu@ru y- a] Agreed your suggestion, even for nested timeouts.
| | | + 15495 [matz@ru y- a] Can you check in?
| | | | 15500 [nobu@ru y- a] What about 1.8?  I think this feature to distinguish particular
| | | + 15504 [mental@ry ia] Is there a way to implement this which doesn't rely on parsing the exception
| | |   15505 [rocky.bernst] ...
| | |   15506 [mental@ry ia] It would help, but frankly this modification of Timeout seems like a
| | + 15471 [dan-ml@da 42] What about the reverse problem? A lazy programmer might do a catch-all
| |   + 15472 [jerith@je it] This is a cryptographically signed message in MIME format.
| |   + 15484 [vjoel@pa h. ] data = [] # Oops!
| + 15431 [pbrannan@at ] I find it particularly annoying that when a unit test times out, the
|   + 15432 [jerith@je it] This is a cryptographically signed message in MIME format.
|   | 15434 [thewoolleyma] The important thing is that Test::Unit should always return a nonzero
|   + 15436 [pbrannan@at ] Actually, I'm looking at the Test::Unit code right now and it appears to
+ 15379 [ jfh@uf .e u] While I tend to agree and ran into the same frustration as you, you can
  15380 [jerith@je it] This is a cryptographically signed message in MIME format.
  15386 [charles.nutt] This is exactly why timeout is a bad idea. Being able to cause the
  15397 [ks@ku ts ep ] We use Timeout frequently for computations that we explicitly want to
  + 15411 [jerith@je it] This is a cryptographically signed message in MIME format.
  + 15422 [mental@ry ia] How hard are your timing constraints?  Remember that the delivery
threads.html
top