17333-18062

17116-30833 subjects 17500-18182

[Ruby 1.8.7 - Bug #193] (Open) Crash during build on Mac OS 10.5.3/PPC when configured with --enable-pthread
17333 [redmine@ru y] Issue #193 has been reported by Florian Ebeling.

Understanding the output of Kernel#caller
17345 [wilsonb@gm i] I am trying to understand what Ruby 1.8 outputs when "caller" is invoked.
17484 [wilsonb@gm i] Unless there is an objection, I will file this on the bug tracker.
17516 [wycats@gm il] ...
17519 [charles.nutt] Why would you expect that? The "calling m_w_b" rescue is the first one

[Ruby 1.8 - Bug #175] (Closed) Rational#power2 raises a NameError or causes infinite loops when passed a Rational
17348 [redmine@ru y] Issue #175 has been updated by Yukihiro Matsumoto.

patches for tests of rubygems
17353 [mame@ts .n .] Twelve test cases for rubygems fail in my environment.
17395 [mame@ts .n .] Any comments?  May I commit them?
17406 [drbrain@se m] Sorry, let me commit them to rubygems then import them to ruby trunk.
17408 [drbrain@se m] Thank you, I have committed to RubyGems and updated ruby trunk.
17442 [mame@ts .n .] In my environment, two types of errors are still output.
17444 [mame@ts .n .] Sorry, above patch doesn't work.
17471 [drbrain@se m] I have applied your patch from [ruby-core:17444] and the runruby.rb
17474 [mame@ts .n .] Thank you, it works!

[Ruby 1.8 - Bug #175] Rational#power2 raises a NameError or causes infinite loops when passed a Rational
17354 [redmine@ru y] Issue #175 has been updated by Arthur Schreiber.
17355 [matz@ru y- a] I'll see.  But note that power2 is not a part of public API of

A faster Array#delete
17356 [djberg96@gm ] Windows XP
+ 17357 [djberg96@gm ] Well, clearly I broke something else, as my change caused a larger test
+ 17362 [djberg96@gm ] for(n = i+1; n < length; n++)
+ 17380 [matz@ru y- a] This patch gave me no significant performance improvement on my Linux
  17381 [Daniel.Berge] =20
  17382 [matz@ru y- a] Your version would have access violation with the following program.
  17399 [Daniel.Berge] =20

1.8.6-p230 test-all fails in test/drb
17364 [luislavena@g] test_07_public_private_protected_missing(TestDRbCore), defined in
17365 [usa@ga ba ec] Hmmm....
17366 [luislavena@g] (I didn't push the p230 changes yet, so it states p114 right now, but

Building on MinGW broken (patch for win32/win32.c)
17367 [larsch@be un] The change in r17376 to win32/win32.c breaks compilation on MinGW/MSYS.
17368 [usa@ga ba ec] Can you test this patch instead of yours?
17369 [larsch@be un] That works too. Thanks!

[PATCH] Improve Class#superclass documentation
17371 [ramos.gaston] I added a bit of explanation and an example on Class#superclass

[PATCH] Add missing Proc#clone documentation
17372 [ramos.gaston] Attached is a patch for add the missing documentation on Proc#clone

[PATCH] Wrong revision.h include on RUBY 1.8 branch
17373 [ramos.gaston] version.h:34:22: error: revision.h: No such file or directory

Strange BigDecimal#add behavior
17374 [vsizikov@gm ] We're writing some more RubySpec tests for Bigdecimal#add, and I'd

[Ruby 1.8.7 - Bug #193] Crash during build on Mac OS 10.5.3/PPC when configured with --enable-pthread
17375 [redmine@ru y] Issue #193 has been updated by Florian Ebeling.

Re: Ruby 1.9.0/1.8.7/1.8.6/1.8.5 new releases (Security Fix)
17377 [billk@ct .c ] I've been following a blog[1] about the security patches, and
+ 17378 [znmeb@ce ma ] Can you run this with "gdb?"
| 17379 [billk@ct .c ] I'm a gdb newbie, but it's high time I learned.
+ 17388 [phil.ross@gm] It looks like this issue has been fixed in revision 17530 in the
+ 17459 [billk@ct .c ] Note, still getting this seg fault as of svn revision 17730
  17460 [shyouhei@ru ] Sorry, cannot reproduce on my machine.  Does anyone have more info?
  17463 [billk@ct .c ] Hmm.. odd.
  17465 [decoux@mo lo] best to see it like this
  17466 [shyouhei@ru ] Thank you Guy, this helped me a lot.  I think I've backported what's
  + 17467 [decoux@mo lo] it seems to work for me
  + 17468 [billk@ct .c ] Ah... so obvious, now.  :D
    17473 [jan.svitok@g] I guess the vulnerability announcement should be updated with links to

[Ruby 1.8.7 - Bug #193] Crash during build on Mac OS 10.5.3/PPC when configured with --enable-pthread
17387 [redmine@ru y] Issue #193 has been updated by Florian Ebeling.

XMLRPC socket patch
17392 [mellon85@gm ] ...
17398 [james@gr yp ] Do you have any idea how common this usage is?  I really don't know,
17400 [mellon85@gm ] ...
17405 [james@gr yp ] Yeah, you're right.  We also just need to weight that against factors =20=
17410 [mellon85@gm ] Thanks for the advices, i've published it on rubyforge.

URGENT: Possible fixes for segfaults and vulnerabilities available for review in ruby-talk
17393 [igal@pr gm t] All currently available versions of MRI Ruby are either vulnerable to
+ 17394 [decoux@mo lo] Please let make matz, he know ruby better than you.
+ 17416 [shyouhei@ru ] Sorry for a late reply but I think I've fixed this issue.  Can someone
  17417 [igal@pr gm t] I'm delighted to hear from you!
  + 17419 [shyouhei@ru ] No, I don't like that menu.  Please take a look at the ChangeLog.  All
  | 17422 [igal@pr gm t] Thank you for the explanation.
  | 17426 [shyouhei@ru ] Yes.  Please write a patch :)
  | 17438 [igal@pr gm t] I'll try to help with the Ruby patches, but I don't know C well enough
  + 17423 [akr@fs j. rg] I read the logs of rubyspec.
    + 17424 [federico.bui] I'm not sure how's the situation with most of these libraries but I
    | + 17425 [akr@fs j. rg] Yes.  REXML has typo.
    | | 17433 [igal@pr gm t] Thank you both for the reviews and for clarifying the situation with the
    | + 17429 [igal@pr gm t] I'm glad you mentioned this. When I was talking about using a stable
    |   17431 [znmeb@ce ma ] Incidentally, Ruby 1.8.7 is in the Portage tree, but it's masked.
    |   17432 [igal@pr gm t] Thanks for the info, Ed.
    |   + 17434 [shyouhei@ru ] 1.8.5 life-span has been scheduled and informed at least since 2006,
    |   | 17435 [igal@pr gm t] Just to confirm, you're terminating support for 1.8.5 and all earlier
    |   | 17441 [shyouhei@ru ] Yes. All releases prior to 1.8.5, including all 1.4 series and all 1.6
    |   + 17446 [znmeb@ce ma ] RHEL *5* is still using 1.8.5? In that case, dropping support for 1.8.5
    |     + 17449 [znmeb@ce ma ] dated 12 November 2007.
    |     + 17454 [shyouhei@ru ] Why on earth?
    |       17456 [vjoel@pa h. ] Bravo.
    + 17436 [igal@pr gm t] Thank you for the help!
      17437 [akr@fs j. rg] ERB tests compare Ruby code generated internaly in ERB.
      17439 [igal@pr gm t] Ah! Correct. I've reviewed that code more carefully and see what you
      17440 [akr@fs j. rg] In your log, which is not guarded?
      17445 [igal@pr gm t] Sorry, I was wrong. I've reviewed all the REXML errors and they seem to
      17447 [federico.bui] The correct guard in this case should be 1.8 as you propose. I'll be

Ruby 1.8.6: Bus Error after revision 17222
17401 [murphy@ru yc] * class.c (clone_method): should copy cref as well.
17402 [igal@pr gm t] All available official versions of Ruby either suffer from

[PATCH] branches/ruby_1_8: Building in separate directory
17407 [larsch@be un] directory from source. Attempts to delete miniruby.exe fails with

Re: Welcome to our (ruby-core ML) You are added automatically
17409 [paul.mcmahon] ...

Time for a release management committee?
17412 [charles.nutt] It seems like recent problems with patchlevel and minor 1.8 releases
+ 17413 [znmeb@ce ma ] Well ... I don't think it's so much a lack of manpower as a lack of
| 17414 [igal@pr gm t] I think this is easier than you think.
| 17418 [dave@pr gp o] Is Charles talking about bugs in specific architectures? I got the
+ 17415 [igal@pr gm t] I believe Charles' proposal has much merit. No one wants to stifle the

1.8 release management
17427 [matz@ru y- a] First of all, sorry for segfaults/troubles you've got.  Let me clear
+ 17428 [znmeb@ce ma ] As I understand it, maintenance has been dropped for 1.8.5. I doubt if
| 17450 [matz@ru y- a] Right.  Now 1.8.6 and 1.8.7 are under stable maintenance.  1.8.8 (to
| 17451 [znmeb@ce ma ] Well ... after I did a little digging and discovered that it appears Red
+ 17430 [shyouhei@ru ] Just FYI, 1.8.7 mentor is now shifting from knu to me.  After that I'm
+ 17443 [igal@pr gm t] That's great news.
+ 17448 [sylvain.joye] Well ... My experience with the Ruby development/release process since 1.8.5 has
| 17453 [matz@ru y- a] <snip unfortunate experience>
| 17461 [sylvain.joye] ... and it has resulted from a four-months breakage of threading.
| 17462 [matz@ru y- a] No, but we at least formed a process that can be a starting-point.  I
| + 17464 [sylvain.joye] Agreed on that. I just did not take the "we don't see any problem here"
| + 17475 [duerst@it ao] I have heard about this. Can somebody please explain how exactly
|   17476 [matz@ru y- a] Of course RedMine can not do this automatically.  A committer puts
|   17478 [dblack@ru yp] If it could, we'd have to change its name from RedMine to ReadMind :-)
+ 17455 [stephen.bann] Let me describe some simple questions about Ruby 1.8.6 that are not
  17458 [shyouhei@ru ] For what I know,
  17547 [wilsonb@gm i] I would like to propose that the 'check' target depend on the
  17549 [igal@pr gm t] That'd be great. I think that RubySpec has more than proven its value
  + 17553 [wilsonb@gm i] Those targets are actually already in the Makefile in the ruby_1_8
  | 17565 [znmeb@ce ma ] ...
  + 17555 [luislavena@g] Don't want to be a pain to someone, but what about a Rakefile instead
    + 17561 [hramrach@ce ] A shell script is also pretty cross platform. After all shell is part
    + 17585 [shyouhei@ru ] Ruby based program to check ruby itself?  Does that really reliable?
      17588 [igal@pr gm t] I think so. Stephen's RubyBuild can build a Ruby interpreter and my
      17589 [shyouhei@ru ] If they fail Ruby is insane.  But their *not* failing do not always
      + 17591 [igal@pr gm t] If the RubyBuild or RubyChecker programs fail to run with only a
      | + 17593 [vsizikov@gm ] RubySpec suite uses "ruby_bug" guards to mark some tests that fail
      | | 17597 [igal@pr gm t] Thanks for writing that up, Vladimir. Sorry for making too broad a
      | | 17601 [vsizikov@gm ] With the very latest RubySpec revision, Arthur and Federico managed to
      | | + 17602 [znmeb@ce ma ] ...
      | | | 17604 [vsizikov@gm ] Here you go. These two been bugging me for a while. They could be seen
      | | | 17626 [matz@ru y- a] According to Nobu, those tests mix buffered and unbuffered IO
      | | | + 17632 [znmeb@ce ma ] ...
      | | | + 17705 [vsizikov@gm ] Thanks, matz,
      | | |   17728 [shyouhei@ru ] Good news.  Though I have a bit more dirty output on my machine.
      | | |   + 17730 [federico.bui] The issues with Socket in your log and in Akira Tanaka's were our
      | | |   + 17738 [akr@fs j. rg] To be accurate, the machine is not mine.
      | | + 17605 [igal@pr gm t] Awesome, good work.
      | | | 17606 [federico.bui] And Vladimir, give yourself some credit.
      | | + 17614 [onepoint@st ] Also Time and IConv failures (attached).
      | |   17628 [onepoint@st ] Update,  this  was due  to  a  bug in  my  ancient  glibc that  caused
      | + 17630 [shyouhei@ru ] Sorry for a late reply,
      + 17599 [stephen.bann] I think as long as there are several forms of test frameworks the likelyhood of running all the tests and having a positive result when the fault still exists AND is being tested for is extremely small.

Ruby 1.8.6 is now patched in Gentoo Linux!!
17452 [znmeb@ce ma ] ...
17469 [nobu@ru y- a] As for memory leak, seems correct, except for that dfree

Ruby status on Gentoo
17457 [graaff@ge to] As Ed mentioned earlier today we now offer a version of ruby 1.8.6_p230
17470 [nobu@ru y- a] The site is complained as the wrong certification.

[Ruby 1.8 - Bug #206] (Open) BigDecimal precision in divmod
17472 [redmine@ru y] Issue #206 has been reported by Anonymous.

[Ruby 1.8 - Bug #211] (Open) "wrong argument type" "expected Proc"
17477 [redmine@ru y] Issue #211 has been reported by Tobi Reif.

[Ruby 1.8 - Bug #212] (Open) Issues with Readline in Mac OS X
17479 [redmine@ru y] Issue #212 has been reported by Federico Builes.

Array#fill behavior
17480 [vsizikov@gm ] I'd like to double-check the Array#fill behavior and to figure out
17481 [akr@fs j. rg] Interesting.  I test more.
17482 [vsizikov@gm ] Thanks for the quick response!
17483 [akr@fs j. rg] I found [ruby-core:12950] (2007-10-26).

[Ruby 1.8 - Bug #211] "wrong argument type" "expected Proc"
17485 [redmine@ru y] Issue #211 has been updated by Eric Hodel.

[Ruby 1.8 - Bug #211] "wrong argument type" "expected Proc"
17486 [redmine@ru y] Issue #211 has been updated by Tobi Reif.

[Ruby 1.8 - Bug #211] "wrong argument type" "expected Proc"
17487 [redmine@ru y] Issue #211 has been updated by Eric Hodel.

HOME and USERPROFILE aliasing under Windows
17488 [jflam@mi ro ] MRI currently expects the HOME environment variable to be set under Windows=
17489 [luislavena@g] USERPROFILE sometimes matches the location of roaming profiles, and
+ 17490 [jflam@mi ro ] Awesome. We'll match the 1.9 behavior then.
| 17492 [jflam@mi ro ] Actually, to clarify this a bit further, we won't *set* HOME, instead we'll=
| 17496 [luislavena@g] ENV['HOME'] || ENV['HOMEDRIVE'] + ENV['HOMEPATH'] ||
+ 17498 [vsizikov@gm ] Ah, good. I also noticed one case where MRI still relies on HOME: irb

[Ruby 1.8.7 - Bug #213] (Open) Different ERB behavior across versions
17491 [redmine@ru y] Issue #213 has been reported by Federico Builes.

[Ruby 1.8 - Bug #212] Issues with Readline in Mac OS X
17493 [redmine@ru y] Issue #212 has been updated by Roger Pack.

[Ruby 1.8 - Bug #212] Issues with Readline in Mac OS X
17494 [redmine@ru y] Issue #212 has been updated by Federico Builes.

[Ruby 1.8 - Bug #211] "wrong argument type" "expected Proc"
17495 [redmine@ru y] Issue #211 has been updated by Shyouhei Urabe.

[Ruby 1.8 - Bug #212] Issues with Readline in Mac OS X
17497 [redmine@ru y] Issue #212 has been updated by Luis Lavena.

We'll release 1.8.6/1.8.7 this Friday
17499 [shyouhei@ru ] to release them again. Expected date for that is Jul. 4 (in Japanese
+ 17502 [znmeb@ce ma ] How do we get these exact versions? Are there tarballs, or do I need
+ 17504 [vsizikov@gm ] * 1.8.6 HEAD has 7 rubyspec falures for StringIO
| + 17506 [charles.nutt] There should not be a release until the failures in rubyspec are addressed.
| | + 17521 [shyouhei@ru ] OK, so all of you can wait until rubyspec passes?  I think I cannot
| | | + 17522 [federico.bui] We'd appreciate it if you could hold on a bit. Most of the current
| | | + 17528 [vsizikov@gm ] It seems to me that the majority of String and StringIO failures is
| | | + 17538 [hramrach@ce ] While it is nice to have a release on the planned date I find the
| | | + 17543 [charles.nutt] Others have already responded, but I think it's of paramount importance
| | | + 17544 [igal@pr gm t] Please delay the release. I don't think we'll be able to resolve the
| | |   17545 [charles.nutt] For what it's worth, this is all a strong argument for having a CI
| | |   + 17546 [igal@pr gm t] Absolutely.
| | |   | + 17551 [phlip2005@gm] CruiseControl.rb doesn't suck.
| | |   | | 17554 [igal@pr gm t] Great, thanks for the recommendation. The manual
| | |   | + 17552 [wilsonb@gm i] Ryan Davis is working on a system for this right now. Last I heard, it
| | |   |   17558 [znmeb@ce ma ] Glad to hear it. I've got some spare cycles (day job starts again 8
| | |   |   17560 [hramrach@ce ] Setting up VMs for Windows or some BSDs should be easy. However, OS X
| | |   |   + 17562 [znmeb@ce ma ] ...
| | |   |   + 17569 [igal@pr gm t] Running OS X in a VM works well, although it's a rather slow and the
| | |   |     17579 [charles.nutt] Honestly I'd almost like to see Apple step up to the plate here. They're
| | |   |     17623 [laurent.sans] Sorry for the late reply, I was discussing with my management.
| | |   + 17559 [znmeb@ce ma ] *And*, some of us who don't have open source day jobs have a long
| | |     17563 [charles.nutt] I've heard of those before.
| | |     17564 [phlip2005@gm] Sez you! At my current day-job, we use Rails via assert{ 2.0 } and assert_xpath.
| | |     17567 [charles.nutt] I meant I've heard of weekends :) I don't really take many "off" days
| | |     17575 [phlip2005@gm] Did I mention we also do pure XP, and have a strictly 7-4 work day with no
| | + 17806 [hramrach@ce ] Could somebody, please, summarize the current status?
| |   17851 [akr@fs j. rg] RubySpec hangs.
| |   17852 [federico.bui] What version of mspec are you using? 1.4.0 was recently released and
| |   + 17853 [akr@fs j. rg] I use -V -f s.
| |   + 17855 [onepoint@st ] I'm not the original poster but I'm seeing hangs too.  If I run
| |     17857 [federico.bui] Tanaka, Jeremy: Thanks for the reports. This particular spec has been
| |     + 17858 [onepoint@st ] Yes, pulling the latest ruby-spec removes the hang.
| |     + 17860 [onepoint@st ] That's  exactly what  I  get now,  except  that I  don't  see the  two
| |       + 17892 [federico.bui] Yup, the value was not big enough for a 64 bit machine but it's been
| |       + 17939 [ks@ku ts ep ] When will we see a new 1.8.6 release?
| |         + 17940 [nobu@ru y- a] Vladimir and Charles stopped Shyouhei from releasing it due to
| |         | 17941 [vsizikov@gm ] The 1.8.6 and 1.8.7 branches have been OK w.r.t. RubySpecs for a while
| |         | 17945 [onepoint@st ] I'm not sure what you mean  by "OK", but I'm still getting errors with
| |         | 17946 [onepoint@st ] To compare,  with the  latest ruby_1_8_6 branch  I get the  same Iconv
| |         | 17947 [federico.bui] Jeremy,
| |         | + 17948 [nobu@ru y- a] SUS defines nothing about the conversion detail, nor encoding
| |         | | 17953 [dev@me ne .c] ...
| |         | | 17955 [federico.bui] Thanks Jeremy, Nobuyoshi and Daniel. Following Daniel's advice I've
| |         | | 17960 [shugo@ru y- ] Currently ruby_1_8_6 has 2 rubyspec failures.
| |         | | + 17965 [onepoint@st ] ... and ruby_1_8_7 has exactly the same failures.
| |         | | | 17968 [vsizikov@gm ] Yes, there was a change in RubySpecs yesterday, that triggers these
| |         | | | 17969 [onepoint@st ] Are there  remote branches  that we could  track to follow  the stable
| |         | | + 17971 [federico.bui] Yes Shugo, that was caused by my commit but as Vladimir said it's a
| |         | |   17992 [shugo@ru y- ] Thank you.
| |         | |   18062 [ks@ku ts ep ] When can we expect a release?
| |         | + 17964 [onepoint@st ] That's fine; I'm  not claiming these are critical  errors, just trying
| |         + 17942 [shyouhei@ru ] When this ML thinks I should.
| + 17532 [schreiber.ar] Most of these 57 rubyspec failures are caused by the convert_type
|   17539 [shyouhei@ru ] Thank you, I merged this revision into 1.8.7.
|   17540 [akr@fs j. rg] It causes following problem.
|   17541 [schreiber.ar] That problem seems to only exist in 1.8.6, r16659 seems to have been
|   17542 [akr@fs j. rg] You are right.  r16659 is for 1.8.6.
+ 17508 [igal@pr gm t] I think so too. However, there are still some problems that may be
+ 17510 [igal@pr gm t] memory leaks.
+ 17515 [vsizikov@gm ] I'd strongly suggest to postpone 1.8.7 release untill te majority of
| 17524 [federico.bui] * Iconv
| 17525 [akr@fs j. rg] What the revision or patchlevel of the "HEAD"?
| 17527 [federico.bui] You're right, I was testing against 1.8.6p255 but p264 seems to fix
| 17529 [akr@fs j. rg] Note that my feeling is not the decision of ruby.
+ 17517 [luislavena@g] I'm not happy to inform that the make test-all procedure still stall
  17526 [usa@ga ba ec] I can't reproduce it with mswin32...
  17530 [luislavena@g] The bad thing is I don't have my VC6 environment to reproduce it
  17533 [usa@ga ba ec] Oh, sorry, I'd mistaken.
  + 17534 [gdefty@at gl] Tim,
  + 17535 [usa@ga ba ec] Now I've decided to revert r17290 from ruby_1_8_6 in consultation
  + 17536 [luislavena@g] Done, the test stall.
    17537 [usa@ga ba ec] Thank you!
threads.html
top