7169-7394 subjects 7585-12269

^ minor patch for fileutils.rb (fu_world_writeable?)
7383 [Daniel.Berge] Ruby 1.8.4
7385 [matz ruby-la] Yes, thank you.
7386 [djberg96 yah] Thanks.  I should add that I'm not trying to be

^ Latest change to rational.rb
7395 [ksruby gmail] Here's a small fix to

^ IO#reopen
7396 [matju arteng] IO#reopen doesn't flush anymore before reopening. Is that normal??
7397 [akr m17n.org] Hm. ruby-1.9 doesn't flush if an argument is a filename.
7398 [matju arteng] Yes. In particular I use -D 2006-02-01 (for reasons previously explained)
7399 [akr m17n.org] Fixed.  Thank you for the report.
7400 [matju arteng] Wow, thanks! that was fast!

^ Symbols overlap ordinary objects, especially on OS X (Was: Re: Problem with weak references on OS X 10.3)
7401 [drbrain segm] OS X's malloc aligns memory on 16 byte boundaries.  This problem is
+ 7412 [drbrain segm] $ cat overlap.rb
+ 7414 [ mfp acm.org] I was being stupid. I forgot that sizeof(RVALUE) % 8 = 4, so the
  + 7416 [ mfp acm.org] Not that it really matters, anyway String#intern stops working after
  + 7427 [matz ruby-la] Since all symbol values have their second LSB set, they should not
    + 7428 [akr m17n.org] It's true in VALUE space but not in object_id space.
    | 7444 [nobu ruby-la] Two thoughts;
    | 7445 [akr m17n.org] c) align objects 20bytes.  map symbols to non-20bytes aligned value.
    | 7447 [akr m17n.org] ===================================================================
    | 7448 [akr m17n.org] Oops.  The original pointer should be keeped to call free.
    | 7465 [evanwebb gma] Just my quick 2 cents...
    | + 7466 [kevin.jackso] Both points are excellently made - especially about the fact that the
    | + 7468 [akr m17n.org] To be fair, no one propose to change the VALUE meaning.
    |   7469 [nobu ruby-la] rb_obj_id() is used as Kernel#hash also.
    |   7471 [akr m17n.org] Hm.  Maybe.
    |   + 7472 [ocean m2.ccs] Sorry for interruption.
    |   | 7474 [matz ruby-la] Sounds nice.  Can you check in, surrounding pragma in the cpp
    |   + 7473 [matz ruby-la] Could you commit this patch, please?
    |     7475 [ryand-ruby z] (and merge to 1.8 please please please please please)
    |     7477 [akr m17n.org] It's committed.
    + 7429 [ mfp acm.org] This solves the crash but doesn't prevent collisions, does it? Normal
    | 7430 [ mfp acm.org] ====
    | 7432 [akr m17n.org] 0x1ff, I think.
    + 7431 [ryand-ruby z] Doesn't fix on OSX. I think Akira's email summarizes what we are

^ Vacation notice (was: Symbols overlap ordinary objects, especially on OS X (Was: Re: Problem with weak references on OS X 10.3))
7402 [wonado donne] Liebe(r) Absender(in),

^ Module#define_method "send hack" fails with Ruby 1.9
7403 [emiel rednod] In trying to get rubygems working with 1.9 I ran accross an error
+ 7404 [g_ogata optu] In current 1.9, #send can't call private methods (unless called
| 7406 [dblack wobbl] I'm still not happy about #funcall.  How is one supposed to remember
| + 7407 [matju arteng] me too. i downvote #funcall. besides, is there a #__funcall__ too?
| | 7409 [tjw omnigrou] I care about private methods actually being private.  One example;
| | 7410 [ara.t.howard] i know what you're driving at - but i used to do alot of work with the java 2d
| + 7408 [meta pobox.c] Presumably both are documented, with their guaranteed behavior specified?
| | 7415 [dblack wobbl] Yes, of course; but a lot of things in Ruby that are documented are
| + 7411 [g_ogata optu] #funcall breaks the rules.  That makes it "fun".  ;-)
|   + 7413 [jgbailey gma] Second! I've disliked 'funcall' since I first saw it. And why break send? I
|   + 7417 [halostatue g] #safe_send/#__safe_send__ ?
+ 7405 [ mfp acm.org] class A

^ Fwd: Launching Ruby scripts and the future of MVM
7418 [headius gmai] to launch Ruby scripts in a platform-independent way. I'm sure

^ ping.rb
7419 [rubypate gma] ping.rb has bothered me for a while for two reasons.  First, it really ough=
+ 7420 [djberg96 yah] No ICMP ping yet, though, mainly because there's
| 7421 [rubypate gma] *sigh*   /me should have checked harder.
+ 7425 [vjoel path.b] It would be good to have all the pings in one ruby basket.

^ [ ruby-Bugs-3654 ] Syck fails with parse error on parsing multiple documents
7422 [noreply ruby] Bugs item #3654, was opened at 2006-02-24 15:45

^ [PATCH] Rubygems/ri integration
7423 [drbrain segm] charset=US-ASCII;
+ 7426 [ryand-ruby z] Please approve. I'd like to merge this with 1.8 as well. This will be
+ 7479 [matz ruby-la] Hmm, it seems not to harm anything.  OK.  I will merge it.

^ [ ruby-Bugs-3658 ] Passing wrong data type to Win32API#call causes segfault
7424 [noreply ruby] Bugs item #3658, was opened at 2006-02-24 19:33

^ [PATCH] Re: Symbols overlap ordinary objects, especially on OS X (Was: Re: Problem with weak references on OS X 10.3)
7433 [ mfp acm.org] Of course, but a second self-reply would have been too much noise.
7443 [ryand-ruby z] I can attest to its ugliness, but not its correctness. Akira, Matz,

^ parser bug, function call without parens across multiple lines
7434 [dsmiley mitr] I have encountered a counter-intuitive bug in Ruby's parser.  The
+ 7435 [decoux moulo] Well, with
+ 7436 [ mfp acm.org] ==============
+ 7437 [langstefan g] It's not the difference between one and multiple lines.
| 7438 [yemi weldfas] When I try code like the following in (ruby 1.8.4 (2005-12-24)
+ 7441 [dsmiley mitr] First of all, I'd like to apologize because my bug report was

^ FYI: ruby-lang.org is on spamcop blacklists
7439 [meta pobox.c] dnsbl/bl.spamcop.net returned deny: for
7440 [jgbailey gma] This is fairly serious as it breaks Google Groups/Ruby Talk list
7464 [halostatue g] It's serious but stupid. I don't know why anyone still pays attentionto spamcop. They're wrong as often as they're right, and by timethey're right, they're usually out of date.

^ GC Question
7442 [zdennis mkte] I have been posting to the ruby-talk mailing list about ruby memory and GC, and I think it's ready
7446 [ocean m2.ccs] Because ruby's GC is conservative. Python uses reference count based GC, so it is guranteed that unreferenced
+ 7449 [ocean m2.ccs] I found interesting phenomenon.
| + 7450 [ocean m2.ccs] Sorry, I lied...
| + 7451 [decoux moulo] Yes this is what it do. Now try it with 1.6.8 and you'll have a
|   7452 [ocean m2.ccs] (2006/02/28 19:35)
|   7453 [decoux moulo] Well, here the example given with 1.6.8
|   7454 [ocean m2.ccs] Hmm, on my machine,
|   7455 [decoux moulo] Well, I don't remember very well but I think that Boehm GC has
+ 7456 [zdennis mkte] zdennis@lima:~/source/downloads/ruby-1.8.4$ patch array.c array.patch
| + 7457 [ocean m2.ccs] I'm using this.
| + 7458 [decoux moulo] Well here a debugging session if you want to try to understand what it do
|   7459 [zdennis mkte] Thank you Guy for posting this with your comments. So this is a bug I am taking it, is this
|   7461 [decoux moulo] Well this is not really a bug, this is how work a conservative GC which
|   7463 [zdennis mkte] From some tests, it appears that ruby will taper off memory usage and reuse unused slots in the
|   7467 [decoux moulo] You have found this , I think (this is the Boehm GC)
+ 7462 [zdennis mkte] zdennis@lima:~/source/languages/ruby$ ~/bin/ruby-1.8.4/bin/ruby gc_mem_testing.rb

^ [ ruby-Bugs-3683 ] rails segfaults at 2nd load with eval.c: No such file or directory.
7460 [noreply ruby] Bugs item #3683, was opened at 2006-02-28 10:33

^ [ ruby-Bugs-3698 ] In a highly dynamic ObjectSpace (loads of created and collected objects) YAML creates erroneous anchors/references
7470 [noreply ruby] Bugs item #3698, was opened at 2006-03-01 19:10

^ Net::HTTP Bug in Ruby 1.8.4?
7476 [james graypr] [ruby-talk:182360]
7482 [james graypr] Well, I've looked into this some more myself.  I do believe this is a
+ 7483 [james graypr] if content_type == "/"
+ 7484 [aamine lover] Thank you very much.

^ Re: [Ruby 1.8.4] Bug report for File.expand_path
7478 [matz ruby-la] I'm afraid it's intended behavior on the DOS derived operating
7481 [usa garbagec] Matz, you are right.

^ (none)
7480 [hasimoto s3.] # chaddr hasimoto@s3.dion.ne.jp hasimoto.kazuhito@gmail.com

^ Bugzilla for ruby?
7485 [hadmut danis] I didn't see any bug management database on the ruby web site.
7486 [matz ruby-la] This is the one I use (note: it's in Japanese).
+ 7487 [meadori gmai] Is the bug tracker at RubyForge (
| 7488 [matz ruby-la] All bugs submitted to the tracker would be registered to the list
+ 7489 [hadmut danis] Sorry, but my knowledge in Japanese is limited to about three
| 7490 [drbrain segm] Yes, it is used.  All reports get sent here to this list.
+ 7491 [kapheine gma] Roundup[1] is pretty nice for cases like these.  You can setup an

^ break semantics from lambdas and methods
7492 [w.kelly qut.] I'm having trouble understanding the logic behind the semantics
7520 [matju arteng] I'm also having that trouble. It's something that is much easier to

^ how to introduce reference objects into ruby
7493 [Geert.Fannes] charset="us-ascii"
+ 7494 [evanwebb gma] Geert,
| 7496 [matju arteng] There aren't that many mis-features: mostly it's just good features that
| 7497 [Geert.Fannes] some explanation of the things I'm trying todo and why I think Ruby as
| + 7498 [Geert.Fannes] 2. It's also possible to construct things that are pointers to variables
| | 7513 [matju arteng] Oh, no, sorry, I don't mean those things as permanent fixtures. It'd so
| + 7499 [hgs dmu.ac.u] # Untested code.
| | + 7500 [Geert.Fannes] The problem with the code you sent is that you have to go through ALL
| | | + 7501 [caleb aei-te] Why not just write a Ruby wrapper around you C extension?
| | | | 7503 [Geert.Fannes] charset="iso-8859-1"
| | | + 7502 [evan falling] See below.
| | | | 7505 [Geert.Fannes] charset="iso-8859-1"
| | | + 7504 [jim weirichh] [... example code elided ...]
| | | | 7506 [Geert.Fannes] charset="iso-8859-1"
| | | | 7507 [jim weirichh] I think you are saying that the rows of your matrix contain the value,
| | | | + 7511 [headius head] If I am understanding correctly, the columns would contain the actual
| | | | | + 7512 [headius head] If you wanted to be even trickier, you could just use [i,j] as the key and
| | | | | + 7514 [headius head] Gah, ok, so I made a small mistake.  That should be @row_number..
| | | | | + 7534 [Geert.Fannes] charset="iso-8859-1"
| | | | + 7535 [Geert.Fannes] charset="iso-8859-1"
| | | + 7537 [meta pobox.c] Well, consider that in a perfect world with zero storage overhead, you'd
| | |   7538 [Geert.Fannes] charset="iso-8859-1"
| | |   7545 [meta pobox.c] Well, I generally prefer an ACID relational database, but I gather that
| | |   7546 [Geert.Fannes] In Ruby, there's the []= and [] operators which you can define together.
| | |   + 7547 [ara.t.howard] this seems like so much work - why not simply use narray - it does nearly all
| | |   | 7548 [matju arteng] Really? How do I do sparse arrays in NArray ???
| | |   + 7556 [matju arteng] In Ruby, a Hash is really a hash and so has lookup in O(1) (statistical
| | + 7510 [matju arteng] If the dim of the matrix is less than 32768 by 32768, replace the [i,j]
| + 7509 [matju arteng] yeah, that amounts to a consider the matrix as the adjacency matrix of a
|   + 7516 [ara.t.howard] have you used narray for any of it?
|   | 7517 [matju arteng] No, I originally considered NArray as a base for building a system I now
|   + 7533 [Geert.Fannes] charset="iso-8859-1"
|     7539 [matju arteng] Ok, I haven't tried it, and doing it efficiently enough might mean using a
|     7540 [matju arteng] i mean the probability for _each_ possible edge to exist, considering each
|     7541 [yemi weldfas] It is funny how that exact method definition was already proposed for making a sparse matrix earlier that was turned down. Maybe a load test would be good just too see if won't do as needed. Some remarks feel like too much theory and not enough practice. I am sure you are all smart enough to optimize AFTER just getting things to work right. Good day.
+ 7495 [matju arteng] Fixnum involves no memory allocation whatsoever, so the concept of

^ [PATCH] small clarification for ruby man page
7508 [dlutter redh] The attached patch (against latest CVS) clarifies that directories from
7515 [matz ruby-la] It will be merged.  Thank you.

^ Proposal: String#notempty?
7518 [lists bertra] I oftenly use `Numeric#nonzero?'. Actually, I miss its
+ 7519 [evan falling] How is this better than doing "not str.empty?".
| 7521 [lists bertra] nickname.notempty? or forename
+ 7522 [nobu ruby-la] Isn't nonempty but notempty?
| 7523 [lists bertra] It should probably be, for consistency.
| 7529 [Daniel.Berge] unless string.empty?
| 7536 [lists bertra] How do you make the unless modifier return a string?
| 7544 [nobu ruby-la] string[/.+/mn] || "default"
| 7549 [lists bertra] puts RUBY_VERSION
| 7550 [dblack wobbl] I see the point about speed, but I'm not clear on what the
| 7552 [lists bertra] I just looked for an amusing way to express that "nonempty?"
+ 7542 [g_ogata optu] foo.bar.baz.blib.blob.blah.!empty?
  7543 [lists bertra] Sorry. I apologize for having caused the inconvenience.

^ Sefe level: bug or feature?
7524 [k.shutemov s] Why cannot do eval with $SAFE=3D3 and can with $SAFE=3D4? Is it bug or
7525 [matz ruby-la] Feature.  All dangerous operation should have been disable at level 4.
+ 7526 [k.shutemov s] Tnx. May tainaed strings be interpreted on the same reason?
+ 7527 [matju arteng] Why would eval still be dangerous if all other dangerous operations have
  + 7528 [chneukirchen] Denial of Service.
  | 7530 [matju arteng] How do you make an application eval your code if that application doesn't
  + 7531 [matz ruby-la] No reason.  It's allowed at level 4, since it's no longer dangerous.
    7532 [chneukirchen] Lession learned: This is not enough for a IRC bot. (hi, flgr :P)
    7562 [florgro gmai] Indeed. A way to limit a thread's memory usage would still be a

^ [ ruby-Bugs-3842 ] xmlrpc client and php xmlrpc servers
7551 [noreply ruby] Bugs item #3842, was opened at 2006-03-15 23:45

^ [ ruby-Bugs-3843 ] "not" operator used in expression that is a method parameter can generate syntax error
7553 [noreply ruby] Bugs item #3843, was opened at 2006-03-15 22:09
7554 [nobu ruby-la] Not a bug.
+ 7555 [matju arteng] Why is there a distinction between statements and expressions? Ruby is so
+ 7557 [shyouhei ice] Nobu, you are not answering to the question.... You have to unveil why
  7559 [matz ruby-la] OK.  Keyword logical operators (and, or, not) are far lower precedence
  7565 [meta pobox.c] I still don't understand why foo(!(1<2)) is OK, but foo(not(1<2)) isn't.
  7566 [binary42 gma] foo(not(1<2), 3)
  7567 [meta pobox.c] Except comma isn't an operator according to the Pickaxe. Could be a
  7568 [binary42 gma] Comma is not an operator, strictly speaking, but when parsing code,
  7582 [robert.dober] Dear all
  + 7586 [evan falling] What you've described is the basic predence difference between
  | + 7587 [vjoel path.b] Actually, it is "&" that can be defined as a method. You are free define
  | | 7588 [evan falling] Huh. Correct you are. In addition, def won't accept && as a method
  | | 7592 [robert.dober] Good thaughts and explanations, I really like to descend to that level of
  | + 7597 [meta pobox.c] ...
  + 7591 [now bitwi.se] Is that perhaps due to the fact that you're mixing the

^ RCR 320: Extend Rational to interoperate better with floats
7558 [dave burt.id] Last October I worked on an extension to the standard Rational library

^ [OT] Rant about keyword logical operators was : (Re: [ ruby-Bugs-3843 ] "not" operator used in expression that is a method parameter can generate syntax error)
7560 [rubyzbibd ub] Here is a small rant about keyword logical operators and usage at my
+ 7561 [now bitwi.se] I'm probably missing something vital here, but why would changing it
| 7563 [rubyzbibd ub] I am embarrassed to say, but in the past we have had situations where
| 7564 [james graypr] This is proper version control usage anyway, in my opinion.  You
| 7573 [rubyzbibd ub] Hehe, like I said it embarrasses me to say we even had this problem,
+ 7569 [vjoel path.b] x = h && h[4]
  7570 [wilsonb gmai] Maybe we just need to rename 'and' to 'and_then' or 'also', then.
  + 7572 [rubyzbibd ub] That would certainly make me happy!  Although, Joel's above examples
  + 7574 [meta pobox.c] What about 'or' and 'not'? Is there a simple renaming to make their
    7575 [jim weirichh] "or_else" works for "or" ("and then" and "or else" are the short circuit
    7576 [dave burt.id] "true_unless"... or "true unless" for that matter.

^ The latest ruby 1.8 CVS breaks openssl compilation on win32
7571 [ville.mattil] This patch seems to fix it.

^ BigDecimal#to_d
7577 [dave burt.id] BigDecimal has no to_d method, although it adds conversion methods for other

^ [ ruby-Bugs-3856 ] strange Kernel#exec behavior with bash's source command
7578 [noreply ruby] Bugs item #3856, was opened at 2006-03-17 17:49
8402 [nobu ruby-la] Not strange.

^ possible bug in parse.y?
7579 [ryand-ruby z] I don't have a repro on this, but I saw it while debugging some of my
7580 [matz ruby-la] The return value of rb_intern() is not used here.  I just wanted to

^ [RCR] Kernel#executed_directly?
7581 [florgro gmai] def executed_directly?()
7583 [dblack wobbl] This message is in MIME format.  The first part should be readable text,

^ [ ruby-Bugs-3866 ] [PATCH] rdoc for String#=~
7584 [noreply ruby] Bugs item #3866, was opened at 2006-03-19 14:18