20979-21712 subjects 21387-21842

What's the best book
21212 [ jmp@no s. r] .... to learn Ruby.
+ 21213 [Stephan.Kaem] The 'pick axe book' also known as
| 21219 [markb1@sy pa] It's online at http://www.rubycentral.com/book/index.html
| 21220 [mrchameleon@] .... but buy a hard copy anyway; it's excellent and the authors deserve our support.
+ 21229 [ryan@ry nf y] ...

Call for discussion for a french speaking ruby group
21215 [Erwan.David@] A call for discussion has just been posted for the creation of

21216 [bobh@hs da o] Any problems getting Ruby to compile on MacOS X? Did you have to add
21230 [lcb134@ps .e] Check out the package at www.entropy.ch/software.  Its a very elegant
22390 [K.Hodges@ft ] ...
22394 [johann@ph si] If anyone's interested, I have packaged Ruby into an OS X framework,
22400 [rich@in oe h] I am interested...where is it?
22470 [johann@ph si] I'm sorry, I tried to get it ready yesterday and failed.

ANN: ns_ruby-0.2.1
21217 [toddg@li ux ] ns_ruby[0], the project to interpret Ruby code within Aolserver[1] has

Ruby objects <-> XML: anyone working on this?
21218 [senderista@h] Are there any Ruby analogs of these two Python modules (xml_pickle,
+ 21255 [maki@op n- e] How about XMarshal?
| + 21264 [green@gr en ] Hey, that's a very cool extension!  For what it's worth, I took a shot at
| + 21275 [senderista@h] Very cool!  I'll have to check this out.  I wonder about the second
+ 21277 [kentda@st d.] I've never been too keen on the format provided by xml_pickle and
  21281 [neumann@s- i] Try SOAP or XML-RPC for marshalling Ruby objects in a language-independent
  21287 [kentda@st d.] I looked at these some months ago, but I had a new look now on SOAP.
  + 21297 [senderista@h] You're saying that SOAP doesn't properly serialize arbitrary graphs of
  | 21319 [kentda@st d.] I'm saying I don't know to do it :-)
  + 21299 [nahi@ke na t] # MN: Thank you for introducing SOAP4R, too.
  | + 21301 [paulp@Ac iv ] We are discussion how to use SOAP and especially WSDL from various
  | + 21304 [nahi@ke na t] I'm very sorry for misunderstanding.
  | + 21320 [kentda@st d.] The result when I serialize an object to file/string/IO.
  + 21308 [neumann@s- i] Yes, the object will be stored twice.

Call Proc in different context
21223 [anders@ig it] Object.instance_eval is useful for evaluation 'raw' source and blocks in
21222 [rtarpine@ho ] I just asked the same basic question within the past week or so, actually.

Redefining constants
21224 [johann@ph si] How do you redefine constants, such as class objects, at the top
21228 [decoux@mo lo] Well, you're trying to change the class hierarchy. See [ruby-talk:20766]
21243 [johann@ph si] Thanks for the reference.  So the answer is that I can't?  This is
+ 21244 [avi@be a4 co] If you change
| + 21249 [elderburn@mi] The answer is you -shouldn't- be able to, per definition of
| + 21285 [pbrannan@at ] Except that if you will still get a warning about "already initialized
+ 21258 [decoux@mo lo] class A < Array

Class of caller
21225 [rtarpine@ho ] Is there a way to get the class of the object that called a certain method?
21226 [decoux@mo lo] There is a RCR to add #call_stack, perhaps it will do what you want. See

21227 [dempsejn@ge ] I've read through some manuals and references, and come from a

Ruby/GTK maintainer?
21234 [nconway@kl m] I mentioned on the list a couple days ago that the Ruby/GTK maintainer
+ 21240 [bombadil@wa ] have been not able of contacting him more.
+ 21256 [maki@op n- e] Did you send a mail to him?  He may not read ruby-talk ML.
| 21267 [nconway@kl m] Yeah -- I emailed him 2 patches a while ago and he never responded.
+ 21283 [michal@cr bo] How about to fork project or start from the scatch work on Gnome
| + 21295 [nconway@kl m] Yep, definately (not the 'fork the project' part -- but bindings for the
| | 21462 [erik@ba fo s] Who has this code?  Can I get my hands on it??  It would really help me
| | 21467 [nconway@kl m] The prototype is @
| + 21463 [erik@ba fo s] I started yesterday to make a defs-parser for ruby (in ruby).
+ 21537 [iga@ru y- an] Sorry for my absence for a long time.

Broken link  in RAA
21235 [hal9000@hy e] The *very first* link in the RAA (MindRuby)
21278 [matz@ru y- a] If you guys want to remove RAA entry, mail me.
21310 [neumann@s- i] Could you remove my entries
21373 [matz@ru y- a] Done.  Thank you.

alternate Download-Site of IOWA?
21236 [horibo@gm .d] I tried to download iowa several times within the last couple of days.
21237 [avi@be a4 co] That's strange - beta4.com is up for me (I would have a hard time writing
21238 [mps@di co sy] You  know, this might be due to communications problems associated with
21239 [toddg@ma l. ] I have had DNS troubles from time to time connecting to beta4.com.  That
21262 [horibo@gm .d] It was no DNS-problem. The connection timed out. But now it's back
21268 [jfitzell@ho ] Yeah, it would wouldn't it?  :)
21269 [toddg@ma l. ] I used to carefully design sites.  Now I just throw out whatever I have on
21270 [hal9000@hy e] Now *there's* a design philosophy. Reminds me of the "worse is better"
21272 [toddg@ma l. ] It works, too.  Keeps one on the straight and narrow path of user

Creating a module hierarchy - advice needed.
21242 [ptkwt@sh ll ] I'm trying to create a library of Win32 functions (most come from C
+ 21260 [decoux@mo lo] Yes, what you want is in the ToDo
+ 21273 [theschof@cs ] Win32 modules)

Embedding Q's
21245 [johann@ph si] Is there a good primer anywhere for embedding Ruby?  The only document
+ 21246 [toddg@ma l. ] No primer, sorry.  I recommend looking at other embedded projects to get
| 21254 [johann@ph si] First off, thanks for the example.  That at least gives me something
+ 21252 [ned@bi e- om] You might like to look at the Vim code for embedding Ruby. Actually, looking

Instance variables of an objecta
21247 [jweirich@on ] Given an object, is there a way of getting a list of the instance
21248 [green@Fr eB ] irb(main):002:0> o = Object.new
21253 [jweirich@on ] Thanks.  For some reason I couldn't find it.

request for help
21250 [theschof@cs ] For a few days now I've been toying with a pet project I came up with when I

Ruby Blend File
21251 [theschof@cs ] Sorry in my earlier post I forgot to actually include Ruby Blend in it, here it

21257 [hal9000@hy e] Networking question for you.
21259 [bdelmee@ad a] Well, I have never programmed anything network-related in ruby,
+ 21271 [hal9000@hy e] THanks for replying. Comments below.  HF
+ 21280 [m_seki@mv .b] ## gossip.rb

Exceptions and Ruby
21263 [alv@pr v. ne] Ruby 1.6.4
+ 21265 [decoux@mo lo] You can use #Integer, #Float
+ 21266 [r2d2@ac .u u] [Alva]
  + 21279 [alv@pr v. ne] Ugh, you got me. What does the operator ||= do?
  + 21286 [pbrannan@at ] Don't do this.  It is very likely to break something.

tree class library?
21274 [DELETEmitche] Aside from Yoki Toshinori's binary tree lib, does anyone have native ruby code for trees? Specifically, I need rooted,
21290 [vjoel@pa h. ] As luck would have it, I just posted v1.3 of Enumerable Tools on the RAA,

Anyone else working with ruby on OS X (or GnuStep?)?
21276 [johann@ph si] Like I said, is anyone out there?  Has any progress been made?

extensions libraries
21282 [michal@cr bo] I'd like to ask you for comments.
21284 [decoux@mo lo] not really a problem.

SV:  Re: Exceptions and Ruby
21288 [hamselv@da s] A much better alternative is to make a new class. ie.

CFV Pointer: comp.software.patterns
21289 [dave@te hn p] Voting is now in progress for the proposed newsgroup

Image libraries
21291 [jjthrash@po ] Is anybody working on Ruby bindings for libpng or libgif?
21309 [neumann@s- i] Some days ago I wrapped the image load functions of Imlib2 (www.enlightment.org/pages/imlib2.html)
21311 [jjthrash@po ] Does loading images mean having an array of some sort in memory that
+ 21313 [neumann@s- i] load textures for OpenGL.
+ 21327 [nconway@kl m] I looked at doing libpng + libjpeg a little while ago but I got confused
  21432 [jjthrash@po ] Well, I have an initial PNG wrapper that can 1) tell you if a file is
  21471 [neumann@s- i] Maybe a general Ruby/Imaging project (could evolve to something like Python's image library)
  21490 [jjthrash@po ] Sounds good to me.  Shall I open a project?  "Ruby/Imaging"?
  21524 [neumann@s- i] Yes, would be nice.
  21526 [aleksei.guze] Please, spend some time on developing the imaging library INTERFACE, not
  21544 [tobiasreif@p] SVG is cool too.

PC Magazine
21292 [rtarpine@ho ] In the October issue of PC Magazine, in his Inside Track column John C.
+ 21293 [toddg@ma l. ] Is 'publicating' a word?
| 21294 [rtarpine@ho ] Aha, caught my little slip there.  I meant 'self-generating,' of course.
+ 21314 [ptkwt@sh ll ] I think this is excellent publicity (well, I must admit that the
+ 21326 [ruby@ja es r] Well, Oscar Wilde said that the only thing worse then having people

nested require files need path internally
21296 [bobgus@mc .c] I had a piece of a program involving date manipulation. This piece started with
21298 [dblack@ca dl] 'require' will look in the order specified in the $: global variable
21302 [bobgus@mc .c] Thanks much - at least there is some logic.
21303 [matz@ru y- a] Too much priority for current directory in the load path sometimes
21306 [larsch@cs au] Should be this way, but I think it would be nice to have a way, similar of
21307 [matz@ru y- a] The big difference is C++ search done in compile time, Ruby search
21331 [pbrannan@at ] The problem with doing this is that if I have a directory structure like
+ 21337 [Dave@Pr gm t] That's because you're giving it a different name each time. The first
+ 21340 [matz@ru y- a] I'm not sure what is your point.  If you really want to read from the
  21353 [pbrannan@at ] The idea is that I want to split a project between multiple files.
  21366 [matz@ru y- a] I think we are talking different issues.  You're talking about a
  21368 [julian-ml@be] But, I think what he is looking for is the ability to require a file from the same directory as the file that does the require, not from the current directory.  I have experienced this problem too...
  + 21369 [rich@in oe h] This seems like a good argument.  In Java, package (namespace) names directly align with directories.  This is not the case with Ruby namespaces (modules/classes).  In Java, a Class does not need to import another Class in the same package (directory).  In Ruby...every additional file you want to use must be loaded/required.  I do understand the extreme difference between the two, but the ability to "require" a ruby file in the same directory as the requesting file seems very good from a library perspective (and I also agree that you need to protect the namespace from the CWD).
  + 21376 [matz@ru y- a] Your statement is understandable.  I think it's trade off between
    + 21401 [kevinbsmith@] That's ok, if we can come up with a simple pattern to allow
    | 21402 [decoux@mo lo] See [ruby-talk:21303]
    + 21406 [pbrannan@at ] I definitely prefer simplicity.
      + 21426 [matz@ru y- a] Every developer who works on multi file program has to decide where to
      | + 21433 [pbrannan@at ] Why is this bad?  It's extraordinarily useful to keep files in the same
      | | 21436 [matz@ru y- a] Because it's against UNIX tradition.  Directories such as /usr/bin are
      | | 21440 [rich@in oe h] ...
      | + 21477 [kevinbsmith@] Now I understand. In the Windows world (where I've lived
      | + 21493 [charleshixsn] This seems to me to be the most preferable choice for a user
      + 21427 [julian-ml@be] I was playing with this method too... definitely seems the best way but I was also unhappy with the fact that it requires parsing text that is destined for humans to read...  I'm planning on implementing this and overriding require (maybe with an old_require alias) in my own projects if nothing else turns up

MetaRuby 0.7
21300 [matju@sy pa ] MetaRuby 0.7

eruby -M on Win98
21305 [hal9000@hy e] I've been playing with eruby and I've found

Name clashes
21312 [alv@pr v. ne] I'm writing a small game using the RUNIT framework. I use

Suggestions for new CGI lib
21315 [anders@jo an] From the comp.lang.ruby thread "Minor cgi.rb question" (2001-09-03), I
+ 21323 [larsch@cs au] I think this looks good. Writing short CGI scripts will become a bit more
+ 21324 [matz@ru y- a] I think they are great.  Here's some questions from me.
| + 21325 [kh.wild@wi o] imho it should be possible to select the values in the shown way.
| | + 21334 [toddg@ma l. ] IMHO, this is dumb b/c a request can only be instantiated with either POST
| | | + 21342 [niklas@ka i.] While it is true that the request method in HTTP is either POST or
| | | | 21343 [toddg@ma l. ] Then that violates the spec.
| | | | 21345 [niklas@ka i.] [Niklas Frykholm]
| | | + 21382 [anders@jo an] POST /script.cgi?user_var=value HTTP/1.1
| | | + 21407 [kh.wild@wi o] sounds usefull.
| | |   21409 [toddg@ma l. ] I'm not conflating cookies with form variables.  They are distinctly
| | + 21335 [matz@ru y- a] Could you explain "for security and to be more flexible" bit more?
| |   + 21379 [chrris@ma l.] I find it quite useful to be able to distinguish between GET, POST and
| |   | 21380 [matz@ru y- a] I think I understand what you meant.  You want them for special cases,
| |   + 21408 [kh.wild@wi o] we prog a web and accept users only if the submit a validation form.
| + 21381 [anders@jo an] cgi.request.var is merged from post, get, cookie and env.
|   21383 [matz@ru y- a] GET /script.cgi?user_var=foo&user_var=bar HTTP/1.1
|   21385 [maki@in c. o] * "bar"           (PHP4)
|   21396 [fgp@ph o. rg] You can also do
|   21400 [eldeerburn@m] Actually, IMO, we should keep cookies separate. From a codemonkey
+ 21341 [domingo@da -] I think that's a good idea have the request joining the cookies, get
+ 21375 [matz@ru y- a] I encourage you to check out http://www.rubygarden.org/ruby?NextGenerationCGI

RUBY NEWSGROUP FAQ -- Welcome to comp.lang.ruby!  (Revised 2001-09-01)
21316 [schneiker@ju] RUBY NEWSGROUP FAQ -- Welcome to comp.lang.ruby!  (Revised 2001-09-01)

PC Magazine
21317 [rtarpine@ho ] In the October issue of PC Magazine, in his Inside Track column John C.
21365 [dblack@ca dl] Hmmmm.... I don't think I'd heard the C++ part before.  Whatever.

SWIG & Ruby
21318 [michal@cr bo] I'd like to ask if somebody is using SWIG for C library and Ruby.
21888 [Craig_Files@] The swig tar ball comes with several examples of using ruby and swig.

iowa and win98
21321 [horibo@gm .d] I'm trying to build a running development environment on my
21322 [avi@be a4 co] Instead of the module, you might just use the iowa.cgi script instead - it
21346 [pit@ca it in] I recently tried to compile mod_iowa on Win 98, too, using VC++
21347 [avi@be a4 co] See http://www.rubytalk.com/17990 for an alternate method (that should
+ 21349 [pit@ca it in] Thanks for the link, I'll have a look at it.
+ 21350 [eli.green@co] But this won't quite work without the URL-parsing patch that Michael Neumann

21328 [pbrannan@at ] X = 'this is a test'         # this is a const reference to an object
+ 21329 [   xm@w3 .r ] Maybe, but 2) seems more reasonable
+ 21330 [Dave@Pr gm t] All you have to do is
  21333 [pbrannan@at ] char * const x;

if before or if after?
21332 [pbrannan@at ] x = '' if not defined?(x)
+ 21336 [decoux@mo lo] At compile time, `x' is resolved as a local variable (ruby has seen an
+ 21338 [Dave@Pr gm t] Just seeing the assignment to 'x' defines it. In
+ 21339 [matz@ru y- a] As a general rule, a local variable start to exist when an assignemt
+ 21367 [yashi@ya hi ] Gang of Three(?) has already posted good explanation but if you wanna

I really get wet when I blow a hot guy.  2858
21344 [ejumxw@ma l.] x-no-archive: yes

Newsgroup strangeness
21359 [harryo@zi .c] I'm just posting this because, on this end, nothing new seems to have been
21358 [harryo@zi .c] Aha!  By re-sorting by name of sender to find my post, I discovered that it

[Ann] JRuby - Ruby interpreter in Java
21362 [japetersen@w] JRuby is the effort to recreate the Ruby interpreter in Java.
21371 [kevinbsmith@] Java.
+ 21372 [chadfowler@y] Great stuff!  I've been looking through the code.  I
| 21391 [japetersen@w] That would be very helpful.
+ 21393 [japetersen@w] Also the design of JRuby (no static fields) make it possible to run more
+ 21582 [benoit.cerri] Sorry but how would this naturally come from having a c++ port.

worm blocker
21363 [joe@vp p. et] tail -f $server_root/logs/error_log | blockworm.rb

How to tell if port is in use...
21377 [hal9000@hy e] How do I tell whether another server already
+ 21386 [csawtell@pa ] netstat -a
+ 21390 [decoux@mo lo] Can you give an example ?
| 21399 [hal9000@hy e] 3333);TCPServer.new("localhost", 3333)'
| 21405 [jonathan.ent] I tried it on mine (win2k, 1.6.4 cygwin), and didn't get an error
| 21476 [wsobel@Ba ra] The problem you are experiencing seems to be intentional. The socket
| 21481 [decoux@mo lo] Windows fatally :-)
| 21638 [wsobel@Ba ra] You are correct! I looked up the definition of REUSEADDR...
| 21639 [decoux@mo lo] You can find it in the socket FAQ, for example at
+ 21464 [erik@ba fo s] #!/usr/bin/ruby

Path walking on windows
21378 [edsin@sw s. ] I've created a little script that recurses through directories and yields
+ 21398 [stephen.hill] I have a very similar function, but I was wondering if this recursive
+ 21429 [edsin@sw s. ] ...
  21485 [niklas@ka i.] [Ed Sinjiashvili]

ruby 1.6.5
21384 [matz@ru y- a] md5sum is