January 2010 - MacRuby-devel - lists.macosforge.org

0.5 status... good to go!
by Laurent Sansonetti 01 Feb '10

01 Feb '10

Hi guys, It has been a long time since I did a status report here, but I have been busy :) Here is a status about what happened since 0.5 beta 2 until we started working on 0.6. I think 0.5 is good enough to be released as final now. I will prepare the release and hopefully it will go out later today. Highlights: - HotCocoa support is back! Also, HotCocoa is no longer developed in the MacRuby repository but on GitHub: http://github.com/richkilmer/hotcocoa . We hope that by being on GitHub, HotCocoa will evolve faster. Be sure to do a 'macrake install hotcocoa'. - AOT compiler is more stable, supports compilation of multiple Ruby files into shared/dynamic libraries. It is also documented in a man- page, be sure to check it out for a detailed list of options and some use case examples. - GCD API is complete, frozen and fully documented. Stay tuned for GCD tutorials on the website. - Better runtime symbolication (backtraces) by emitting DWARF annotations. - Numerous miscellaneous bug fixes! Changes: - macrubyc: better CPU architecture check - ruby catch/throw statements now use specialized C++ exceptions instead of setjmp/longjmp - macrubyc: make sure compilation works if paths contain spaces - macrubyc: no longer use llvm-config for C++ compilation flags but rbconfig - fixed Module#method_defined? - rake build: ignore RUBYOPT - templates: updated rb_main.rb to honor .rbo files - GCD: ignore exceptions raised inside GCD blocks - do not cache blocks for functions - openssl extention: misc fixes - #throw must clear the current VM exception before unwinding the stack - GCD: fixed a few race conditions - define the GC and ObjectSpace API as module functions - start tagging non-RObject classes to not use ivar slots - added Module#__properties__ to return the list of Objective-C properties - never overwrite -[NSObject hash] - when registering a method as a BS informal protocol, pad the signature with @ in case the ruby arity is greater than the objc one - property handle missing 64-bit bridgesupport annotations in informal protocols for pointer types - fixed a bug when inserting an object into a KVO array - str.split('') should split into characters - honor nil backtraces when printing an uncaught exception - misc KVC bug fixes - templates: switched from .nib to .xib - GCD: block shared/dynamic variables are now copied - GCD: Queue#dispatch and Group#dispatch are gone. Replacements: Queue#async(group=nil) and Queue#sync - macrubyc: don't forgot to initialize load path and program name - GCD: aliased Queue#to_s to Queue#label - emit DWARF metadata at compilation time and use it to establish backtracing later at runtime - fixed Time marshalling - GCD: don't try to release a NULL semaphore - fixed a whole lot of crashes on 32-bit - macrubyc: added support for AOT compiled source code objects in #require - macrubyc: added manpage - GCD: when finalizing a dispatch semaphore, re-equilibrate before release to avoid a crash - fixed #method_missing and #respond_to? for NSProxy-based objects - fixed some class variables bugs - when changing the visibility of a method and duplicating its entry in the lookup table, be sure to keep the same flags too - added Pointer out-of-bounds access check - optimized #to_path conversion - fixed Fixnum#odd? and Fixnum#even? - fixed a thread-safety bug in RoxorCore::retype_method() - #extend_object: honor nested modules - added some basic support for multibyte regexps - FFI::Library#attach_function: the first argument can also be a symbol - macrubyc: generate better tmpfiles to allow multiple compilations of the same project - GCD: added rdoc documentation for all APIs - fixed a bunch of super-dispatch bugs - added preliminary (not complete) support for objc methods aliasing - fixed a bug when accessing regexp match special variables - fixed a bug in String#dump in case the receiver is a bytestring - fixed a bug in the #send method inliner where protected methods would not be callable - removed HotCocoa from the repository. It is now living on github as a separate gem. - implemented Thread.kill and Thread.exit - rb_str_dup(): in case the receiver is not a pure CF string, send - mutableCopy - Thread.list: do not return the internal thread list but a copy instead - don't use #delete to remove a thread from the internal list since it might trigger custom methods which might acquire the core lock and therefore cause a deadlock - implemented #[] and #[]= as shortcuts for objectForKey: and setObject:forKey: - GCD: added sources API - do not uncache a block if the context (self) is different - templates: do not load rb_main.rb twice Laurent

6 7

[MacRuby] #372: Error while requiring rss module
by MacRuby 01 Feb '10

01 Feb '10

#372: Error while requiring rss module ------------------------------+--------------------------------------------- Reporter: dmlary@… | Owner: lsansonetti@… Type: defect | Status: new Priority: minor | Milestone: MacRuby 0.5 Component: MacRuby | Keywords: rss, macruby 0.5 beta 1 ------------------------------+--------------------------------------------- Using macruby 0.5 beta 1, {{{ ~> macruby --version MacRuby version 0.5 (ruby 1.9.0) [universal-darwin10.0, x86_64] ~> cat rss.rb require 'rss' ~> macruby rss.rb rss.rb:1:in `<main>': undefined method `split' for nil:NilClass (NoMethodError) }}} -- Ticket URL: <http://www.macruby.org/trac/ticket/372> MacRuby <http://macruby.org/>

1 2

[MacRuby] #581: sudo macgem install hotcocoa fails on latest nightly
by MacRuby 31 Jan '10

31 Jan '10

#581: sudo macgem install hotcocoa fails on latest nightly -----------------------------------+---------------------------------------- Reporter: isaackearse@… | Owner: lsansonetti@… Type: defect | Status: new Priority: major | Milestone: Component: MacRuby | Keywords: -----------------------------------+---------------------------------------- Hi Guys, sudo macgem install hotcocoa not in gzip format not in gzip format (Zlib::GzipFile::Error) This is after installing the latest MacRuby nightly (http://macruby.icoretech.org/details/138) It worked for me a week ago before rubygems was updated to 1.3.5 Cheers, Isaac -- Ticket URL: <http://www.macruby.org/trac/ticket/581> MacRuby <http://macruby.org/>

1 2

[MacRuby] #290: Hashes do NOT stay ordered
by MacRuby 31 Jan '10

31 Jan '10

#290: Hashes do NOT stay ordered -----------------------------------+---------------------------------------- Reporter: martin.ploeger@… | Owner: lsansonetti@… Type: defect | Status: new Priority: blocker | Milestone: Component: MacRuby | Keywords: Hash order -----------------------------------+---------------------------------------- Hashes in Ruby 1.9 keep their order. In MacRuby (macirb): {:aa=>1, :a=>1} => {:a =>1, :aa=>1} According to Ruby 1.9 the result should be in the input-order: => {:aa =>1, :a=>1} The same is true for methods on a Hash like Hash#keys: The should return the keys in the input-order. -- Ticket URL: <http://www.macruby.org/trac/ticket/290> MacRuby <http://macruby.org/>

1 9

[MacRuby] #563: Retrieving the singleton class of a class seems to return Class.
by MacRuby 31 Jan '10

31 Jan '10

#563: Retrieving the singleton class of a class seems to return Class. -------------------------------------+-------------------------------------- Reporter: eloy.de.enige@… | Owner: lsansonetti@… Type: defect | Status: new Priority: blocker | Milestone: Component: MacRuby | Keywords: mocha -------------------------------------+-------------------------------------- For example: {{{ % cat t.rb class A def self.singleton_class (class << self; self; end) end end p A.singleton_class % ruby19 t.rb #<Class:A> % macruby t.rb Class }}} -- Ticket URL: <http://www.macruby.org/trac/ticket/563> MacRuby <http://macruby.org/>

1 1

[MacRuby] #523: require of 'rss/2.0' fails.
by MacRuby 30 Jan '10

30 Jan '10

#523: require of 'rss/2.0' fails. -----------------------------+---------------------------------------------- Reporter: rob@… | Owner: lsansonetti@… Type: defect | Status: new Priority: blocker | Milestone: Component: MacRuby | Keywords: -----------------------------+---------------------------------------------- robert-gleesons-iMac:~ robertgleeson$ macirb irb(main):001:0> require 'rss/2.0' NoMethodError: undefined method `get_file_and_line_from_caller' for RSS::Rss:Class from (irb):1:in `<main>' from core:in `eval:' from core:in `evaluate:' The same error is reported with require 'rss/1.0' Thanks Rob -- Ticket URL: <http://www.macruby.org/trac/ticket/523> MacRuby <http://macruby.org/>

1 3

OpenCL experiment (request for comments)
by Ruben Fonseca 30 Jan '10

30 Jan '10

Hi @all! I promised Laurent I would start a discussion on OpenCL and Macruby on the ML, and here I am :) I wrote a small hack for MacRuby that adds basic (and hacky) support for OpenCL kernels running on your GPUs or CPUs. You can read about it on my blog post here http://blog.0x82.com/2010/1/23/opencl-in-macruby-hack-not-very-useful The branch with the code is located here on github http://github.com/rubenfonseca/macruby. If you are interested on the actual hacky implementation, please look at “opencl.c” file. Now there are a couple of things I need help. It was my very first (mac)ruby C extension, and I’m not really sure about many details on the implementation (I basically copy-pasted code from other modules eheh). I’ll raise a couple of questions here, I’m sure you can answer some of them :) - How and were to store primitive values? For instance, OpenCL::Device has somewhere inside a “cl_device_id” pointer. However, on other classes (think OpenCL::Context#new) I’ll need to have a reference to that “cl_device_id” pointer somewhere. What’s the best way to store the pointer? Inside the Object struct? As an instance var? As an accessor? These later options doesn’t make sense to me, 'cause I’m never interested on getting the “cl_device_id” pointer on a IRB shell for instance... Hope I’m making myself clear. - How the memory should be managed? As I said, I never wrote a MacRuby extension before. When writing the extension, I needed to do a couple of memory allocations. I used “xmalloc” (discovered by looking at other macurby *.c files). However, when I called “free” after I don’t need the memory anymore, all sorts of warnings happened at runtime. After I deleted the “free” calls, it all worked, but I’m not sure if I’m leaking memory somehow. On the other hand, maybe the memory is automatically GC’ed :) Can you clear this for me and show me the best practices? - How to turn the OpenCL API more “Rubyish" I have no clue on this one. OpenCL seems like a huge API, and I don’t have a really background knowledge of GPU programming. There are all sorts of variations on each call, and a number of different entities (classes). Any suggestion on how to make this more pleasant to write in Ruby would be very very welcome. I even saw Laurent talk about a “Ruby -> OpenCL direct compiler via LLVM bitcode”, but I’m definitely not qualified to even consider that hipotesys. I tried twice creating a very simple compiler with LLVM and failed completely :P Anyhow, sorry for the long email, but I would be very thankful for any help I can get :) Cheers, Rúben Fonseca

2 1

[MacRuby] #578: Garbage Collector in 0.6
by MacRuby 30 Jan '10

30 Jan '10

#578: Garbage Collector in 0.6 ---------------------------+------------------------------------------------ Reporter: jsn@… | Owner: lsansonetti@… Type: defect | Status: new Priority: major | Milestone: MacRuby 0.6 Component: MacRuby | Keywords: ---------------------------+------------------------------------------------ Reporting this as a bug, since i'm fairly sure this is not the way it's intended to work. Running the following code on a bunch of 600mb - 2gb files will (eventually) cause a SIGSEGV when memory allocation fails. {{{ def get_digest(file) digest = Digest::MD5.new() fil = File.open(file,'r') while((l = fil.read(READ_BUFFER_SIZE))!=nil) digest << l end fil.close() digest.hexdigest end }}} If I run it on a smaller set of files, 1.8gb to be exact, the memory allocated to the process jumps to 700-900mb, and just stays there. I let the process sit there for 30 mins, without noticing any drop in allocated memory. I've tried with various sizes for READ_BUFFER_SIZE, from 16kb to 32mb, and also tried with File.read(), and it all behaves the same way, although the higher read buffer, the faster memory usage shoots up. setting "l=nil" for every read operation and/or "digest=nil" before returning doesn't seem to make a difference. Using Digest::MD5.file(fname).hexdigest results in "writing to non- bytestrings is not supported at this time." I'm not sure if this is a "string cache" thing gone wrong, or it's simply not collecting garbage. -- Ticket URL: <http://www.macruby.org/trac/ticket/578> MacRuby <http://macruby.org/>

1 1

is there a way to use ruby namespaces in IB?
by Michael Johnston 29 Jan '10

29 Jan '10

I found that if I make an object controller in and set its class to "Bar", it will find any Bar class regardless of nesting in modules. If I define: class Bar;end # in Bar.rb && module Foo; class Bar; end; end; # in Foo/Bar.rb it seems at runtime to randomly choose one of the two (doesn't seem to depend on require order) Is there any way to use ruby namespacing with classes that are to be referred to in IB or passed to Objective C methods? Cheerio, Michael Johnston lastobelus(a)mac.com

2 1

Question about current design
by Emmanuel Pinault 29 Jan '10

29 Jan '10

Hi, I am looking through the repository and although I understand that some class are directly mapping to Object-C class instead, I am curious about how do you go to bring a new version of Ruby 1.9.x? Since you are replacing most of the code in string.c for example, will you to backport all the time changes from 1.9.x? Also, how do you go about embedding some gem ? I see how you embed the framework for MacRuby but what about gems? Do you just have to copy the source code of the gem as well and replace some Path? thanks Emmanuel

2 1