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

[MacRuby] #625: macirb failing with latest trunk
by MacRuby 26 Feb '10

26 Feb '10

#625: macirb failing with latest trunk ----------------------------------------+----------------------------------- Reporter: pontus.stromdahl@… | Owner: lsansonetti@… Type: defect | Status: new Priority: major | Milestone: MacRuby 0.6 Component: MacRuby | Keywords: macirb irb_binding workspace.rb ----------------------------------------+----------------------------------- Get the following trace when trying to start macirb: ---- pontus> macruby -v MacRuby version 0.6 (ruby 1.9.0) [universal-darwin10.0, x86_64] pontus> macirb -f /Users/pontus/Workcopies/spike/macruby/setFrame_Fail/lib/irb/workspace.rb:51:in `<main>': undefined local variable or method `irb_binding' for main:TopLevel (NameError) ---- I tried debugging it but so far I haven't gotten anywhere. -- Ticket URL: <http://www.macruby.org/trac/ticket/625> MacRuby <http://macruby.org/>

1 3

[MacRuby] #622: Overriding setRepresentedObject on NSViewController subclass doesn't seem to work
by MacRuby 24 Feb '10

24 Feb '10

#622: Overriding setRepresentedObject on NSViewController subclass doesn't seem to work -----------------------------+---------------------------------------------- Reporter: ninh@… | Owner: lsansonetti@… Type: defect | Status: new Priority: blocker | Milestone: MacRuby 0.6 Component: MacRuby | Keywords: -----------------------------+---------------------------------------------- Overriding setRepresentedObject in a subclass of NSViewController seems to have no effect. In particular, in macrubybind.rb, it shows that a custom setMyRepresentedObject is being used, but when you just override setRepresentedObject, it's not being used in conjunction with cocoa bindings. In NinhBindingTest.zip you can see that in objective C it works as expected though, i.e. overriding setRepresentedObject yields the expected behavior. -- Ticket URL: <http://www.macruby.org/trac/ticket/622> MacRuby <http://macruby.org/>

1 1

[MacRuby] #621: Overriding setRepresentedObject on NSViewController subclass doesn't seem to work
by MacRuby 24 Feb '10

24 Feb '10

#621: Overriding setRepresentedObject on NSViewController subclass doesn't seem to work -----------------------------+---------------------------------------------- Reporter: ninh@… | Owner: lsansonetti@… Type: defect | Status: new Priority: blocker | Milestone: Component: MacRuby | Keywords: -----------------------------+---------------------------------------------- Overriding setRepresentedObject in a subclass of NSViewController seems to have no effect. In particular, in macrubybind.rb, it shows that a custom setMyRepresentedObject is being used, but when you just override setRepresentedObject, it's not being used in conjunction with cocoa bindings. In NinhBindingTest.zip you can see that in objective C it works as expected though, i.e. overriding setRepresentedObject yields the expected behavior. -- Ticket URL: <http://www.macruby.org/trac/ticket/621> MacRuby <http://macruby.org/>

1 1

[MacRuby] #423: #respond_to? returns true, actual method call raises NoMethodError
by MacRuby 23 Feb '10

23 Feb '10

#423: #respond_to? returns true, actual method call raises NoMethodError ---------------------------------+------------------------------------------ Reporter: ecin@… | Owner: lsansonetti@… Type: defect | Status: new Priority: major | Milestone: Component: MacRuby | Keywords: ---------------------------------+------------------------------------------ framework 'Cocoa' a = NSMetadataQuery.new a.respond_to? :setPredicate # => true a.setPredicate # => NoMethodError: undefined method `setPredicate' for #<NSMetadataQuery:0x2002a1e20> a.setPredicate "a" # => RuntimeError: NSInvalidArgumentException: -[NSCFString generateMetadataDescription]: unrecognized selector sent to instance 0x2002309c0 -- Ticket URL: <http://www.macruby.org/trac/ticket/423> MacRuby <http://macruby.org/>

1 5

[MacRuby] #512: MacRuby misreports #initialize argument count in error messages
by MacRuby 23 Feb '10

23 Feb '10

#512: MacRuby misreports #initialize argument count in error messages -----------------------------------+---------------------------------------- Reporter: matthew@… | Owner: lsansonetti@… Type: defect | Status: new Priority: major | Milestone: Component: MacRuby | Keywords: -----------------------------------+---------------------------------------- Given the following code, {{{ class A def initialize(a, b, c) end end A.new }}} '''Expected result (Ruby 1.9.1)''' {{{ArgumentError: wrong number of arguments (0 for 3)}}} '''Actual result (MacRuby 0.5b2)''' {{{ArgumentError: wrong number of arguments (0 for 1)}}} -- Ticket URL: <http://www.macruby.org/trac/ticket/512> MacRuby <http://macruby.org/>

1 2

macruby/hotcocoa questions
by Gary Weaver 19 Feb '10

19 Feb '10

Hello, I wrote a small application in HotCocoa/MacRuby just to get familiar with it ( http://github.com/garysweaver/hourz ). MacRuby and HotCocoa are awesome! I did have a few issues during development that I thought I'd share in case anyone can assist. I'm using MacRuby 0.5 and HotCocoa 0.5.1 on OS X 10.6.2, was using TextMate vs. XCode to develop it. I have some experience with Ruby, but consider me a newbie to MacRuby, HotCocoa, and Cocoa (and Obj-C for that matter). Here is what I ran into: * If I take a stringValue from a text field, Marshal.dump it to a file, and then Marshal.load it, and put it back into the text field, it puts wierd characters in the end of the value like 'Òÿýÿÿÿ'. I think this might be a result of the class that I'm Marshalling that contains "include HotCocoa::Behaviors"? Not sure. I wasn't able to reproduce the issue via macirb using simple array or array of simple custom class Marshalling to file and back, but I'm able to reproduce the issue everytime in my app, but the class is more complicated. * I wasn't able to get File.copy to work so I wrote my own method to copy a file. Could you provide an example that can copy a file in MacRuby using File.copy or using FileUtils? * I wasn't able to set the (HotCocoa) layout_view frame or hidden properties successfully via calls by a button on_action. I could set them fine as long as they were called during startup of the application. For example, I wanted to do the following to hide a layout_view and replace it with another layout_view and vice versa in the same area of the window, but these methods only worked if being called during the initial load. I'm thinking maybe there is some sort of refresh method I need to call? def in_add_mode @edit_view.frame = [0, 0, 0, 0] @edit_view.hidden = true @add_view.frame = [0, 0, 0, 40] @add_view.hidden = false end def in_edit_mode @add_view.frame = [0, 0, 0, 0] @add_view.hidden = true @edit_view.frame = [0, 0, 0, 40] @edit_view.hidden = false # ... end * (window).will_miniaturize { exit } works great to keep the app active (in the dock) after closing the window, but I don't know what to call such that I could close the window and then click on the icon in the dock it then show the window. Also when I close it, it doesn't appear to be in the hidden state, because the dock menu indicates that I can hide the window (but it is hidden). * I couldn't figure out how to access the dock_menu in HotCocoa/MacRuby (I'm not using XCode or http://www.echographia.com/blog/2009/02/08/dynamic-dock-menus-in-macruby/ would have helped). I'd like to be able to manage it dynamically in HotCocoa if possible to be able to choose which Task I'm working on at the moment without showing the window. It would also be cool to dynamically change the dock icon when this happens to somehow indicate which task is being worked on (similar to Thunderbird showing the number of new messages, maybe I could show a number or some brief text?). * I couldn't figure out how to alter the "About" part of the application to provide the authorship, license, and link to the project, although I know the version number is specified in config/build.yml. It would also be cool to have full control over that area and define the window, etc. * I wasn't sure what the best practice would be to ensure that the window could be resized if someone somehow is using larger/smaller fonts, but to keep it from being resized so much that it is unusable. Maybe the right thing to do is to just specify a defined size/frame for the window that cannot be changed? * Note that I might also look into storing the task data as xml or json. I'd rather it be hand-editable if needed. (I assume I could just put the array into a hash and use to_xml/from_xml or to_json/from_json.) Again, thanks so much for all of your work on this. It is really cool to develop apps for OS X this quickly! (<16 hours total dev time, not including time to post to GitHub, and a lot of that was research, documentation, making an icon, etc.) Thanks in advance for any assistance! Gary

2 2

is active_support supported in MacRuby 0.5?
by Gary Weaver 18 Feb '10

18 Feb '10

Is active_support supported in MacRuby 0.5? $ macirb irb(main):001:0> require 'rubygems' => true irb(main):002:0> require 'active_support' NoMethodError: undefined method `alias_method_chain' for Range:Class Thanks in advance, Gary

2 1

[MacRuby] #613: Dispatch classes should be wrapped in an Data struct to allow subclassing
by MacRuby 18 Feb '10

18 Feb '10

#613: Dispatch classes should be wrapped in an Data struct to allow subclassing ----------------------------------------+----------------------------------- Reporter: ernest.prabhakar@… | Owner: lsansonetti@… Type: defect | Status: new Priority: blocker | Milestone: Component: MacRuby | Keywords: ----------------------------------------+----------------------------------- The gcd.c implementation has this cryptic comment: // TODO: These structures need to be wrapped in a Data struct, // otherwise there are crashes when one tries to add an instance // variable to a queue. (Not that that is a good idea.) At first I ignored this, but later it occurred to me that this may be why I am unable to subclass Dispatch::Group (to use for a Future, which contains an instance variable). Is that correct? If so, is there a particular reason why subclassing these would be bad idea? -- Ticket URL: <http://www.macruby.org/trac/ticket/613> MacRuby <http://macruby.org/>

1 5

[MacRuby] #555: Redefinition of method gives unexpected behavior
by MacRuby 18 Feb '10

18 Feb '10

#555: Redefinition of method gives unexpected behavior --------------------------------+------------------------------------------- Reporter: maccheck@… | Owner: lsansonetti@… Type: defect | Status: new Priority: major | Milestone: MacRuby 0.5 Component: MacRuby | Keywords: --------------------------------+------------------------------------------- When I run following test program with Ruby 1.9.1/1.8.7 and MacRuby 0.5b2 I get an unexpected result: {{{ class TestClass def initialize; @a_var = "content"; end end a = TestClass.new; puts "a: %s" % a.inspect class TestClass def initialize(k); @a_var = k; end end b = TestClass.new "valueGiven"; puts "b: %s" % b.inspect c = TestClass.new; puts "c: %s" % c.inspect }}} results in Ruby 1.9.1/1.8.7 with the expected behavior: {{{ a: #<TestClass:0x100155d50 @a_var="content"> b: #<TestClass:0x100155b20 @a_var="valueGiven"> initTest.rb:13:in `initialize': wrong number of arguments (0 for 1) (ArgumentError) from initTest.rb:13:in `new' from initTest.rb:13 }}} In MacRuby I get following output: {{{ a: ##-<TestClass:0x200020b60 @a_var="content"> b: ##-<TestClass:0x2000213e0 @a_var="valueGiven"> c: ##-<TestClass:0x200021820 @a_var="content"> }}} It seems that {{{#initialize}}} (without arguments) is still floating around... -- Ticket URL: <http://www.macruby.org/trac/ticket/555> MacRuby <http://macruby.org/>

1 3

Immutable foundation objects
by Erik Österlund 18 Feb '10

18 Feb '10

Hello. This is just a thought, but shouldn't for instance NSArray.alloc.init.frozen? return true, as it is immutable and cannot be changed? It is like a frozen version of NSMutableArray, isn't it? It's a bit confusing when asking an immutable Foundation object if it is frozen, and it says no... On the other hand, I suppose it's possible to create categories that add behavior that somehow modifies the object. Any thoughts on this? Thanks.

5 11