rubycocoa-talk — talking about RubyCocoa

Re: [Rubycocoa-talk] Not able to install

[Laurent S. <lsa...@ap...>]

Hi James,

This is interesting. Do you have by any chance a custom Ruby  
installation in /usr?

Could you also attach the installer logs?

Thanks,
Laurent

On Sep 5, 2007, at 11:20 AM, James Sosontovich wrote:

> Unable to install RubyCocoa-0.12.0-OSX10.4universal.pkg on Mac OS X
> version 10.4.10 (This version of OS X was updated from 10.4).  When
> the installer finished writing the files, 100%, the system seems to
> hang, and then two brief messages flash on screen: ... writing
> packages ... Running ..., and then it crashes.  The message:  Please
> try installing again.   Tried, but same results.  Is my upgrade from
> 10.4 to 10.4.10, patched, a problem?
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a  
> browser.
> Download your FREE copy of Splunk now >>  http://get.splunk.com/
> _______________________________________________
> Rubycocoa-talk mailing list
> Rub...@li...
> https://lists.sourceforge.net/lists/listinfo/rubycocoa-talk

Re: [Rubycocoa-talk] having an image button rotating as iSync

[Laurent S. <lsa...@ap...>]

Yvon,

What you need to do is write a custom view that would draw a rotated  
image. Then you can use NSAnimation or NSTimer to force the display of  
the control.

There are no problems specific to RubyCocoa.

Laurent

On Sep 5, 2007, at 12:06 PM, Yvon Thoraval wrote:

> i have a rubycocoa application similar to iSync and i'd like having  
> the sync button rotating.
>
> i able from a thread to make it rotating when the sync button isn't  
> activated.
>
> the same when i use an NSTimer in place of a ruby thread.
>
> the sync button stays with an image pressed and darker.
>
> no other action are possible until the action of this button is  
> terminated.
>
> are those probs specific to rubycocoa ?
>
> -- 
> yvon  
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a  
> browser.
> Download your FREE copy of Splunk now >>  http://get.splunk.com/_______________________________________________
> Rubycocoa-talk mailing list
> Rub...@li...
> https://lists.sourceforge.net/lists/listinfo/rubycocoa-talk

[Rubycocoa-talk] having an image button rotating as iSync

[Yvon T. <yvo...@gm...>]

i have a rubycocoa application similar to iSync and i'd like having the sync
button rotating.

i able from a thread to make it rotating when the sync button isn't
activated.

the same when i use an NSTimer in place of a ruby thread.

the sync button stays with an image pressed and darker.

no other action are possible until the action of this button is terminated.

are those probs specific to rubycocoa ?

-- 
yvon

[Rubycocoa-talk] Not able to install

[James S. <jj...@pp...>]

Unable to install RubyCocoa-0.12.0-OSX10.4universal.pkg on Mac OS X  
version 10.4.10 (This version of OS X was updated from 10.4).  When  
the installer finished writing the files, 100%, the system seems to  
hang, and then two brief messages flash on screen: ... writing  
packages ... Running ..., and then it crashes.  The message:  Please  
try installing again.   Tried, but same results.  Is my upgrade from  
10.4 to 10.4.10, patched, a problem?

Re: [Rubycocoa-talk] RubyCocoa stanalonification ???

[Ferhat A. <fer...@ya...>]

Hi list,

I'm new to RubyCocoa and would like to know how I can 
write  a standalone application to publish it with the
rest of the world.
Is there a doc which passed? Google just gives me the
standalonify script but I really don't know whether
this is sufficient. Has the user to install RubyCocoa
too?

Thanks and sorry if this is a 'very newbie' question.
RubyCocoa is very promising!

Ferhat

what are your advice for standalonification of a
RubyCocoa app?

i'm nearly to need that for beta testing

best,

-- 
yvon 




       
____________________________________________________________________________________Ready for the edge of your seat? 
Check out tonight's top picks on Yahoo! TV. 
http://tv.yahoo.com/

Re: [Rubycocoa-talk] making an NSButton image rotating ???

[Yvon T. <yvo...@gm...>]

2007/9/4, apparao <ap...@ef...>:
>
> Hi,
>
>         This functionality is achieved by displaying images  by using
> Timer.
>
> - Apparao.



OK, thanks i'll have a look upon Timler...



-- 
yvon

Re: [Rubycocoa-talk] making an NSButton image rotating ???

[apparao <ap...@ef...>]

Hi,

	This functionality is achieved by displaying images  by using Timer.

- Apparao.

On 03-Sep-07, at 11:23 PM, Yvon Thoraval wrote:

>
> i'm able to make the image of a button (the same as in eSync)  
> rotating when the button isn't pressed.
>
> then the lines of code are correct.
>
> however after the button has been activated, it stays darker  
> without rotating.
>
> obviously i've tried to setImage and/or setAlternateImage without  
> success...
>
> the rotation of the image is done in a thread apart from main one.
>
> any idae ???
>
> the thread is perfectly working when used apart from the true  
> action of the button...
>
> what i'd like to do :
>
> have a similar button as in eSync
> -- 
> yvon
> ---------------------------------------------------------------------- 
> ---
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a  
> browser.
> Download your FREE copy of Splunk now >>  http://get.splunk.com/ 
> _______________________________________________
> Rubycocoa-talk mailing list
> Rub...@li...
> https://lists.sourceforge.net/lists/listinfo/rubycocoa-talk

[Rubycocoa-talk] making an NSButton image rotating ???

[Yvon T. <yvo...@gm...>]

i'm able to make the image of a button (the same as in eSync) rotating when
the button isn't pressed.

then the lines of code are correct.

however after the button has been activated, it stays darker without
rotating.

obviously i've tried to setImage and/or setAlternateImage without success...

the rotation of the image is done in a thread apart from main one.

any idae ???

the thread is perfectly working when used apart from the true action of the
button...

what i'd like to do :

have a similar button as in eSync
-- 
yvon

[Rubycocoa-talk] RubyCocoa stanalonification ???

[Yvon T. <yvo...@gm...>]

what are your advice for standalonification of a RubyCocoa app?

i'm nearly to need that for beta testing

best,

-- 
yvon

[Rubycocoa-talk] No more Cocoa Ruby Application in XCode after update

[Yvon T. <yvo...@gm...>]

just to say i do have no more Cocoa Ruby Application in the chooser of XCode
after install of the latest apple dev update.

this is true even if i re-install RubyCocoa afterwards...

best,

-- 
yvon

Re: [Rubycocoa-talk] booting a rails app from RC

[Eloy D. <elo...@gm...>]

Hey Tim,

I'm also using the solution mentioned by Laurent atm in an application.
Unless you need to be able to interact more than via HTTP this is a
good viable option.

My 2 cents :)

Cheers,
Eloy

On 8/31/07, Laurent Sansonetti <lsa...@ap...> wrote:
> Hey Tim,
>
> NSTask could also be used to launch mongrel within the appropriate
> Rails app directory.
>
> http://developer.apple.com/documentation/Cocoa/Reference/Foundation/Classes/NSTask_Class/Reference/Reference.html
>
> NSTalk might be more reliable than the Ruby core equivalents,
> especially since you target OSX.
>
> Laurent
>
> On Aug 30, 2007, at 9:35 PM, Tim Perrett wrote:
>
> > Hey chaps
> >
> > I dont suppose if anyone knows if there is a way of booting a rails
> > app (with mongrel) from a RC application is there? I guess I just
> > need some programatic way of starting starting mongrel and then to
> > check periodically if it is still running?
> >
> > Cheers
> >
> > Tim
> >
> >
> >
> >
> > -------------------------------------------------------------------------
> > This SF.net email is sponsored by: Splunk Inc.
> > Still grepping through log files to find problems?  Stop.
> > Now Search log events and configuration files using AJAX and a
> > browser.
> > Download your FREE copy of Splunk now >>  http://get.splunk.com/
> > _______________________________________________
> > Rubycocoa-talk mailing list
> > Rub...@li...
> > https://lists.sourceforge.net/lists/listinfo/rubycocoa-talk
>
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a browser.
> Download your FREE copy of Splunk now >>  http://get.splunk.com/
> _______________________________________________
> Rubycocoa-talk mailing list
> Rub...@li...
> https://lists.sourceforge.net/lists/listinfo/rubycocoa-talk
>

Re: [Rubycocoa-talk] booting a rails app from RC

[Jon B. <jo...@gm...>]

Not necessarily true ... Rails will boot the server in the order  
listed below, I think you may be looking @ older source(?) ...

server = case ARGV.first
   when "lighttpd", "mongrel", "webrick"
     ARGV.shift
   else
     if defined?(Mongrel)
       "mongrel"
     elsif RUBY_PLATFORM !~ /(:?mswin|mingw)/ && !silence_stderr  
{ `lighttpd -version` }.blank? && defined?(FCGI)
       "lighttpd"
     else
       "webrick"
     end
end

On Aug 31, 2007, at 4:28 AM, Tim Perrett wrote:

> Hey Jon
>
> As far as im aware there using Lighttpd/FCGI rather than mongrel, so
> the setup would be quite a bit more complicated I should have
> thought. I have been looking at the rails "script/server" command
> boots its mongrel instance, and they seem to be using (from the rails
> source) a method called "load", with this code:
>
>    trap(:INT) { exit }
>
>    begin
>      silence_warnings { ARGV = parameters }
>      load("mongrel_rails")
>    ensure
>      tail_thread.kill if tail_thread
>      puts 'Exiting'
>    end
>
> They do some other configuration before hand of course, but the thing
> is, I am not sure what load() its actually calling? Sorry thats out
> of a left field but I just wondered if anyone knew?
>
> Cheers
>
> Tim
>
>
> On 30 Aug 2007, at 20:55, Jon Baer wrote:
>
>> Might want to check + see how Locomotive guys are doing it
>> underneath ...
>>
>> http://locomotive.raaum.org
>>
>> - Jon
>
>
> ---------------------------------------------------------------------- 
> ---
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a  
> browser.
> Download your FREE copy of Splunk now >>  http://get.splunk.com/
> _______________________________________________
> Rubycocoa-talk mailing list
> Rub...@li...
> https://lists.sourceforge.net/lists/listinfo/rubycocoa-talk

Re: [Rubycocoa-talk] booting a rails app from RC

[Laurent S. <lsa...@ap...>]

Hey Tim,

NSTask could also be used to launch mongrel within the appropriate  
Rails app directory.

http://developer.apple.com/documentation/Cocoa/Reference/Foundation/Classes/NSTask_Class/Reference/Reference.html

NSTalk might be more reliable than the Ruby core equivalents,  
especially since you target OSX.

Laurent

On Aug 30, 2007, at 9:35 PM, Tim Perrett wrote:

> Hey chaps
>
> I dont suppose if anyone knows if there is a way of booting a rails
> app (with mongrel) from a RC application is there? I guess I just
> need some programatic way of starting starting mongrel and then to
> check periodically if it is still running?
>
> Cheers
>
> Tim
>
>
>
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a  
> browser.
> Download your FREE copy of Splunk now >>  http://get.splunk.com/
> _______________________________________________
> Rubycocoa-talk mailing list
> Rub...@li...
> https://lists.sourceforge.net/lists/listinfo/rubycocoa-talk

Re: [Rubycocoa-talk] Programmatically overriding 0-arity ObjC methods

[Laurent S. <lsa...@ap...>]

On Aug 29, 2007, at 7:50 PM, sergio wrote:
> I did some googling and I found that the arity conventions in ruby
> 1.9 could change... I don't know what is the official source of
> information about 1.9, but I found this interesting:
>
> (1) http://blade.nagaokaut.ac.jp/cgi-bin/scat.rb/ruby/ruby-talk/120252
> (2) http://blade.nagaokaut.ac.jp/cgi-bin/scat.rb/ruby/ruby-core/11029?
> help-en
> (3) http://blade.nagaokaut.ac.jp/cgi-bin/scat.rb/ruby/ruby-talk/249149
>
> in short, it seems that in 1.9 both proc{}.arity and proc{||}.arity
> shall return 0 -- by definition (1) -- although it seems (3) that in
> some not so old ruby 1.9.0 release both returned -1. in any case,
> this would be a change from how it works now. if what I am saying is
> just crap, simply ignore it... :-)

Interesting, I didn't know that at least the core team decided to  
investigate a new behavior for 1.9.

Let's keep the RubyCocoa exception for -1 arity methods then, it  
doesn't really harm and we have this workaround :)

Laurent

Re: [Rubycocoa-talk] booting a rails app from RC

[Bryan L. <br...@os...>]

On Aug 31, 2007, at 4:28 AM, Tim Perrett wrote:
>
>    trap(:INT) { exit }
>
>    begin
>      silence_warnings { ARGV = parameters }
>      load("mongrel_rails")
>    ensure
>      tail_thread.kill if tail_thread
>      puts 'Exiting'
>    end
>
> They do some other configuration before hand of course, but the thing
> is, I am not sure what load() its actually calling? Sorry thats out
> of a left field but I just wondered if anyone knew?
>

http://corelib.rubyonrails.org/classes/Kernel.html#M002059 (rdoc for  
Kernel#load)

mongrel_rails is in the bin/ directory of your mongrel gem install  
install.

Re: [Rubycocoa-talk] booting a rails app from RC

[Tim P. <he...@ti...>]

Hey Jon

As far as im aware there using Lighttpd/FCGI rather than mongrel, so  
the setup would be quite a bit more complicated I should have  
thought. I have been looking at the rails "script/server" command  
boots its mongrel instance, and they seem to be using (from the rails  
source) a method called "load", with this code:

   trap(:INT) { exit }

   begin
     silence_warnings { ARGV = parameters }
     load("mongrel_rails")
   ensure
     tail_thread.kill if tail_thread
     puts 'Exiting'
   end

They do some other configuration before hand of course, but the thing  
is, I am not sure what load() its actually calling? Sorry thats out  
of a left field but I just wondered if anyone knew?

Cheers

Tim


On 30 Aug 2007, at 20:55, Jon Baer wrote:

> Might want to check + see how Locomotive guys are doing it  
> underneath ...
>
> http://locomotive.raaum.org
>
> - Jon

Re: [Rubycocoa-talk] booting a rails app from RC

[Jon B. <jo...@gm...>]

Might want to check + see how Locomotive guys are doing it  
underneath ...

http://locomotive.raaum.org

- Jon

On Aug 30, 2007, at 3:35 PM, Tim Perrett wrote:

> Hey chaps
>
> I dont suppose if anyone knows if there is a way of booting a rails
> app (with mongrel) from a RC application is there? I guess I just
> need some programatic way of starting starting mongrel and then to
> check periodically if it is still running?
>
> Cheers
>
> Tim

[Rubycocoa-talk] booting a rails app from RC

[Tim P. <he...@ti...>]

Hey chaps

I dont suppose if anyone knows if there is a way of booting a rails  
app (with mongrel) from a RC application is there? I guess I just  
need some programatic way of starting starting mongrel and then to  
check periodically if it is still running?

Cheers

Tim

Re: [Rubycocoa-talk] Programmatically overriding 0-arity ObjC methods

[sergio <sde...@gm...>]

On 29/ago/07, at 01:18, Laurent Sansonetti wrote:

>> this is my first post to this list.
>
> Welcome!

thanks!

>> [...]
>> /Library/Frameworks/RubyCocoa.framework/Versions/A/Resources/ruby/ 
>> osx/
>> objc/oc_import.rb:318:in `_ns_behavior_method_added': Cannot override
>> Objective-C method 'awakeFromInsert' with Ruby method
>> #awakeFromInsert, they should both have the same number of arguments.
>> (expected arity 0, got -1) (RuntimeError)

> The problem here is that an empty block (proc {}) has an arity of -1,
> for an unknown reason.  But a block with an empty argument list (proc
> {||}) will have an arity of 0. So normally if you call define_method
> with blocks with an empty argument list, you should not get the
> exception.

perfect! it works with an empty argument list...

I did some googling and I found that the arity conventions in ruby  
1.9 could change... I don't know what is the official source of  
information about 1.9, but I found this interesting:

(1) http://blade.nagaokaut.ac.jp/cgi-bin/scat.rb/ruby/ruby-talk/120252
(2) http://blade.nagaokaut.ac.jp/cgi-bin/scat.rb/ruby/ruby-core/11029? 
help-en
(3) http://blade.nagaokaut.ac.jp/cgi-bin/scat.rb/ruby/ruby-talk/249149

in short, it seems that in 1.9 both proc{}.arity and proc{||}.arity  
shall return 0 -- by definition (1) -- although it seems (3) that in  
some not so old ruby 1.9.0 release both returned -1. in any case,  
this would be a change from how it works now. if what I am saying is  
just crap, simply ignore it... :-)

best
sergio

Re: [Rubycocoa-talk] NSOutlineView crash

[Laurent S. <lsa...@ap...>]

Hi Hans,

Actually the problem here is that you return a pure Ruby object as the  
node name to NSOutlineView.

When you pass pure Ruby objects to the ObjC world, RubyCocoa will  
build an ObjC proxy for you. However, the proxy object is built as  
autoreleased, so that it will be soonly collected if not retained.

NSOutlineView expects you to always pass the same objects (in the ObjC  
world, the same pointer address) as data source items. However, since  
NSOutlineView doesn't retain the items, the RubyCocoa proxies will be  
remade, and then NSOutlineView crashes.

There is unfortunately no reliable way for RubyCocoa to keep the proxy  
objects and release them once the embedded Ruby object dies (I tried  
several code paths unsuccessfully). Fortunately, except for  
NSOutlineView, the current mechanism seems to always work.

To work around this problem, you always have to return NSObject-based  
objects as NSOutlineView data source items.

In your project, I changed Node#setItemName:

   def setItemName(name)
     @itemName = NSString.stringWithString(name)
   end

and also #init to call

   setItemName('node')

And I was able to edit cells.

We should definitely write this in a Wiki page.

Laurent

On Aug 29, 2007, at 4:40 PM, Hans-Georg wrote:

> RubyCocoa crashes if I a cell in a NSOutlineView is edited. To verify
> this download the following file from rubycocoa.com:
>
> http://rubycocoa.com/files/outline-views/OutlineMeRuby.zip
>
> Run the project, double-click "Root"  In the NSOutlineView and edit
> the cell content. If you then click outside the cell, the app will
> crash.
>
> Hans-Georg
>
>
> Exception:  EXC_BAD_ACCESS (0x0001)
> Codes:      KERN_PROTECTION_FAILURE (0x0002) at 0x00000000
>
> Thread 0 Crashed:
> 0   libobjc.A.dylib                	0x90a40100 objc_msgSend + 32
> 1   com.apple.rubycocoa            	0x0002f26c rbobj_to_nsobj + 1436
> 2   com.apple.rubycocoa            	0x0002f5f8 rbobj_to_ocdata + 776
> 3   com.apple.rubycocoa            	0x00030ea8 sel_to_rbobj + 1544
> 4   com.apple.rubycocoa            	0x00042064
> ffi_closure_helper_DARWIN + 644
> 5   com.apple.rubycocoa            	0x00042354 ffi_closure_ASM + 116
> 6   com.apple.AppKit               	0x93850dd0 -[NSTableView
> _drawContentsAtRow:column:clipRect:] + 252
> 7   com.apple.AppKit               	0x938f6520 -[NSOutlineView
> _drawContentsAtRow:column:clipRect:] + 244
> 8   com.apple.AppKit               	0x93850854 -[NSTableView
> drawRow:clipRect:] + 220
> 9   com.apple.AppKit               	0x938f6418 -[NSOutlineView
> drawRow:clipRect:] + 184
> 10  com.apple.AppKit               	0x93850610 -[NSTableView
> drawRowIndexes:clipRect:] + 108
> 11  com.apple.AppKit               	0x9384fc98 -[NSTableView
> drawRect:] + 1144
> 12  com.apple.AppKit               	0x937e4858 -[NSView
> _drawRect:clip:] + 2128
> 13  com.apple.AppKit               	0x937e3e18 -[NSView
> _recursiveDisplayAllDirtyWithLockFocus:visRect:] + 404
> 14  com.apple.AppKit               	0x937e6b60
> _recursiveDisplayInRect2 + 84
> 15  com.apple.CoreFoundation       	0x907ed3ec CFArrayApplyFunction +
> 416
> 16  com.apple.AppKit               	0x937e3f2c -[NSView
> _recursiveDisplayAllDirtyWithLockFocus:visRect:] + 680
> 17  com.apple.AppKit               	0x937e6b60
> _recursiveDisplayInRect2 + 84
> 18  com.apple.CoreFoundation       	0x907ed3ec CFArrayApplyFunction +
> 416
> 19  com.apple.AppKit               	0x937e3f2c -[NSView
> _recursiveDisplayAllDirtyWithLockFocus:visRect:] + 680
> 20  com.apple.AppKit               	0x937e33e0 -[NSView
> _recursiveDisplayRectIfNeededIgnoringOpacity:isVisibleRect:rectIsVisible
> RectForView:topView:] + 196
> 21  com.apple.AppKit               	0x937e39a8 -[NSView
> _recursiveDisplayRectIfNeededIgnoringOpacity:isVisibleRect:rectIsVisible
> RectForView:topView:] + 1676
> 22  com.apple.AppKit               	0x937e39a8 -[NSView
> _recursiveDisplayRectIfNeededIgnoringOpacity:isVisibleRect:rectIsVisible
> RectForView:topView:] + 1676
> 23  com.apple.AppKit               	0x93804044 -[NSThemeFrame
> _recursiveDisplayRectIfNeededIgnoringOpacity:isVisibleRect:rectIsVisible
> RectForView:topView:] + 192
> 24  com.apple.AppKit               	0x937dd054 -[NSView
> _displayRectIgnoringOpacity:isVisibleRect:rectIsVisibleRectForView:]
> + 384
> 25  com.apple.AppKit               	0x937d2348 -[NSView
> displayIfNeeded] + 248
> 26  com.apple.AppKit               	0x937d21b8 -[NSWindow
> displayIfNeeded] + 180
> 27  com.apple.AppKit               	0x937d2064
> _handleWindowNeedsDisplay + 200
> 28  com.apple.CoreFoundation       	0x907dd76c __CFRunLoopDoObservers
> + 352
> 29  com.apple.CoreFoundation       	0x907dda0c __CFRunLoopRun + 420
> 30  com.apple.CoreFoundation       	0x907dd4ac CFRunLoopRunSpecific +
> 268
> 31  com.apple.HIToolbox            	0x93298b20
> RunCurrentEventLoopInMode + 264
> 32  com.apple.HIToolbox            	0x9329812c ReceiveNextEventCommon
> + 244
> 33  com.apple.HIToolbox            	0x93298020
> BlockUntilNextEventMatchingListInMode + 96
> 34  com.apple.AppKit               	0x9379eae4 _DPSNextEvent + 384
> 35  com.apple.AppKit               	0x9379e7a8 -[NSApplication
> nextEventMatchingMask:untilDate:inMode:dequeue:] + 116
> 36  com.apple.AppKit               	0x9379acec -[NSApplication run] +
> 472
> 37  com.apple.AppKit               	0x9388b87c NSApplicationMain + 452
> 38  com.apple.rubycocoa            	0x0004227c ffi_call_DARWIN + 208
> 39  com.apple.rubycocoa            	0x00041db4 ffi_call + 212
> 40  com.apple.rubycocoa            	0x0003fbfc rb_ffi_dispatch + 5180
> 41  com.apple.rubycocoa            	0x00038488
> find_bs_boxed_by_encoding + 1176
> 42  libruby.dylib                  	0x0022686c rb_call0 + 1180
> (eval.c:5848)
> 43  libruby.dylib                  	0x002273e4 rb_call + 612 (eval.c:
> 6096)
> 44  libruby.dylib                  	0x002241ec rb_eval + 7468 (eval.c:
> 3473)
> 45  libruby.dylib                  	0x00234bbc ruby_exec_internal +
> 172 (eval.c:1636)
> 46  libruby.dylib                  	0x00234c08 ruby_exec + 24 (eval.c:
> 1655)
> 47  libruby.dylib                  	0x00234c58 ruby_run + 56 (eval.c:
> 1665)
> 48  com.apple.rubycocoa            	0x000356f8 RBApplicationMain + 344
> 49  ....rubycocoa.OutlineMeRubyApp 	0x000019c8 _start + 760
> 50  ....rubycocoa.OutlineMeRubyApp 	0x000016cc start + 48
>
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a  
> browser.
> Download your FREE copy of Splunk now >>  http://get.splunk.com/
> _______________________________________________
> Rubycocoa-talk mailing list
> Rub...@li...
> https://lists.sourceforge.net/lists/listinfo/rubycocoa-talk

[Rubycocoa-talk] NSOutlineView crash

[Hans-Georg <pan...@ma...>]

RubyCocoa crashes if I a cell in a NSOutlineView is edited. To verify  
this download the following file from rubycocoa.com:

http://rubycocoa.com/files/outline-views/OutlineMeRuby.zip

Run the project, double-click "Root"  In the NSOutlineView and edit  
the cell content. If you then click outside the cell, the app will  
crash.

Hans-Georg


Exception:  EXC_BAD_ACCESS (0x0001)
Codes:      KERN_PROTECTION_FAILURE (0x0002) at 0x00000000

Thread 0 Crashed:
0   libobjc.A.dylib                	0x90a40100 objc_msgSend + 32
1   com.apple.rubycocoa            	0x0002f26c rbobj_to_nsobj + 1436
2   com.apple.rubycocoa            	0x0002f5f8 rbobj_to_ocdata + 776
3   com.apple.rubycocoa            	0x00030ea8 sel_to_rbobj + 1544
4   com.apple.rubycocoa            	0x00042064  
ffi_closure_helper_DARWIN + 644
5   com.apple.rubycocoa            	0x00042354 ffi_closure_ASM + 116
6   com.apple.AppKit               	0x93850dd0 -[NSTableView  
_drawContentsAtRow:column:clipRect:] + 252
7   com.apple.AppKit               	0x938f6520 -[NSOutlineView  
_drawContentsAtRow:column:clipRect:] + 244
8   com.apple.AppKit               	0x93850854 -[NSTableView  
drawRow:clipRect:] + 220
9   com.apple.AppKit               	0x938f6418 -[NSOutlineView  
drawRow:clipRect:] + 184
10  com.apple.AppKit               	0x93850610 -[NSTableView  
drawRowIndexes:clipRect:] + 108
11  com.apple.AppKit               	0x9384fc98 -[NSTableView  
drawRect:] + 1144
12  com.apple.AppKit               	0x937e4858 -[NSView  
_drawRect:clip:] + 2128
13  com.apple.AppKit               	0x937e3e18 -[NSView  
_recursiveDisplayAllDirtyWithLockFocus:visRect:] + 404
14  com.apple.AppKit               	0x937e6b60  
_recursiveDisplayInRect2 + 84
15  com.apple.CoreFoundation       	0x907ed3ec CFArrayApplyFunction +  
416
16  com.apple.AppKit               	0x937e3f2c -[NSView  
_recursiveDisplayAllDirtyWithLockFocus:visRect:] + 680
17  com.apple.AppKit               	0x937e6b60  
_recursiveDisplayInRect2 + 84
18  com.apple.CoreFoundation       	0x907ed3ec CFArrayApplyFunction +  
416
19  com.apple.AppKit               	0x937e3f2c -[NSView  
_recursiveDisplayAllDirtyWithLockFocus:visRect:] + 680
20  com.apple.AppKit               	0x937e33e0 -[NSView  
_recursiveDisplayRectIfNeededIgnoringOpacity:isVisibleRect:rectIsVisible 
RectForView:topView:] + 196
21  com.apple.AppKit               	0x937e39a8 -[NSView  
_recursiveDisplayRectIfNeededIgnoringOpacity:isVisibleRect:rectIsVisible 
RectForView:topView:] + 1676
22  com.apple.AppKit               	0x937e39a8 -[NSView  
_recursiveDisplayRectIfNeededIgnoringOpacity:isVisibleRect:rectIsVisible 
RectForView:topView:] + 1676
23  com.apple.AppKit               	0x93804044 -[NSThemeFrame  
_recursiveDisplayRectIfNeededIgnoringOpacity:isVisibleRect:rectIsVisible 
RectForView:topView:] + 192
24  com.apple.AppKit               	0x937dd054 -[NSView  
_displayRectIgnoringOpacity:isVisibleRect:rectIsVisibleRectForView:]  
+ 384
25  com.apple.AppKit               	0x937d2348 -[NSView  
displayIfNeeded] + 248
26  com.apple.AppKit               	0x937d21b8 -[NSWindow  
displayIfNeeded] + 180
27  com.apple.AppKit               	0x937d2064  
_handleWindowNeedsDisplay + 200
28  com.apple.CoreFoundation       	0x907dd76c __CFRunLoopDoObservers  
+ 352
29  com.apple.CoreFoundation       	0x907dda0c __CFRunLoopRun + 420
30  com.apple.CoreFoundation       	0x907dd4ac CFRunLoopRunSpecific +  
268
31  com.apple.HIToolbox            	0x93298b20  
RunCurrentEventLoopInMode + 264
32  com.apple.HIToolbox            	0x9329812c ReceiveNextEventCommon  
+ 244
33  com.apple.HIToolbox            	0x93298020  
BlockUntilNextEventMatchingListInMode + 96
34  com.apple.AppKit               	0x9379eae4 _DPSNextEvent + 384
35  com.apple.AppKit               	0x9379e7a8 -[NSApplication  
nextEventMatchingMask:untilDate:inMode:dequeue:] + 116
36  com.apple.AppKit               	0x9379acec -[NSApplication run] +  
472
37  com.apple.AppKit               	0x9388b87c NSApplicationMain + 452
38  com.apple.rubycocoa            	0x0004227c ffi_call_DARWIN + 208
39  com.apple.rubycocoa            	0x00041db4 ffi_call + 212
40  com.apple.rubycocoa            	0x0003fbfc rb_ffi_dispatch + 5180
41  com.apple.rubycocoa            	0x00038488  
find_bs_boxed_by_encoding + 1176
42  libruby.dylib                  	0x0022686c rb_call0 + 1180  
(eval.c:5848)
43  libruby.dylib                  	0x002273e4 rb_call + 612 (eval.c: 
6096)
44  libruby.dylib                  	0x002241ec rb_eval + 7468 (eval.c: 
3473)
45  libruby.dylib                  	0x00234bbc ruby_exec_internal +  
172 (eval.c:1636)
46  libruby.dylib                  	0x00234c08 ruby_exec + 24 (eval.c: 
1655)
47  libruby.dylib                  	0x00234c58 ruby_run + 56 (eval.c: 
1665)
48  com.apple.rubycocoa            	0x000356f8 RBApplicationMain + 344
49  ....rubycocoa.OutlineMeRubyApp 	0x000019c8 _start + 760
50  ....rubycocoa.OutlineMeRubyApp 	0x000016cc start + 48

Re: [Rubycocoa-talk] Programmatically overriding 0-arity ObjC methods

[Laurent S. <lsa...@ap...>]

Hi,

On Aug 28, 2007, at 10:19 PM, sergio wrote:

> Hi all,
>
> this is my first post to this list.

Welcome!

> [...]
> /Library/Frameworks/RubyCocoa.framework/Versions/A/Resources/ruby/osx/
> objc/oc_import.rb:318:in `_ns_behavior_method_added': Cannot override
> Objective-C method 'awakeFromInsert' with Ruby method
> #awakeFromInsert, they should both have the same number of arguments.
> (expected arity 0, got -1) (RuntimeError)
>
> I am getting this behaviour in RubyCocoa 0.12.0 and 0.11.1. The same
> code works under RubyCocoa 0.10.1 (and previous versions that used
> the ns_override directive). Furthermore, this behaviour only is shown
> when overriding methods that take no parameters; everything works
> fine when overriding, say, willChangeValueForKey, which takes one
> parameter. I tried to override the didSave method as well (no
> parameters), and it also failed, so it seems to be related to the
> number of parameters.
>
> Of course, overriding always works when I am not doing it via
> define_method but plainly in the class definition.
>
> Any clues anyone? I am willing to explore further this issue on my
> own, provided there is no easy explanation, if someone is so kind to
> point me in the right direction....

The exception you get was added to avoid users overriding methods with  
the wrong arity (which could cause crashes when calling back the Ruby  
method from ObjC).

The problem here is that an empty block (proc {}) has an arity of -1,  
for an unknown reason.  But a block with an empty argument list (proc  
{||}) will have an arity of 0. So normally if you call define_method  
with blocks with an empty argument list, you should not get the  
exception.

Typically, we don't really care about methods with a -1 arity, since  
they are usually not written accidentally. So we could just not throw  
the exception for -1 arity methods.

Laurent

Re: [Rubycocoa-talk] [Slightly OT] Licensing

[Tim P. <he...@ti...>]

Hey all,

Thanks for the replys - that is most interesting, I will take a good  
look at the aquaticmac system and see how I could potentially adapt  
to fit so I dont need to re-invent the wheel :)

Cheers

Tim

On 27 Aug 2007, at 23:46, Jon Baer wrote:

> I was subscribed to #macsb mailing list for a while and the  
> conversation *always* seemed to come up (I think that is where the  
> post by Allan generated from as well) and there were like a few  
> dozen good ideas.  You might have more luck subscribing + posting  
> this question to their mailing list ...
>
> http://macsb.ironcoder.org/wiki/Main_Page
> http://macsb.ironcoder.org/wiki/WhoIsWho
>
> I have not thought about it yet but from what I have read your idea  
> is pretty much like what the Aquatic Prime key generation does.
>
> http://www.aquaticmac.com/
>
> - Jon

[Rubycocoa-talk] Programmatically overriding 0-arity ObjC methods

[sergio <sde...@gm...>]

Hi all,

this is my first post to this list.

I have been recently playing with metaprogramming a RubyCocoa app  
using Core Data, and I got into a behaviour I don't fully understand  
when programmatically overriding an NSManagedObject method that takes  
no parameters, as didSave or awakeFromInsert.

Say I have the following bit of code :

class RCEntity < OSX::NSManagedObject

     ....

     def self.makeEntity(...)

        define_method(:awakeFromInsert) do
            super_awakeFromInsert
            # more stuff here
        end

     end
end

class TestEntity < RCEntity
     makeEntity(...)
end

(this code is not meant to be run)

What I expect is that, due to the call to makeEntity, awakeFromInsert  
is  overriden by TestEntity. What I get is the following error message:

/Library/Frameworks/RubyCocoa.framework/Versions/A/Resources/ruby/osx/ 
objc/oc_import.rb:318:in `_ns_behavior_method_added': Cannot override  
Objective-C method 'awakeFromInsert' with Ruby method  
#awakeFromInsert, they should both have the same number of arguments.  
(expected arity 0, got -1) (RuntimeError)

I am getting this behaviour in RubyCocoa 0.12.0 and 0.11.1. The same  
code works under RubyCocoa 0.10.1 (and previous versions that used  
the ns_override directive). Furthermore, this behaviour only is shown  
when overriding methods that take no parameters; everything works  
fine when overriding, say, willChangeValueForKey, which takes one  
parameter. I tried to override the didSave method as well (no  
parameters), and it also failed, so it seems to be related to the  
number of parameters.

Of course, overriding always works when I am not doing it via  
define_method but plainly in the class definition.

Any clues anyone? I am willing to explore further this issue on my  
own, provided there is no easy explanation, if someone is so kind to  
point me in the right direction....

Thanks in advance and best
sergio

Re: [Rubycocoa-talk] Can't generate bridge metadata

[Laurent S. <lsa...@ap...>]

On Aug 28, 2007, at 9:26 PM, Hans-Georg wrote:

> Am 28.08.2007 um 21:11 schrieb Laurent Sansonetti:
>
>>  Did you install any build of WebKit on your Tiger
>> machine, by any chance?
>
> I installed the beta of Safari 3.0.3.

Okay, these are unfortunately not supported, yet. In the meantime, you  
can either revert to the Tiger version, or hack the build.rb script to  
skip the WebKit metadata generation.

Laurent