Re: bypass NSApp

2009-07-27 Thread Julien Isorce
t so then we are moving to the other solution: > bypass > > NSApp, > > making a classic NSRunLoop but add to it the ability to perform/handle > the > > events that the NSApp usually is the only one able to do. > > > > I am open to any solution, even if it requires t

Re: bypass NSApp

2009-07-10 Thread Stephen J. Butler
On Fri, Jul 10, 2009 at 6:50 PM, Julien Isorce wrote: > you are enterely right so then we are moving to the other solution: bypass > NSApp, > making a classic NSRunLoop but add to it the ability to perform/handle the > events that the NSApp usually is the only one able to do. > >

Re: bypass NSApp

2009-07-10 Thread Scott Thompson
sure but in the previous mails I tried to explain that I am making a plugin (which consists of a dynamic library) of a program which I have not the source code. OK.. the application for which you are writing the plugin... is it a Cocoa application, a Carbon application, or something else..

Re: bypass NSApp

2009-07-10 Thread Stephen J. Butler
On Fri, Jul 10, 2009 at 6:33 PM, Julien Isorce wrote: > So the final question: is there a way to make [NSThread isMainThread] return > YES in a thread different that the main thread of a process, on MacOSX ? We don't have the source for the Foundation Framework, but there is a pthreads pthread_mai

Re: bypass NSApp

2009-07-10 Thread Julien Isorce
t;> >> yup something like that. >> >> >>> >>> There may be a way to solve your problem properly in Mac OS X, >>> >> >> >> I experimented on MacOSX and it seems that a pthread (I mean a thread >> created using native C api) is

Re: bypass NSApp

2009-07-10 Thread Mike Abdullah
On 11 Jul 2009, at 00:33, Julien Isorce wrote: Hi, First, thx for your reply. 2009/7/10 Scott Thompson On Jul 10, 2009, at 3:40 PM, Julien Isorce wrote: In GNUstep it's possible to decide which pthread is the "main thread". I mean the first pthread which call GSRegisterCurrentThread, i

Re: bypass NSApp

2009-07-10 Thread Julien Isorce
Hi, First, thx for your reply. 2009/7/10 Scott Thompson > > On Jul 10, 2009, at 3:40 PM, Julien Isorce wrote: > > In GNUstep it's possible to decide which pthread is the "main thread". >> I mean the first pthread which call GSRegisterCurrentThread, is the main >> thread. >> >> And then the NSA

Re: bypass NSApp

2009-07-10 Thread I. Savant
On Jul 10, 2009, at 5:49 PM, Scott Thompson wrote: What exactly is it that are you trying to do on Mac OS X? What is the environment in which your code is running and what effect are you trying to achieve? And here we have the winning question (asked for the second time in this thread)

Re: bypass NSApp

2009-07-10 Thread Scott Thompson
On Jul 10, 2009, at 3:40 PM, Julien Isorce wrote: In GNUstep it's possible to decide which pthread is the "main thread". I mean the first pthread which call GSRegisterCurrentThread, is the main thread. And then the NSApp must be init and run in this main thread, as it's required on MacOSX.

Re: bypass NSApp

2009-07-10 Thread Julien Isorce
well ok, In GNUstep it's possible to decide which pthread is the "main thread". I mean the first pthread which call GSRegisterCurrentThread, is the main thread. And then the NSApp must be init and run in this main thread, as it's required on MacOSX. Is it possible to change the "main thread" as f

Re: bypass NSApp

2009-07-09 Thread Stephen J. Butler
On Thu, Jul 9, 2009 at 6:09 PM, Scott Thompson wrote: > To empasize this point. Consider, for example, the method > performSelectorOnMainThread:. We use this API in several place to ensure > that functionality is performed on the same thread that is processing user > events. If code were able to

Re: bypass NSApp

2009-07-09 Thread Scott Thompson
Sent from my iPhone On Jul 9, 2009, at 10:04 AM, Bill Bumgarner wrote: On Jul 9, 2009, at 2:46 AM, Julien Isorce wrote: I am trying to isolate this difference in order to avoid to run the main loop in the main thread. Don't bother. Cocoa is quite explicitly designed to not support ev

Re: bypass NSApp

2009-07-09 Thread Bill Bumgarner
On Jul 9, 2009, at 2:46 AM, Julien Isorce wrote: I am trying to isolate this difference in order to avoid to run the main loop in the main thread. Don't bother. Cocoa is quite explicitly designed to not support event processing on any thread other than the main thread. No amount of "tr

Re: bypass NSApp

2009-07-09 Thread Julien Isorce
Hi, Thx for the reply. First I think you did not understood my questions. In other platform, it's possible to run the GUI main loop in any thread. On macosx, there is a difference between the main run loop and the other NSRunLoop. I am trying to isolate this difference in order to avoid to run th

bypass NSApp

2009-07-08 Thread Julien Isorce
Hi, I am in a case where I have no access to the main thread since I am making a plugin which consists of a dynamic library loaded by a third party program. This third party program has nothing to deal with Cocoa and only uses standard C API. And it has it's own main loop. In the plugin I create