[Openmcl-devel] loading ccl-ide-init

[Paul Krueger]

Yep, I believe all that. It wasn't until I started to look at why my IDE init file was no longer doing what it did previously that I figured this out. It loaded fine when I manually did it immediately after starting the IDE, so I looked at where it was being loaded and then researched the #/applicationWillFinishLaunching: method. The latest documentation is explicit that it is called prior to initialization of the Application object and I don't remember reading that previously, so I'm guessing you're absolutely correct that the functionality changed somewhere along the line. If you're going to change this, you'll have to figure out whether you also want to move the other stuff that is done there now, but I'm betting you might. It should be a pretty easy change I expect.

If you think this change will be coming soon I'll wait for it, otherwise I'm tempted to just change "Will" to "Did" and recompile to make my stuff work again.

Thanks for the analysis.

On Oct 31, 2012, at 6:46 PM, Gary Byers <gb at clozure.com> wrote:

> As far as I can tell/remember, loading the IDE init file has always been done
> in the application delegate's #/applicationWillFinishLaunching: method (instead
> of in #/applicationDidFinishLaunching:).
> 
> IIRC, that was about the earliest point in time that it was practical to load that file: the NSApplication object was in fact (largely) initialized
> and at least in theory loading the file then made it possible to customize
> what happens between the points when those notifications are posted.
> 
> The code to load the IDE init file was added in early May of 2009; it's
> certainly possible that the state of the NSApplication at the point of
> the earlier notification isn't what it was ~3.5 years (and ~3 OS releases)
> ago because of Cocoa changes.  If that's true, it seems to argue in favor
> of loading the init file later (when we know that the application DID finish
> launching and is fully initialized), since we can't really count on how
> much initialization remains to be done when we're told that it WILL "finish
> launching."
> 
> 
> On Wed, 31 Oct 2012, Paul Krueger wrote:
> 
>> Just upgraded to version 1.9-dev-r15490M-trunk  (DarwinX8664) and discovered that loading the ccl-ide-init file has been put into the #/applicationWillFinishLaunching: method. I don't know when that happened because it's been a LONG time since I upgraded last. The problem with this is that the things that I have previously been doing in my ccl-ide-init require that the Application object already be initialized, which is not the case when this method is called. Is there some reason loading the ccl-ide-init file couldn't be put into a #/applicationDidFinishLaunching: method which is called after the Application object is initialized?
>> 
>> _______________________________________________
>> Openmcl-devel mailing list
>> Openmcl-devel at clozure.com
>> http://clozure.com/mailman/listinfo/openmcl-devel
>> 
>>