[MPlayer-dev-eng] [REPOST] [PATCH] Mac OS X fixes + Darwin AltiVec detection

[Dan Christiansen]

On Thursday, April 3, 2003, at 08:09 PM, Alex Beregszaszi wrote:

> Hi,
>
>> Yes, I believe it's needed since QuickTime depends on it. IIRC, I also
>>
>> had to include Carbon since some of the items present in QuickTime for
>>
>> Windows are part of Carbon on Mac OS X. (Apple ported a small subset
>> of Mac OS to Windows along with QuickTime.) Another reason for having
>> it is that simply checking for QuickTime could lead to false
>> positives. AFAIK, no CoreFoundation API is used within MPlayer, so
>> it's irrelevant.
> osdep/timer-macosx.c uses that CF stuff

Yeah, I just noticed :)

>
>> About the -framework switch; no it isn't Objective-C only. A framework
> ok
>
> attached a patch, please test that, if it's ok, i will commit it, now 
> :)

Your patch works fine, but I have some questions:
- You removed "#include <Carbon/Carbon.h>" from the test. Is there a 
particular reason for this? Simply checking for it's presence would 
ensure that there would be no false positives on another system with 
QuickTime installed. Another option would be to check for 
CoreFoundation instead. This could be done with:
	CFRunLoopRunInMode(kCFRunLoopDefaultMode,  0, false);
- You removed the $_codecmodules part. There is feedback for almost 
every other imaginable group in configure, so would having it for this 
one as well make sense?
- You kept " -x objective-c". Unless checking for Cocoa, there isn't 
really any reason for it to be there...
- You're patch won't add "#undef MACOSX" to config.h whenever Mac OS X 
support isn't enabled. Isn't it the normal practise to always have it?

- Dan