Peter Castine

extern-works-on-intel-not-on-ppc-runtime

Regarding the crash report - I was getting something pretty similar building a standlone on PPC using externals that were NOT in the C74 folder ... the build process reported copying the external into the collective, but actually wasn't, or was mangling it in the process. Any help?

This seems possibly related to another question I planned to ask:

The externals I've hacked generally perform a bundle-load of some kind, 

and I'm now in the process of thinking about UB versions of them. The 

loading technique I had been using revolves around using 

CFBundleLoadExecutable() and finding the appropriate entry-points. I had 

a student running [rtcmix~] under UB 4.6 which seemed to be finding 

strange entry-points (like the 'bang'-handler caused another specific

message-handler to be invoked, weird...), and I'm wondering if the 

Xcode/mach-o versions of these should be done using NSLinkModule()instead? 

How compatible/uncompatible is this with the Future of Max/MSP? Could 

Peter's woes be attributable to some purgatory of dynamic loading we are 

The externals I've hacked generally perform a bundle-load of some kind, 
and I'm now in the process of thinking about UB versions of them.  The 
loading technique I had  been using revolves around using 
CFBundleLoadExecutable() and finding the appropriate entry-points.  I had 
a student running [rtcmix~] under UB 4.6 which seemed to be finding 
strange entry-points (like the 'bang'-handler caused another specific
message-handler to be invoked, weird...), and I'm wondering if the 
Xcode/mach-o versions of these should be done using NSLinkModule()instead? 
How compatible/uncompatible is this with the Future of Max/MSP?  Could 
Peter's woes be attributable to some purgatory of dynamic loading we are 
presently inhabiting?

I'm not sure if my experience applies to Brad's and John's problem, 

David took the time to look at my object and noticed that "Show 

Package Contents" appeared to give an empty folder. It transpired 

that the Bundle's Contents/ folder had its Invisible bit set. This 

seems to cause problems on some processor/OS combinations.

This is probably not what's plaguing other people, but it is a point 

worth checking. It can be corrected with the SetFile command 

(assuming you have Apple's Developer Tools installed).

On 24-Sep-2006, at 20:16, Peter Castine wrote:

> Situation: build a standalone or collective using, say, lp.scampf 

> on Intel Mac w/Max 4.6 and run it on Intel Mac. Everything works 

> fine. Take the same collective to a PPC Mac with Max 4.6 Runtime. 

Peter Castine +--> Litter Power & Litter Bundle for Jitter

Interface Building for |home | chez nous|

Max/MSP Extremely cool |bei uns | i nostri|

I'm not sure if my experience applies to Brad's and John's problem,  
but just to let you know:

David took the time to look at my object and noticed that "Show  
Package Contents" appeared to give an empty folder. It transpired  
that the Bundle's Contents/ folder had its Invisible bit set. This  
seems to cause problems on some processor/OS combinations.

This is probably not what's plaguing other people, but it is a point  
worth checking. It can be corrected with the SetFile command  
(assuming you have Apple's Developer Tools installed).

On 24-Sep-2006, at 20:16, Peter Castine wrote:
> Situation: build a standalone or collective using, say, lp.scampf  
> on Intel Mac w/Max 4.6 and run it on Intel Mac. Everything works  
> fine. Take the same collective to a PPC Mac with Max 4.6 Runtime.  
> All Litter objects are listed as bogus
[schnipp]

--------------    http://www.bek.no/~pcastine/Litter/    -------------
Peter Castine             +--> Litter Power & Litter Bundle for Jitter
                                Universal Binaries on the way
iCE:  Sequencing,  Recording &
       Interface  Building  for                   |home    | chez nous|
       Max/MSP   Extremely cool                   |bei uns |  i nostri|
       http://www.dspaudio.com/                   http://www.castine.de

> I'm not sure if my experience applies to Brad's and John's problem, but 

I reported the collective problem already, it seems there is some 

conceptual problem of how collectives have to look like (which might not 

affect standalones, as the come with a complete set of externals and are 

What a mess. You would need to include three sorts of externals to make 

it run, or would need to have three different collectives...

Maybe C74 is already working on a solution... But thats tricky...

Peter Castine wrote:
> I'm not sure if my experience applies to Brad's and John's problem,  but 
> just to let you know:

I reported the collective problem already, it seems there is some 
conceptual problem of how collectives have to look like (which might not 
affect standalones, as the come with a complete set of externals and are 
bound to a certain Platform).
What a mess. You would need to include three sorts of externals to make 
it run, or would need to have three different collectives...

-- 
Stefan Tiedje------------x-------
--_____-----------|--------------
--(_|_ ----|-----|-----()-------
-- _|_)----|-----()--------------
----------()--------www.ccmix.com


Extern works on Intel, not on PPC Runtime