Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!lll-winken!lll-lcc!ames!hc!beta!rmc
From: rmc@beta.UUCP (R. Martin Chavez)
Newsgroups: comp.sys.mac.hypercard
Subject: Urgent XCMD/Hypercard Questions
Message-ID: <11956@beta.UUCP>
Date: Fri, 6-Nov-87 05:58:46 EST
Article-I.D.: beta.11956
Posted: Fri Nov  6 05:58:46 1987
Date-Received: Sun, 8-Nov-87 16:13:02 EST
Organization: Los Alamos Natl Lab, Los Alamos, N.M.
Lines: 44
Keywords: XCMD Hypercard MPW segmentation

I'm working against some deadlines here (who isn't), so I'll
be very grateful for timely responses.

(1) I have a rather large MacApp program that I'd like to use
as an XFCN with Hypercard.  Even with some consolidation at 
link time, I have around ten segments.  I couldn't possibly 
cram EVERYTHING (especially all the MacApp stuff) into one segment.

	a.  When I try to link with "-rt XFCN=O", I get the
	following error:

	### Link: Error  Linker error: uninitialized jump table offset =  -1

	Does this have something to do with multiple segments?
	I can't find anything in the APDA docs that outlaws multiple
	segments, but I know that DAs can only have one segment,
	so maybe the same restriction applies.

	(I also had a nasty problem with unresolved references to
	%_CLASSINFO in %_INIT2.  I fixed it by assuming a 128K Rom
	and re-writing %_INIT2 to load the proper table without 
	reference to %_CLASSINFO.)

	b.  Are XFCN's also prohibited from having global variable
	references?  If so, then MacApp seems to be out of the
	question.

(2) If it's impossible to use MacApp with the XCMD feature of
Hypercard (to tell the truth, I never expected such a wild & wonderful
thing!), I have two options.  The first: rewrite my code in raw
MPW.  NO WAY: I won't go back, even under torture.

The second: I could use the Multifinder.  I've already decided
to get that second meg for my Mac II, so $$$ isn't the issue.
But I'd REALLY like a way for one application (mine) to tell
another (Hypercard) to activate itself.  That's all I ask; the
rest (e.g., communication) can be hacked.  Is there a trap
with which one application can reactivate the application from which
it was sublaunched?

			Many thanks,

			R. Martin Chavez
			Knowledge Systems Laboratory
			Stanford University