Path: utzoo!attcan!uunet!mcvax!hp4nl!philmds!leo
From: leo@philmds.UUCP (Leo de Wit)
Newsgroups: comp.sys.atari.st
Subject: Re: Resource / fsel_input() problem
Message-ID: <913@philmds.UUCP>
Date: 7 Jan 89 11:14:45 GMT
References: <908@philmds.UUCP> <826@laura.UUCP>
Reply-To: leo@philmds.UUCP (Leo de Wit)
Organization: Philips I&E DTS Eindhoven
Lines: 48

In article <826@laura.UUCP> klute%trillian.irb@unido.UUCP (Rainer Klute) writes:
|In article <908@philmds.UUCP> leo@philmds.UUCP (Leo de Wit) [that's me] writes:
|>Using the resource functions (notably rsrc_load(), rsrc_gaddr(),
|>rsrc_free()) I can load resources that, for instance, contain object
|>trees like  forms and menus.  Nice.
|>
|>Using fsel_input() I have a standard way for the user to search and
|>specify a pathname (directory and filename). Also nice.
|>
|>But trying to use both in the same program failed; when an fsel_input()
|>was done after a rsrc_load() the resource tree contained garbage.
|>It looks as if fsel_input() uses the rsrc_load() function itself, and
|>since a program can only have one resource file loaded at a time (am I
|>correct ?), it put its own resource on top of mine.
|
|I'm quite surprised to hear about such a behavior of your
|program. I use the same sequence (rsrc_load, fsel_input) over
|and over again in my programs and *never* encountered such a
|bug. I guess it is a bug in your own code.

Trying and retrying with new and old versions of my program (the new
version did NOT have the problem) I found the cause of the trouble:

The initial startup code that Lattice C (the compiler I'm using) adds
to a program, ensures that the program uses all but 4K of memory. This
is the default; it can be overruled by specifying a value for the
(global) variable _mneed, which sets the memory needed explicitly, or 
by supplying a commandline parameter for the program of the form
%workspace.

The old version (the one with the problem) used the default memory
scheme ("all but 4K"). The resource file I was using was about 2K.  If
I'm correct, resource files are loaded into remaining memory, so that
was the cause of the trouble (4K - 2K == 2K remains, what proved to be
too little for whatever tricks fsel_input had to perform).
(Funny detail: when I used the default memory scheme on the new
version, fsel_input returned immediately).

Moral: always ensure that your program using resource files leaves
enough remaining memory for these files to get loaded (and some 4K
extra too). Perhaps it had been better if rscr_load refused to load
into such a small remaining amount of memory, or fsel_input complained
about lack of memory. Well, you can't have it all...

Sorry if I caused any confusion on the net, but I think this is
a useful lesson anyway. Cheers!

                                  Leo.