In addition to the subtle problem with FrmReturnToForm(0), we've run
into one other strange problem with the 3.5 d4 roms posted for
quick-click (whatever...) download.

On 3.5 Color Release only, we have a sequence of code which is:

bmpPtr = MemHandleLock(bmpHandle);
RenderSomeDataToTheBitmap(bmpPtr);
WinDrawBitmap(bmpPtr, 0, 15);
MemPtrUnlock(bmpPtr);

POSE a3 and a4 report:  <application> 1.0 has just read directly from
memory manager data structures.  If you hit debug, you are in
MemPtrUnlock.  Continue seems fine and the app does not crash under
stress testing with the warning disabled.

Here is the interesting part.  Comment out "RenderSomeDataToTheBitmap" -
no change.  Comment out "WinDrawBitmap" - the error disappears.  And
this is only in 3.5 Color _release_.  Color debug, EZ debug/release,
non-Ez debug/release, 3.3 EZ debug/release, 2.0 debug/release, 3.1
release - all fine.

The 3.0 SDK sample Beamer exhibits this sample problem - under Color
Release only.  That sample allocates heap memory and creates its own
Bitmap.  Presumably, something did not get initialized correctly in the
bitmap header which freaks WinDrawBitmap on the color/release version
crazy.

If I could get the 3.5 SDK, I could see what changed in Beamer, or even
test for 3.5 and use the new create-bitmap function I've seen people
refer to...  I'm trying to be patient, but between waiting weeks and
weeks for paperwork to get processed and waiting weeks and weeks for a
working CW (the primary reason we upgraded to R6 was for new resource
support)...

Thanks in advance for the help,
-jjf


-- 
For information on using the Palm Developer Forums, or to unsubscribe, please see 
http://www.palm.com/devzone/mailinglists.html

Reply via email to