See the following threads in the archives:
http://www.escribe.com/computing/pcpqa/index.html?by=OneThread&t=OS%204.0dr4
%20-%20DataMgr.c%2C%20Line%3A6915%2C%20Record%20isn%27t%20busy
http://www.escribe.com/computing/pcpqa/index.html?by=OneThread&t=DmGetRecord
%20vs%20DmResizeRecord%20handle%20%28again%29
All this trouble over one little assert. That'll teach me... :)
--
Danny Epstein
OS Engineer, Palm Inc.
"James Preston" <[EMAIL PROTECTED]> wrote in message
news:42372@palm-dev-forum...
> I just tried testing my application with the 4.0 ROMs, and I've hit an
oddity
> that's either a bug in 4.0 or my misunderstanding. I'm hoping someone can
> clue me in.
>
> Here's the part of my application that's in question:
>
> VoidHand nameRecHan = DmResizeRecord(gNameDB, recIdx, sizeof(NameInDB) +
StrLen(newName));
> if (nameRecHan) {
> NameInDBPtr nameRecPtr = MemHandleLock(nameRecHan);
> DmStrCopy(nameRecPtr, offsetof(NameInDB, projectName), newName);
> MemPtrUnlock(nameRecPtr);
> DmReleaseRecord(gNameDB, recIdx, true);
> }
>
> This worked fine in 3.5, but with 4.0, the emulator throws an error at the
> call to DmReleaseRecord that the record is not busy. I've scoured the API
> docs, and it does not explicitly say that DmResizeRecord sets the busy
bit,
> but I had thought that that was the case. Have I been mistaken all this
time?
> Given that DmResizeRecord returns a handle to the new record, it doesn't
> really make sense to me that it wouldn't mark it busy. I mean, I would
think
> that the vast majority of the time, when you resize a record you're going
to
> want to write into it next.
>
> The only way I've found to fix the above is to add a call to DmGetRecord
> before the resize, in order to force the busy bit to be set, like this:
>
> VoidHand nameRecHan = DmGetRecord(gNameDB, recIdx);
> nameRecHan = DmResizeRecord(gNameDB, recIdx, sizeof(NameInDB) +
StrLen(newName));
> if (nameRecHan) {
> NameInDBPtr nameRecPtr = MemHandleLock(nameRecHan);
> DmStrCopy(nameRecPtr, offsetof(NameInDB, projectName), newName);
> MemPtrUnlock(nameRecPtr);
> DmReleaseRecord(gNameDB, recIdx, true);
> }
>
> But that seems like a wasted call.
>
> So is DmResizeRecord supposed to set the busy bit, and this is really a
bug
> in 4.0, or have I been writing buggy code all this time and just got lucky
> up until now? And if that's the case, is there a better way to do this
than
> to make that extraneous call to DmGetRecord?
>
> Thanks,
> --Jim Preston
>
>
>
>
--
For information on using the Palm Developer Forums, or to unsubscribe, please see
http://www.palmos.com/dev/tech/support/forums/
