>From the docs, the only difference between the DmGetRecord and DmResizeRecord
is the fact that DmGetRecord sets the busy bit (apart from the resizing
question, of course). The busy bit is used for two things:
1) If a record is busy (busy bit set), then you can't call DmGetRecord on
it. It will fail.
2) Forces you to call DmReleaseRecord, allowing you to set the dirty bit,
which can be used upon synchronization.
Apart from that, the handle will always be valid, and consequently the
pointer will be valid while the handle is locked.
Danko Radic wrote:
> I'm posting this question again because I still got no answer. I can't
> belive that nobody knows about it. On the other hand, I apologize if the
> question is that stupid that it couldn't attract anyone's attention, but
> it's not the worst thing that was posted to this forum for sure...
>
> Regards,
>
> -DR
>
> #######
>
> When I have to resize record, I use DmResizeRecord. It returns handle to
> data chunk. Now, is it safe to lock that handle and do DmWrite business
> with such pointer the same way I do when I lock handle gotten by
> DmGetRecord? What I mean is that DmResizeRecord doesn't set busy bit (at
> least I didn't find it in OS sources) like DmGetRecord and im not sure if
> the pointer is valid all the time.
> If not, would it be safe to remember record uniqueID, resize it, find
> index by DmFindRecordByID and then use DmGetRecord and MemHandleLock?
--
Sergio Carvalho
---------------
[EMAIL PROTECTED]
[EMAIL PROTECTED]
If at first you don't succeed, skydiving is not for you
