My database has typically 300 or so records where each records is from 100
byts to 65535 bytes long, with the typical size being around 10,000 bytes.
The first 20 or so bytes of each record is a table that describes what's in
the rest of the record, so when searching the database I often only need
Rich,
What happens when you incrementally take things out of the loop you
show below? I'm wondering because searching through 607 records in
my Address Book takes only a second or so on my Treo 600 when
searching for my name. I figure it probably uses a loop similar to
what you want, so perh
recHandle = DmGetRecord(dbHandle, index);
You can use DmQueryRecord (assuming you don't need to change the
record). That should be a tiny bit faster.
DmRecordInfo( dbHandle, index, NULL, &recID, NULL);
No idea how much time this takes up. Is there some way to code around it?
// do
You might want to rethink how you search the records, and binary search is
your friend but your key will need to be sorted.
I have a 15k records db and the average search is about, IIRC, 6ms.
elv
"Richard Coutts" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
> My database has typ
> What happens when you incrementally take things out of the loop you
> show below?
That was the problem -- other code was slow, not the record reading. I was
jumping to conclusions.
Thanks!
Rich
--
For information on using the Palm Developer Forums, or to unsubscribe, please see
http://www.
\I'm wondering if I'm doing something wrong or if there's a way to speed
things up. To load the first 20 bytes of each record, I loop through and
open each record like this:
for (index = 0; index < numRecs; index++)
{
recHandle = DmGetRecord(dbHandle, index);
Use DmQueryRecord here. It is a
Ben Combee a écrit :
> >\I'm wondering if I'm doing something wrong or if there's a way to speed
> >things up. To load the first 20 bytes of each record, I loop through and
> >open each record like this:
> >
> >for (index = 0; index < numRecs; index++)
> >{
> > recHandle = DmGetRecord(dbHand
What is the difference with DmGetRecord if after a Query, you can nonetheless
write the record and release it? Only the Busy bit setting?
Basically.
Regards,
Steve Mann
--
For information on using the Palm Developer Forums, or to unsubscribe, please see
http://www.palmos.com/dev/support/forums/
MemHandleLock is surprisingly slow function. I've benchmarked it many
times and got awful results - it was a times slower than everything else
in my program. I found no way to avoid it - even if I use the the
handle as a pointer to pointer (and the value it points is the actual
value I receive
> MemHandleLock is surprisingly slow function. I've benchmarked it many
> times and got awful results - it was a times slower than everything else
> in my program. I found no way to avoid it - even if I use the the
> handle as a pointer to pointer (and the value it points is the actual
> value I r
I suggest you do a shell or heap sorts - forget the bubble :)
I've made some mix between them and optimized the logic and arithmetic -
now 3 records are sorted in less than half a second on T|C :)
Richard Coutts wrote:
MemHandleLock is surprisingly slow function. I've benchmarked it many
tim
11 matches
Mail list logo