In article <93787@palm-dev-forum>, Robert McKenzie
<[EMAIL PROTECTED]> wrote:

> You should note that if you want to store your own IDs in the Palm record
> header that:
> 
> (1) Current Palm record IDs are 24-bit, non-zero numbers, NOT 32 bit numbers.
> (2) This is subject to change.
> 
> The only speed savings attained by puting IDs in the header is in finding
> records by IDs.  However, dmFindRecordByID uses a top down linear search,
> thus is (for large DBs) pretty much pig slow anyway...
> 

Based on this, and other recent posts, I think I am back to square one. 
I am thinking that for future compatibility the best thing may be to
use my own unique ID system.


In article <93784@palm-dev-forum>, Stringer <[EMAIL PROTECTED]> wrote:
> They are not useful for a persistent record pointer, but
> there are other techniques which work well for that.
> But they do assure a unique record ID while your program
> is running.
> 
> Roger Stringer
> Marietta Systems, Inc.
> 


Can you give any suggestions as to these "other techniques"?  Anything
better than simply adding a custom unique ID number to each record?

Thanks again for all the help.

-Sean

-- 
For information on using the Palm Developer Forums, or to unsubscribe, please see 
http://www.palmos.com/dev/support/forums/

Reply via email to