Re: Timestamps

[Jim Langston <jimlangston@xxxxxxxxxxxxxxxx>]

Eric,

Did you read the discussion on how a UUID is generated?  Curiously
enough, part of the UUID is a timestamp.  And there is discussion
that it is possible to have a UUID with the same value if you get
two values before the clock cycles.  And that you then have to 
increment a value.  It is very interesting readying, and in OS400's
case those 3 zeros at the end could be incremented.

So, basically, if you decide to generate a UUID instead of using a
time stamp, you are back to the exact same problem and still have to
find a work around.

Regards,

Jim Langston

"Eric N. Wilson" wrote:
> 
> Hello,
> 
> Using the API is a bit of overkill in most situations. Doing the Time
> opcode into a timestamp field gives you everything but the last 3
> digits of the timestamp... Which I believe to be the same result as
> the API.
> 
> Using a logical file with the time stamp omitted is the only way I
> know of to actually get the last three digits of the timestamp...
> 
> Even if a person were able to get the last three digits of a timestamp
> there will come a day when more than one record will be able to be
> added in that interval (Trust me it will especially if you rely on a
> timestamp to be your unique key). There has been discussion on the
> validity of using timestamps and methods of retrieving them recently
> and should be accessible via the search facilities.
> 
> A approach that may be more viable is to use a UUID as has also been
> discussed recently. The value is not pretty to the eyes, but unique
> keys should not have any embedded meaning (other than to uniquely
> identify a record) in my opinion.
>
+---
| This is the RPG/400 Mailing List!
| To submit a new message, send your mail to RPG400-L@midrange.com.
| To subscribe to this list send email to RPG400-L-SUB@midrange.com.
| To unsubscribe from this list send email to RPG400-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---