Re: %TIMESTAMP renders only milliseconds?

[Booth Martin <booth@xxxxxxxxxxxx>]

But how many of the decimal characters were non-zero?  I did a test to 
see if I could get more than 4 valid decimal places using every on-board 
method of establishing current date/time that I could think of.

I could add seconds, like %timestamp(*SYS: 12) + 
%seconds(0.000056789012), and that worked just fine, but I could not 
seed current date with valid micro seconds beyond 4 places.


On 12/7/2019 2:01 PM, Bruce Vining wrote:
> Just for fun I threw together a test case of SQL current timestamp vs
> QWCCVTDT (*YYMD, 20-byte receiver).  Both were accessing microseconds and
> the loop for each was 1,000,000 iterations.  The API "won" every time
> though not by a whole lot.  The percentage longer using the SQL route  was
> 13%, 7%, and 7% over the API approach.
>
> Note all the typical disclaimers of the system not being dedicated, etc.
> I'll also add that the*very first time*  I ran the SQL part of the program
> it took substantially longer and so I did not include that very first run
> in the above timings.  The percentages are based on subsequent runs of the
> program.