Stephen wrote:
>On Mon, 15 May 2000, Aengus Lawlor wrote:
>>
>> To be fair, it's a Working Draft, not a spec.
>
>Oh, I agree. But that's no reason to ignore it without a reason.
They had a reason. Being Microsoft, they just didn't explain it to
anyone :-)
>> It's a pity Microsoft didn't document why the ignored the draft (the use
>> of milliseconds ... are improvements, in my opinion).
>
>I disagree here. Putting everything in seconds with decimals allows
>different servers to choose different granularities. One can log in
>milliseconds, one in centiseconds, and one in seconds, if they
>want.
The use of milliseconds doesn't preclude measurements in
centiseconds or seconds (just add a couple of zeroes). I suppose
that it's theoretically possible that someone might want to log in
sub-millisecond units, but I'm not sure why (sub-millisecond
responses might be practical, but would you want to log
sub-millisecond times). I'd also assume that logging a integer value
is less expensive than converting to a float.
Anyway, this explains why Analog requires the #Software: header, which
confused me when I was playing with some test logs.
Aengus
------------------------------------------------------------------------
This is the analog-help mailing list. To unsubscribe from this
mailing list, send mail to [EMAIL PROTECTED]
with "unsubscribe" in the main BODY OF THE MESSAGE.
List archived at http://www.mail-archive.com/analog-help@lists.isite.net/
------------------------------------------------------------------------
