On 11/2/99 3:50 PM Jeremy Wadsack ([EMAIL PROTECTED]) wrote:
>This is an old thread that I wanted to revive for an important note.
>
>Under IIS, Microsoft appears to log processing time in units of 1/100th
>of a second. Not in seconds, as it should be anmd not in seconds to the
>100th precision and Stephen suggested previously. This means reports of
>processing time running under IIS will appear to be 100 times longer
>than they should be. I don't know if Analog should make a special case
>for IIS (not that it can necessarily tell what server generated the
>logs) so all Processing Time reports are proper or if IIS users should
>just be aware of this when reading the Procesing Time report (i.e. a
>note to that effect in the docs).
>
>If anyone has seen differently on this I'd like to know.
I've been through quite a collection of different servers, though not
all, and I have never seen anyone using the W3C log token TIME-TAKEN
except Microsoft. Microsoft has always logged that token as 1/100ths of a
second. Of the major servers, only Microsoft IIS, WebSTAR, and Apache log
transfer time. WebSTAR uses TIME_TAKEN or TRANSFER_TIME, and neither is
included by default. Apache doesn't use W3C format and only logs seconds,
which is essentialy useless. That means that the majority of people
interested in the Processing Time report (really Transfer Time) are
getting the wrong numbers.
Given the incomplete nature of the W3C documentation, the spec. was never
completed even to rough draft level, I don't see any reason to use that
as the reference. Microsoft IIS is the only software using that
particular token, and we can expect that future implimentations will
follow Microsoft more often than the spec.
Jason
-----------------
[EMAIL PROTECTED]
-----------------
Dr. Seuss books . . . can be read and enjoyed on several levels. For
example, 'One Fish Two Fish, Red Fish Blue Fish' can be deconstructed
as a searing indictment of the narrow-minded binary counting system.
-- Peter van der Linden, Expert C Programming, Deep C Secrets
------------------------------------------------------------------------
This is the analog-help mailing list. To unsubscribe from this
mailing list, send mail to [EMAIL PROTECTED]
with "unsubscribe analog-help" in the main BODY OF THE MESSAGE.
List archived at http://www.mail-archive.com/[email protected]/
------------------------------------------------------------------------
