On Dec 13, 2005, at 6:05 AM, Hannes Gredler wrote:
Guy Harris wrote:
Do we want relative time stamps (-ttt, for secs/usecs since
previous packet, and -ttttt, for secs/usecs since first packet) to
be printed as HH:MM:SS.UUUUUU, or seconds.UUUUUU?
in the sense of having an indicator wether this is a relative/or
absolute timestamp ?
No, just in the sense of whether to print seconds, or
hours:minutes:seconds, before the decimal point.
saw that and i deliberatly decided against it i.e.
to print HH:MM:SS.UUUUUU as it IMHO makes things easier to read for
long-term traces ...
...although the person who asked for it said
Guy, I would personally like to see 124.093988 where 124 ==
seconds. I suppose one could also
easily add a switch which converts this to the HH:MM:SS format :)
although he might not have been thinking of long-term traces.
I suppose we could get really ambitious and support strftime()-like
formats ("strftime()-like" because, for relative times, you don't
have any date fields, just time fields).
-
This is the tcpdump-workers list.
Visit https://lists.sandelman.ca/ to unsubscribe.
