Hi,
I think that I am experiencing the same bug since several months.
Pedro's analysis looks convincing to me. I already noticed the
regular periodicity for the manifestation of the bug. Furthermore, it
occurs simultaneously on several machines, all running up-to-date
Etch. I once started to write down the dates at which the bug
manifests. The last dates were :
Nov 02 2007
Dec 21 2007
Feb 09 2008
Mar 30 2008
Note that there can be an error of one day or two, depending on when I
noticed the problem. The periodicity is not far from the 49days
17hours period of wraparound for a 32bits milliseconds counter.
Moreover, the dates roughly coincide with the expected dates of
wraparound, if the timestamp is a 32bits unsigned counter,
corresponding to the number of milliseconds since 1970-01-01 00:00:00
UTC, what it is AFAICT.
I however did not check Pedro's patch.
Regards.
--
Arnaud Giersch
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
