- From: Wilfredo Sánchez Vega <wsanchez@wsanchez.net>
- Date: Thu, 22 Dec 2005 17:24:51 -0800
- To: Julian Reschke <julian.reschke@gmx.de>
- Cc: Dan Brotsky <dbrotsky@adobe.com>, Jim Whitehead <ejw@soe.ucsc.edu>, WebDav WG <w3c-dist-auth@w3.org>
Well, if it's fixable it's fixable down to one millisecond instead of second. I'm not sure why the window is one second. I agree that would be better, but you still have a (shorter) window. I can ask on the httpd list to see if anyone knows why our window is one second when apr_time_t is in milliseconds... but I'll wait to do that after the holiday break. I'm betting there's some filesystem with a one second resolution out there... If it's an OS limitation (as opposed to a filesystem one) we can probably contain the damage to that OS. -wsv On Dec 21, 2005, at 7:23 AM, Julian Reschke wrote: > I'm not sure why that would be, but the code in http_etag.c > certainly uses a hardwired constant for this. Maybe this is > fixable, in which case the problem would go away. Wilfredo?
Received on Friday, 23 December 2005 01:25:06 UTC