RE: JCVSlet bug...

[Brian Nelson <brian@PocketScience.com>]

Erp...  I certainly didnt mean to come off as thought I was demanding
anything, but it was merely a suggestion.   BTW:  This is a great product,
and I thank you for putting it out there.

-----Original Message-----
From: Tim Endres [mailto:time@ice.com]
Sent: Friday, August 20, 1999 3:49 PM
To: Brian Nelson
Cc: jcvs@gjt.org
Subject: Re: JCVSlet bug...


On Fri, Aug 20 1999, Brian Nelson wrote:
Brian> Seems that the modification time, when viewing the contents of a
tree, are
Brian> the OS file date tag, and not the actual modification time in cvs.
Perhaps
Brian> it should be reading the modification time from the cvs files, and
not the
Brian> date of the file itself?

Hi Brian,

The mod time is of the local file that is in the working directory. The
reason
for this is that is the only time that I have at that point. I have not yet
queried the file for its history, and the Entries file only contains the
local
mod time. However, when you click on the file to get the history page, you
will
see the correct timestamp on each revision.

I am not sure that I want to add the overhead of the history query, as that
can hit the (possible very slow) wire, whereas the Entries file is local.

At some point, I will be adding the code to cache the history information
whenever it is retrieved. The difficulty is, however, that I would then have
correct information for some, but possibly not all, files. That seems too
confusing, so I am likely to leave the code as is. The current timestamp
does
actually have merit, in that it reflects the last time the file was updated
locally.

Thanks for the report,
tim.
Tim Endres  -  time@ice.com
ICE Engineering, Inc.  -  http://www.ice.com/
"USENET - a slow moving self parody." - Peter Honeyman