On Wed, Jul 05, 2000 at 08:46:21AM +0200, Markus Niebel wrote:
> I played and tested a little bit more - it seems to be
> caused by the logic of cvs itself: it relies on filetimes.
> When I change the time for dir1/File1.hpp everything works
> fine. As I noted, the two files was committed exactly at
> the same time, and (I think this is the problem) the
> revision dates of dir1/File1.hpp and
> dir1/subdir1/File1.hpp are exactly the same. I remember a
> discussion some weeks ago about using checksums instead of
> filetimes or in addition to filetimes.
Ok, now I finally got your point. That the two files have
the same timestamp is not the problem, but the fact, that
your copying preserved the original timestamp of
subdir1/File1.hpp. Therefore the timestamp of
dir1/File1.hpp, which is kept in dir1/CVS/Entries is the
same as the one of the just copied!
If you use a normal copy command or just edit a file, the
new timestamp will differ to that one in CVS/Entries.
With Linux you can use '-p' flag to the 'cp' command to
preserve the attributes of a file, including their original
timestamp, but default is to create a new one for the copy.
So, yes, there are unreliable cases when relying on
timestamps only.
Regards,
Matthias
--
Matthias Kranz [EMAIL PROTECTED]
http://www.belug.org/~kranz
"Ever tried. Ever failed. No matter. Try again.
Fail again. Fail better." (Samuel Beckett)
