On Fri, May 06, 2005 at 05:56:19PM +0200, Lele Gaifax wrote:
> >>>>> "Nathan" == Nathan Gray <[EMAIL PROTECTED]> writes:
>
> Nathan> While pulling from CVS to darcs, using tailor.py, I came
> Nathan> across this error. I believe it is because a commit has
> Nathan> been applied to CVS with a date prior to my last tailor
> Nathan> sync.
>
> Uhm, no, tailor itself does not care about the sequence of patches.
Ah, good.
> Nathan> e = CvsEntry(entry) File
> Nathan> "/var/projects/cvsync/vcpx/cvs.py", line 351, in __init__
> Nathan> y,m,d,hh,mm,ss,d1,d2,d3 = strptime(ts, "%a %b %d %H:%M:%S
> Nathan> %%Y") File "/usr/lib/python2.3/_strptime.py", line 425, in
> Nathan> strptime raise ValueError("time data did not match format:
> Nathan> data=%s fmt=%s" % ValueError: time data did not match
> Nathan> format: data=dummy timestamp fmt=%a %b %d %%H:%M:%S %%Y
>
> The problem here is that CVS put something strange ("dummy" in the
> particular case) in its CVS/Entries, where usually there's a
> timestamp. Could you check the related file, and send it to me?
Aha! Found it. Just as you say, the CVS/Entries file in one directory
contained three files with 'dummy' dates:
/MANIFEST/0/dummy timestamp//
/META.yml/0/dummy timestamp//
/Makefile.PL/0/dummy timestamp//
Investigating further, I find that these files were somehow added (cvs
add) in my tailor working directory, but not committed into CVS.
As soon as I ran a manual 'cvs commit', tailor was able to sync again.
Thanks for your prompt (and accurate) insights!
-kolibrie
_______________________________________________
darcs-users mailing list
[email protected]
http://www.abridgegame.org/mailman/listinfo/darcs-users
