Re: [viewvc-users] Only *some* diffs give exception on Windows

"C. Michael Pilato" <[email protected]>
Newsgroups gmane.comp.version-control.cvs.viewcvs.user
Organization CollabNet, Inc.
Message-ID <[email protected]>
Jeff Hoover wrote:
> I have:
> CollabNetSubversion-server-1.6.2-1.win32 - Apache, not svnsvr
> python-2.5.4
> pywin32-213.win32-py2.5
> diffutils-2.8.7-1 - installed at C:\GnuWin32\bin and in PATH
> 
> When I am at a page like this:
> http://<ipaddr>/viewvc/repository/trunk/proj/ExampleFile1.txt?view=log
> 
> and I click "Diff to previous XX", *some* links take me to a successful Colored Diff page, while others give me the exception below. 
> When a link fails, I can successfully diff each of the 2 revisions involved in the failed link against different revisions, so I don't think either revision is corrupt.
> 
> For example, a link that fails is:
> http://<ipaddr>/viewvc/repository/trunk/proj/ExampleFile1.txt?r1=21&r2=22
> but these succeed:
> ...r1=20&r2=21
> ...r1=22&r2=23
> ...r1=23&r2=22
> ...r1=21&r2=20
> 
> Interestingly, if I change the url by hand to swap the order of the revisions in the offending link, the swap-ordered link works. Following my previous example:
> ...r1=22&r2=21
> is OK.
> 
> I thought the exception below is typically the result of python not being able to find "diff" in the PATH, but why do *some* of my links diff properly?
> 
> Any thoughts?

At the risk of offering positively nothing of real value to this thread, I'd
just like to a) confirm that another human has read your mail, and b) said
human thinks your situation is really really weird.  That's not really a
game-changing bit of feedback, I know, but I guess in some ways it might be.
  Some questions I'd have:

1.  Are you using ViewVC's cache control logic?  If so, set
http_expiration_time=0 and generate_etags=0 for the purposes of further
testing, and clear your browser's cache.  My other questions assume we don't
have browser caches to worry about while debugging.

2.  Are the successes/failures self-consistent?  Does hitting r1=20&r2=21
*always* succeed no matter how many times you hit it?  Likewise, do the
failing URLs always fail?

3.  Is there any chance that you are exhausting (perhaps temporarily) some
limitable system resource, like the allowed number of open file handles or
somesuch?  (I'm not so much a Windows guy, so this question could be more
idiotic than throught-provoking.)

-- 
C. Michael Pilato <[email protected]>
CollabNet   <>   www.collab.net   <>   Distributed Development On Demand

------------------------------------------------------
http://viewvc.tigris.org/ds/viewMessage.do?dsForumId=4255&dsMessageId=2371012

To unsubscribe from this discussion, e-mail: [[email protected]].
signature.asc (application/pgp-signature, 197 B)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)

iEYEARECAAYFAkpbbvQACgkQokEGqRcG/W6e2gCgg7YM+C/Jl0GvjfdUXoAbT1q4
gKgAoILyt+5uFgUzkk7s2w/nwvEb6Z9P
=gZo6
-----END PGP SIGNATURE-----
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.