On Sun, 2006-02-12 at 22:13 +0000, Malcolm Rowe wrote:
> Could anyone explain the rationale behind our treatment of
> schedule-deleted files encountered during 'svn diff'? (I realise that
> using 'rationale' and 'diff' in the same sentence may provoke some humour,
> but presumably there is some reason for the current arrangement).
>
> Specifically, we seem to treat such files as if they are empty, rather
> than as if they are absent. This is rather strange, since diffs once
> we commit will (of course) treat the file as absent.
>
> A reproduction recipe is below, which should hopefully make things a
> little clearer.
>
> I'd like to fix repos-BASE diffs so that they don't have this odd
> behaviour (and fix another bug in exactly the same area), but before I do
> that, I'd like to understand what the correct behaviour for diffs against
> WORKING is supposed to be, since that might affect the way I do this.
>
> [DannyB, I've cc:d you because you committed the original regression test,
> and so presumably you've got some idea about this.]
The only intention is that schedule delete files show as a diff from the
file to empty, which is what should be happening (because the entire
file is being deleted).
This matches expected user behavior (I received a large number of
complaints when it used to treat them as absent not to mention it caused
*errors* from diff).
Why do you believe deleting the file should not show up at all in diff?
If you apply the diff that results using patch, it won't DTRT, for
example.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Mon Feb 13 00:15:17 2006