Ben Collins-Sussman wrote:
>
> On Aug 8, 2005, at 7:00 AM, A.T.Hofkamp wrote:
>
>>
>> + svn mv --force x y
>> svn: Cannot copy or move 'x': it's not in the repository yet; try
>> committing first
>>
>> Is this a known problem?
>
>
>
> Sure, because 'svn mv' is just 'svn cp; svn rm'. You cannot 'svn cp' a
> schedule-add file. Instead, 'svn revert' the schedule-add directory,
> 'mv' it with unix, then 'svn add' it again.
Nice solution, I didn't consider that.
From an implementation point-of-view, your solution makes sense. However,
that implies that users should be aware of the inner workings of svn,
something that I hesitate to accept. Imho a tool should work transparently, ie
the following (user orientated) reasoning should hold imho:
svn owns x, svn allows moving, and thus I should be able to mv x to y at any
time I like.
That the move only needs to be performed in the working copy should not be a
user problem imho.
Albert
PS in that respect, I don't understand why 'svn cp' would not be possible
here. Is there some reason why svn cannot create a copy of x in the working
copy (ie is there something fundamental against performing 'cp x y' in unix,
then 'svn add y'?)
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Mon Aug 8 15:43:44 2005