You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by to...@gmail.com on 2006/09/09 01:48:19 UTC
It's possible to create a completely blank check-in (in 1.3.2)
C:\test>svn co file:///c:/svn/trunk .
A b.txt
A a.txt
Checked out revision 30.
C:\test>svn rm file:///c:/svn/trunk/b.txt -m "Deleted on server side."
Committed revision 31.
C:\test>svn rm b.txt
D b.txt
C:\test>svn ci -m "Deleted on client side."
Deleting b.txt
Committed revision 32.
C:\test>svn log -r 32:31 -v file:///c:/svn/
------------------------------------------------------------------------
r32 | Tom | 2006-09-09 02:40:05 +0100 (Sat, 09 Sep 2006) | 1 line
Changed paths:
Deleted on client side.
------------------------------------------------------------------------
r31 | Tom | 2006-09-09 02:39:52 +0100 (Sat, 09 Sep 2006) | 1 line
Changed paths:
D /trunk/b.txt
Deleted on server side.
------------------------------------------------------------------------
C:\test>svn di -r 31:32 file:///c:/svn/
C:\test>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: It's possible to create a completely blank check-in (in 1.3.2)
Posted by Garrett Rooney <ro...@electricjellyfish.net>.
On 9/8/06, tommyjb@gmail.com <to...@gmail.com> wrote:
> Also note that 'moving' behaves strangely in this kind of situation
> too -- in the following, the 'move' made in the WC seems to change
> into a 'copy':
That's because a "move" in svn is currently just a copy + a delete,
and if the delete has already occurred, all you've got left is a copy.
-garrett
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: It's possible to create a completely blank check-in (in 1.3.2)
Posted by to...@gmail.com.
Also note that 'moving' behaves strangely in this kind of situation
too -- in the following, the 'move' made in the WC seems to change
into a 'copy':
C:\test>svn mv file:///c:/svn/trunk/a.txt file:///c:/svn/trunk/foo.txt
-m "Moved on server side."
Committed revision 33.
C:\test>svn mv a.txt bar.txt
A bar.txt
D a.txt
C:\test>svn ci -m "Moved on client side."
Deleting a.txt
Adding bar.txt
Committed revision 34.
C:\test>svn log -r 34:33 file:///c:/svn/ -v
------------------------------------------------------------------------
r34 | Tom | 2006-09-09 02:52:07 +0100 (Sat, 09 Sep 2006) | 1 line
Changed paths:
A /trunk/bar.txt (from /trunk/a.txt:32)
Moved on client side.
------------------------------------------------------------------------
r33 | Tom | 2006-09-09 02:51:05 +0100 (Sat, 09 Sep 2006) | 1 line
Changed paths:
D /trunk/a.txt
A /trunk/foo.txt (from /trunk/a.txt:32)
Moving on server side.
------------------------------------------------------------------------
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: It's possible to create a completely blank check-in (in 1.3.2)
Posted by Garrett Rooney <ro...@electricjellyfish.net>.
On 9/8/06, tommyjb@gmail.com <to...@gmail.com> wrote:
> C:\test>svn co file:///c:/svn/trunk .
> A b.txt
> A a.txt
> Checked out revision 30.
>
> C:\test>svn rm file:///c:/svn/trunk/b.txt -m "Deleted on server side."
>
> Committed revision 31.
>
> C:\test>svn rm b.txt
> D b.txt
>
> C:\test>svn ci -m "Deleted on client side."
> Deleting b.txt
>
> Committed revision 32.
>
> C:\test>svn log -r 32:31 -v file:///c:/svn/
> ------------------------------------------------------------------------
> r32 | Tom | 2006-09-09 02:40:05 +0100 (Sat, 09 Sep 2006) | 1 line
> Changed paths:
>
> Deleted on client side.
> ------------------------------------------------------------------------
> r31 | Tom | 2006-09-09 02:39:52 +0100 (Sat, 09 Sep 2006) | 1 line
> Changed paths:
> D /trunk/b.txt
>
> Deleted on server side.
> ------------------------------------------------------------------------
>
> C:\test>svn di -r 31:32 file:///c:/svn/
>
> C:\test>
This is known, and IIRC was discussed on this list fairly recently.
It's kinda weird, but I believe that's intentional behavior. We don't
error out on a delete of an already deleted item, and it is possible
to have empty commits, although nontrivial to actually make it happen
via the client.
-garrett
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org