Re: r16398 for backport?

David James wrote:
> On 10/6/05, Peter N. Lundblad <peter@famlundblad.se> wrote:
>> r16398 is proposed for backport, fixing a python bindings segfault. I'm
>> reviewing the core changes in this commit (yes, it contains both core and
>> binging changes, which I don't like, but let's leave that aside).

> Peter, I'm planning to create a merge branch for the core changes in
> r16398, because it will make the core changes easier to review.

Why? This is unnecessary complication.

People can just look at r16398, and ignore paths starting with
subversion/bindings/

>> My concern is that it remove an internal, but exported function:
>> svn_txdelta__copy_windows. It was used, however, by libsvn_fs_base. I'm
>> not sure about our policy for mixing core svn libs. Do we support linking
>> libsvn_fs_base from before 1.3 against libsvn_delta 1.3 or newer? In that
>> case, we need to keep the internal name for compatibility.

> No, we don't support mixing versions of core svn libs. See the
> following thread for an explanation:
> http://svn.haxx.se/dev/archive-2004-12/0156.shtml

Explanation later, because real life is pulling be away from the keyboard
right now, but I disagree.

Max.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Oct 6 14:43:35 2005