On Mon, Dec 31, 2012 at 03:18:41PM +0100, Branko Čibej wrote:
> On 31.12.2012 15:05, Stefan Sperling wrote:
> > Anything more involved than the foreign-repos merge crutch above would be
> > far from trivial. And there is no point in making Subversion a distributed
> > version control system.
>
> I think you're misinterpreting what "foreign repository branch" means.
> It does /not/ imply that Subversion would suddenly become distributed.
Well... what does this term really mean to you? You seem to be
using it to describe some idea you have about some proposed future
behaviour of Subversion.
I was answering Ziemowit's question about inter-repository copies.
This has nothing to do with whatever definition we apply to the term
"foreign repository copies" now or in the future.
In my mind, "foreign repository merge" is a term we use to describe
the existing hack of making 'svn merge' cope with merge sources from
foreign repositories by dropping mergeinfo and copyfrom information.
It is already documented as such in 'svn help merge'.
> Rather, it would do away with all the hoops you have to jump through in
> order to work with vendor branches. It would also remove 99% of the use
> cases for out-of-repo externals.
>
> Both are worthy goals IMO.
I agree that sounds good, though I'm not sure what details you have in
mind. You might want to call your proposed advanced foreign merge something
else ("inter-repository merge", perhaps?) to avoid confusion with existing
terminology. Or maybe for some future release we can redefine "foreign
repository merge" to mean whatever you think it means.
Received on 2012-12-31 15:39:15 CET