[NOTICE] Branch naming scheme in external repositories

[Paul Davis <pa...@gmail.com>]

For those of you following along at home I've just gone through and
renamed all of the branches in the new external repositories so that
we have a sane order to things. When I first did those extractions I
was silly and just named the branches "import" without thinking too
hard. As it turns out there are three sources of code that we need to
be able to merge:

* CouchDB master
* BigCouch
* RCouch

For each external repository there are three possible branches:

* import-master - Code from current Apache CouchDB master branch
* 1843-feature-bigcouch  - Code being merged in from BigCouch
* 1994-merge-rcouch - Code being merged in from RCouch

Generally speaking I haven't worked too hard on making sure that
import-master branches have a full history. These branches mostly just
exist so that we have a common base to compare against while doing
merge work. This mostly affects dependencies we pulled in (oauth,
ibrowse. mochiweb) because I pulled those histories in the
1843-feature-bigcouch branches from upstream (ie, those repos are now
a true mirror instead of snapshots checked into SVN).

The two other branches should be self explanatory.

Its important to note that not every has all three (or even two of the
three) branches. In particular BigCouch and RCouch both split some
code into external applications some of which are unique to each
codebase.

That's about all I've got for now. I'm gonna do some work to try and
get the main couchdb-couch repo into the same shape as we have in
1843-feature-bigcouch and 1994-merge-rcouch so code diffs are easier
but that'll be it for the night.

Re: [NOTICE] Branch naming scheme in external repositories

[Adam Kocoloski <ko...@apache.org>]

On Feb 12, 2014, at 1:31 AM, Paul Davis <pa...@gmail.com> wrote:

> For those of you following along at home I've just gone through and
> renamed all of the branches in the new external repositories so that
> we have a sane order to things. When I first did those extractions I
> was silly and just named the branches "import" without thinking too
> hard. As it turns out there are three sources of code that we need to
> be able to merge:
> 
> * CouchDB master
> * BigCouch
> * RCouch
> 
> For each external repository there are three possible branches:
> 
> * import-master - Code from current Apache CouchDB master branch
> * 1843-feature-bigcouch  - Code being merged in from BigCouch
> * 1994-merge-rcouch - Code being merged in from RCouch
> 
> Generally speaking I haven't worked too hard on making sure that
> import-master branches have a full history. These branches mostly just
> exist so that we have a common base to compare against while doing
> merge work. This mostly affects dependencies we pulled in (oauth,
> ibrowse. mochiweb) because I pulled those histories in the
> 1843-feature-bigcouch branches from upstream (ie, those repos are now
> a true mirror instead of snapshots checked into SVN).
> 
> The two other branches should be self explanatory.
> 
> Its important to note that not every has all three (or even two of the
> three) branches. In particular BigCouch and RCouch both split some
> code into external applications some of which are unique to each
> codebase.
> 
> That's about all I've got for now. I'm gonna do some work to try and
> get the main couchdb-couch repo into the same shape as we have in
> 1843-feature-bigcouch and 1994-merge-rcouch so code diffs are easier
> but that'll be it for the night.

Thanks Paul!  Looking forward to seeing all this surgery pay off.

Adam