What's next for pbranch?

[Peter Arrenbrecht]

On Tue, Nov 24, 2009 at 10:30 PM, Peter Arrenbrecht
<peter.arrenbrecht at gmail.com> wrote:
> On Tue, Nov 24, 2009 at 4:17 PM, Martin Geisler <mg at lazybytes.net> wrote:
[snip]
>> However, I've always thought of pbranch as the most well-polished
>> third-party extension out there (wow, it has documentation!) and I would
>> like to include it in Mercurial proper.
>
> Martin, thanks. However, last time this came up seriously, Matt wasn't
> happy about having two different ways for managing patches (mq and
> pbranch) in the endorsed extensions. Don't know what his stance is
> now. I still think they differ sufficiently in their approaches and
> aims.

Another thing is, the pbranch documentation is heavily doctest
oriented (doctest for shell scripts, really). Don't know how that fits
in with the rest of Hg's documentation. Would a move in this direction
be appreciated? Or would this rather go into the Hg book? Or remain
separate?

I myself would actually rather like similar support for Hg's tests,
too. Meaning that command output is interleaved with commands in the
test scripts so that correlating commands and output is trivial. The
tests then read like a narrative with expected output inline. This
could simply be an option for new tests.

This is my current script for shell-doctests:

  http://bitbucket.org/parren/hg-pbranch/src/tip/tut/docbash.py

It could easily be adapted to handle rst instead of textile, or
something more primitive yet for pure tests.

-parren