[PATCH 0 of 6] Refactor localrepo.commit() to read from changectx

[Christian Ebert]

* Patrick Mezard on Sunday, June 15, 2008 at 19:21:36 +0200
> Here is a followup of "[PATCH 0 of 5] Support committing without a working
> directory" posted the 2008/06/02.
> 
> This series focuses on rewriting localrepo.commit() to take a changectx like
> object as single argument, and remove all dirstate management code. The second
> point is not complete yet because some parts are deeply embedded in commit
> code, but the result is good enough so memory contexts can be built and passed
> to _commitctx().
> 
> I would like to have some feedback about this approach before starting with the
> memctx and refactoring convert.
> 
> BTW, these changes have the good side effect of failing the commit upon bad
> username or date *before* asking for the commit message (or starting the
> transaction).

With current crew tip I'm getting into serious trouble ;) :

$ hg -q --version
Mercurial Distributed SCM (version 8251ffb35725)
$ hg ci -m test
$ hg tip --template='{date|isodate}\n'
2008-06-21 10:34 +0200
$ hg rollback
rolling back last transaction
$ hg -q --version
Mercurial Distributed SCM (version 2011bb8ada9a)
$ hg ci -m test
$ hg tip --template='{date|isodate}\n'
1970-01-01 00:00 +0000


c
-- 
  Was heißt hier Dogma, ich bin Underdogma!
[ What the hell do you mean dogma, I am underdogma. ]

_F R E E_  _V I D E O S_  -->>  http://www.blacktrash.org/underdogma/