From: Carl Worth <cworth@cworth.org>
To: Michal Sojka <sojkam1@fel.cvut.cz>,
 Sebastian Spaeth <Sebastian@SSpaeth.de>,
	Jesse Rosenthal <jrosenthal@jhu.edu>,
	Notmuch development list <notmuch@notmuchmail.org>
Subject: Re: sort order regression
In-Reply-To: <87tyr237qd.fsf@steelpick.2x.cz>
References: <87bpdbmvj0.fsf@SSpaeth.de> <87wrvz7ex3.fsf@jhu.edu>
	<87vdbj7elk.fsf@jhu.edu> <87tyr37e4p.fsf@jhu.edu>
	<87ljcfqjl3.fsf@yoom.home.cworth.org>
	<87mxwur5jl.fsf@SSpaeth.de> <87tyr237qd.fsf@steelpick.2x.cz>
Date: Fri, 23 Apr 2010 08:18:42 -0700
Message-ID: <87y6gefb8t.fsf@yoom.home.cworth.org>
MIME-Version: 1.0
Content-Type: multipart/signed; boundary="=-=-=";
	micalg=pgp-sha1; protocol="application/pgp-signature"
Precedence: list

--=-=-=
Content-Transfer-Encoding: quoted-printable

On Fri, 23 Apr 2010 10:15:22 +0200, Michal Sojka <sojkam1@fel.cvut.cz> wrot=
e:
> On Fri, 23 Apr 2010, Sebastian Spaeth wrote:
> > Ohh, that is a good point. Maybe I should write some :-). Is the test
> > suite going to be changed any day now or does it still make sense to
> > write tests for the "monolitic" test suite?

I would still like to make things more modular in the future. That is,
I'd like the tests to be categorized and grouped so that one can run a
single group of tests, (or even a single test). I'd also like to have
some dependencies between tests so that when an early test fails, later
tests that depend on that functionality will simply report UNTESTED due
to the earlier failure.

But there's certainly no need to wait for that before adding additional
tests.

> I do not have a plan to modularize the test suite in a near future.

Michal, thanks for all your effort to modularize based on the git test
suite. I know you've had to put continual effort to keep your patches up
to date as things have changed. Sorry that you got blocked by some
relicensing that hasn't quite worked out.

> I guess that currently we are not able to relicense the test library
> from git. There is still one missing ack. I've just sent a mail to Junio
> about his opinion.

Want to replay all the git test-suite commits other than any from the
person with the missing ack? That should get us pretty close to the
current state in git, would give us code we could use, and might even
make it possible for us to submit improvements directly to the upstream
git repository.

Should be a pretty simple "git rebase -i" or so and edit out the lines
From=20the author of interest. (Simple, assuming it doesn't fall over
completely due to conflicts.)

Just an idea anyway. I also understand if you don't have motivation to
pursue this further. And we can just do our own "modularization" as we
want new features.

=2DCarl

--=-=-=
Content-Type: application/pgp-signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)

iD8DBQFL0bpS6JDdNq8qSWgRAk+yAJ9uWKZo7My6jGWFSuxB/tDLYfAWrwCfa7ER
zb1GALG3oDS16TRW1rye6Uc=
=h7qH
-----END PGP SIGNATURE-----
--=-=-=--