Pulling with --dont-allow-conflicts when there is a conflict doesn't
fail as it should, but asks me to resolve it.  (--skip-conflicts does
work.)  I'll send a test.

Uh-oh!  Would you be able to help us out by sending a regression testing
script for this bug?  It's really easy.

See http://wiki.darcs.net/Development/RegressionTests

Thanks! 

PS. this is my first or second attempt at using a semi-canned response
(I think it was Trent who gave me the idea)

Eric Kow <bugs@darcs.net> writes:

> Eric Kow <kowey@darcs.net> added the comment:
>
> Uh-oh!  Would you be able to help us out by sending a regression testing
> script for this bug?  It's really easy.

I already did, once I got the issue number (patch212).

Thanks for the test case (rah! that really helps us move things along).
Adding Ganesh because I think he may be particularly interested.

Assuming the test case is accepted, we need investigation to work out
what the underlying problem is.

On Mon, 26 Apr 2010, Eric Kow wrote:

>
> Eric Kow <kowey@darcs.net> added the comment:
>
> Thanks for the test case (rah! that really helps us move things along).
> Adding Ganesh because I think he may be particularly interested.
>
> Assuming the test case is accepted, we need investigation to work out
> what the underlying problem is.

>From memory, --external-merge trumps --dont-allow-conflicts (the two 
don't really make sense together). I haven't checked the code to confirm 
that though.

Ganesh

Ganesh Sittampalam <ganesh@earth.li> writes:

> From memory, --external-merge trumps --dont-allow-conflicts (the two 
> don't really make sense together). I haven't checked the code to confirm 
> that though.

OK.  That's not documented.  I think it's the wrong way round,
particularly as you don't seem to be able to do something like
--no-external-merge to turn off what you have in ~/.darcs/defaults.

On Mon, 26 Apr 2010, Dave Love wrote:

> Ganesh Sittampalam <ganesh@earth.li> writes:
>
>> From memory, --external-merge trumps --dont-allow-conflicts (the two
>> don't really make sense together). I haven't checked the code to confirm
>> that though.
>
> OK.  That's not documented.  I think it's the wrong way round,
> particularly as you don't seem to be able to do something like
> --no-external-merge to turn off what you have in ~/.darcs/defaults.

I agree that the interactions are confusing and inconsistent and that you 
need a way to be able to disable options from the defaults etc. This was 
discussed during the --skip-conflicts thread late last year, but noone has 
done much about it since. (Though I thought someone did at least 
add negative versions of all existing options, but I could be 
misremembering.)

On Mon, Apr 26, 2010 at 23:42:24 +0100, Ganesh Sittampalam wrote:
> I agree that the interactions are confusing and inconsistent and that you 
> need a way to be able to disable options from the defaults etc. This was  
> discussed during the --skip-conflicts thread late last year, but noone 
> has done much about it since. (Though I thought someone did at least add 
> negative versions of all existing options, but I could be  
> misremembering.)

Florent added a lot of --no-foo flags.  It seems like we need a
--no-external-merge flag along that vein of work.

Related bugs:
- http://bugs.darcs.net/issue1457
- http://bugs.darcs.net/issue1550

Aside from implementing --no-external-merge, do we need to do anything
else in particular?

-- 
Eric Kow <http://www.nltg.brighton.ac.uk/home/Eric.Kow>
PGP Key ID: 08AC04F9

On Tue, 27 Apr 2010, Eric Kow wrote:

> On Mon, Apr 26, 2010 at 23:42:24 +0100, Ganesh Sittampalam wrote:
>> I agree that the interactions are confusing and inconsistent and that you
>> need a way to be able to disable options from the defaults etc. This was
>> discussed during the --skip-conflicts thread late last year, but noone
>> has done much about it since. (Though I thought someone did at least add
>> negative versions of all existing options, but I could be
>> misremembering.)
>
> Florent added a lot of --no-foo flags.  It seems like we need a
> --no-external-merge flag along that vein of work.
>
> Related bugs:
> - http://bugs.darcs.net/issue1457
> - http://bugs.darcs.net/issue1550
>
> Aside from implementing --no-external-merge, do we need to do anything
> else in particular?

I'm not sure if last option currently wins (with defaults coming before 
the command-line). If it doesn't, it should.

There's also been some discussion about cleaning up the conflict-allowing 
options, in this thread:

http://lists.osuosl.org/pipermail/darcs-users/2009-October/021664.html

I don't really know the right answer so am not inclined to do anything in 
a rush, but it should be kept in mind when looking at this area.

OK, I gather this is just about implementation then (plus tweaking the
test comments to reflect the real nature of the bug).  Thanks!

See patch2233