David Kastrup <[email protected]> writes:

> Colin Hall <[email protected]> writes:
>
>> On Thu, Jun 07, 2012 at 12:49:31PM +0200, David Kastrup wrote:
>
> [...]
>>> > What is your point?
>>> 
>>> That automatically enforced indentation is not necessarily the best
>>> choice for colloboration.
>>
>> Thanks, that's clear. Our experiences differ.
>>
>>> You don't have the choice to make indentation compromises while a
>>> file is actively being worked on.  Reverts and merges often fail.
>>
>> Yes, that's a feature. One only gets to check in code that meets the
>> agreed standard. Compromise the indentation on your own branch.
>>
>> I attach a naive example. Comments welcome.
>
> You are using Emacs diff.  That's not relevant.  The tool to consult
> about merge conflicts is git, and using git in a colloborative setting
> was what the thread was supposed to be about.

Just as a minor anecdote: my recent issue 2584 work starts like the
following in git:

commit 76eeafd00669ac2b3eac1fa325a9a50a46bf8faf
Author: David Kastrup <[email protected]>
Date:   Wed Jun 6 13:12:38 2012 +0200

    Revert "Allow multiple identical slurs (issue 1967)"
    
    This reverts commit b986b38f14195f31e26b0a929c8ac23a8ecfc204.
    
    Conflicts:
    
        lily/slur-engraver.cc

The "conflicts" line is a run of astyle that needed manual resolution.
It was just a few lines in this case, but since you don't really have a
three-way diff available for conflict resolution, it took half an hour
to make sure that the state was corresponding to the original.

When you have a case where not just a few lines but whole passages got
reindented, doing that kind of work is decidedly unfunny.

-- 
David Kastrup


_______________________________________________
lilypond-user mailing list
[email protected]
https://lists.gnu.org/mailman/listinfo/lilypond-user

Reply via email to