Workflow question, advice...

[David Champion]

> I think you buys are missing the original poster's point.  It's not that
> typing an extra command to pull or push is onerous, it's that half of
> your change sets are merges.  From a CVS user's perspective, it makes
> things messy.  Plus you may pull, merge, ensure it compiles, and then go
> to push and find out that you have to pull and merge again.  It seems
> like you're trying to optimize the wrong thing.

I frequently miss the point, especially where massively collaborative
version control is concerned.  However: while I value merge logs for
history's sake, I admit that for most -review- tasks they add more noise
than signal.  It seems that if that bothers a lot of people, then it
should be possible to somehow mark in metadata commits of merges that
went without a user resolution, and hide those from 'hg log' if, say,
hidesafemerges=True in [log].

Is that a feasible compromise?

-- 
 -D.    dgc at uchicago.edu    NSIT    University of Chicago