Concatenating changesets

[John D. Mitchell]

On Dec 15, 2007 8:07 AM, Matt Mackall <mpm at selenic.com> wrote:
> On Fri, Dec 14, 2007 at 12:36:49PM -0800, John D. Mitchell wrote:
[...]
> > Why do you think it's a bad idea?  Especially if you like the use of
> > mq (in general, if not for this specific use case)?
>
> A changeset should be a single well-defined change. Which means it
> generally should be small. Combining changesets is generally (but not
> always) a step in the wrong direction.

Indeed, as with so many things, it's often a question of discipline.
Lots of tools/languages/etc. have features that can be abused if used
to extremes.  Comparatively speaking, things like mq's strip command
are much easier to abuse.

In this case, IMHO, the reason to use this feature is primarily a
question of granularity (and the attendant clarity).  TDD'ers like to
have lots of micro-commits while they are actually working on a
particular feature but the "right" granularity for communicating that
new/fixed feature to the rest of the team/world is usually in a very
small number of rollup commits.

Thanks,
John