On 11/18/2016 10:32 AM, John Jefferies wrote:
On 18/11/2016 08:51, Arne Babenhauserheide wrote:
Nicolas Pinault writes:
Testing just needs you to use it and to report any problems you
experience. Often not having deep skills is actually an advantage for
that :)
Time is another matter… maybe you’ll have a project at some point where
it can come in helpful - for example writing a patch you’ll have to
revise a few times.
I have used the evolve extension every day for several years now and
find it wonderful for the work I do. While I accept that there may be
rough edges that the devs want to clear up before it moves from
experimental, I don't see them in the work I do. If people urged it to
move from experimental as it is now, it's my belief that documentation
explaining the limitations is maybe all it needs.
It is great that evolution work great for you. But it get more
complicated for multiple other. I even met some on them today. There is
a defined set of thing we know we need to be better before can safely
release it to more user. We'll eventually get there.
Thanks for your positive and detailed feedback ☺
Also, I'm sure you know, TortoiseHg supports most of the evolve
commands, and is very good at displaying the history of obsolete and
troubled changesets. I can't be the only one who uses these features so
it must have had lots of testing by now. :-)
OK, one problem I've had is where I wanted to evolve a branch containing
a merge. I had to manually rebase stuff, but that was no more difficult
than if I wasn't using the evolve extension. There were problems in the
early days that have been fixed, such as difficulties working out why a
changeset was troubled.
FWIW. The commands I regularly use are:
commit --amend (lots)
rebase
histedit
evolve
prune
touch
Did you had a look at `hg uncommit` and `hg split`?
Cheers,
--
Pierre-Yves David
_______________________________________________
Mercurial mailing list
[email protected]
https://www.mercurial-scm.org/mailman/listinfo/mercurial
