MQ and push/pull

Paul Moore p.f.moore at gmail.com
Sat Jan 12 10:46:31 UTC 2008


I am trying to do some work on a repository, using MQ. I have a clone
of the mainline, and I have done the normal hg qinit, hg qnew
my-changes, edit, hg qrefresh sequence.

However, I now want to do some work on this code at home. I can copy
the repo to a USB drive, and take it home to continue work, but if I
then forget to take it back in to work, I'm in trouble.

If I wasn't using MQ, I would just hg clone the repo to the USB drive,
hg clone at home, and merge the 2 independent copies back, when I
finally remember to bring the home copy in. However, with MQ I can't
do this - if I clone a repo with patches applied, the changes come
over, but the patch series doesn't. And I get in a mess if I change,
pull back and merge:

>hg  qrefresh
abort: working directory revision is not qtip

Could anyone help? What is the recommended workflow here?

Just to summarise what I want to do:

1. Work on a MQ patch in repo1
2. Take a copy of repo1, and work on the patch elsewhere
3. Work on the same patch in repo1 (I may not be able to merge back at
this point)
4. Finally get simultaneous access to the 2 repos at once, again
5. Get the 2 copies back in sync, with the patch containing all changes

I'd rather that the sequence of commands in (5) were reasonably simple
and intuitive - I'm likely to do this a lot, and a complex process
would hamper things a lot.

Thanks,
Paul.



More information about the Mercurial mailing list