[petsc-dev] strangness in Chebyshev estimate of eigenvalues

[Jed Brown]

Mark Adams <mfadams at lbl.gov> writes:

>>
>>
>> Firstly - we should not be cherry-picking commits to next. Its an
>> 'integration branch' i.e any required fixes go into the feature branch
>> that might have triggered the issue.
>>
>
> So I should just (re)merge my branch with my local next, as I fix it, pull
> from next, push to next.

I'm having trouble comprehending the confusion of ideas that would
provoke such a question.  You never pull from 'next'.  This is nothing
new; it's been written in bold in the developer instructions since day 1
and we've talked about it before.

Do the work in your branch.  Get it done.  If you need something
specific that someone else developed, you can merge their topic branch
(e.g., Matt's branch with some fix that you need) and provide an
accurate commit message (put your rationale below the auto-generated
merge subject line in the commit message).  Only when you're completely
done with your branch should it be merged to 'next'.  (Because of the
mess and thus rewound 'next', you'll need to reset your local 'next'
before merging.)

Remember that 'next' is for testing and eager users; anything you put in
'next' that isn't in a topic branch will be thrown away.  So don't ever
put something there except by merging topic branches.  Committing
directly on 'next' is like cooking food inside a dumpster.

If you're ever unsure, just make a PR.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 818 bytes
Desc: not available
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20160525/0d54d4ef/attachment.sig>