[U-Boot] Please pull u-boot-atmel/master

[Wolfgang Denk]

Dear Albert ARIBAUD,

In message <4DFFB961.7070406 at aribaud.net> you wrote:
>
> > No. I understand MV next branch is not intended to go into this
> > release. When -rc3 is out, we may create a next branch, where this
> > can be pulled into.
> 
> Wolfgang: I am not as convinced as you are about the MV next branch pull 
> request not being intended for this release: the custodian wiki 

Prafulla, can you please comment what your actual intentions were,
i. e. wether Albert or me are interpreting your pull request
correctly?

> addresses pull requests for master branches only, and considers next 
> branches only as temp storage for commits accepted outside a merge 
> window. ...

Right, but that means (or is supposed to mean - sorry if this is not
documented clear enough) that all custodians can do this as well,
i. e. they can pull patches into their own "next" branches and send
pull requests for the next branch.

>     ... Next branches are supposed to be used only to rebase the master 
> the same repo; there is no process, nor reason, reason to ask for a pull 
> of a next branch. ...

I disagree here.

It makes perfecxt sense to keep the stack of open patches small, and a
next branch is a good way to handle stuff that is waiting for the next
merge window to open.

The only issue with this is that I am not very strict with this
process; theoretically we should (1) create -rc1 as soon as the merge
window closes and all patches have checked in, and (2) create a "next"
branch as soon as we have -rc1.

My problems with this is that usually there is a lot of unprocesses
stuff queued when the MW closes, so I wait with the -rc1 until at
least most of this has gone in;  additionally I wait with "next" until
I feel the tree is reasonable stable - in the current release for
example I'm waiting for all these ARM build fixes, i. e. especially
the AT91 stuff.

>               ... So if Prafulla asked for a pull, it may be that he 
> actually wanted a pull into ARM master, somehow mentally squezing the 
> 'rebase MV master onto MV next then ask for a pull of master' process.

I don't think so.


Best regards,

Wolfgang Denk

-- 
DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
There are two ways of constructing a software design. One way  is  to
make  it  so  simple that there are obviously no deficiencies and the
other is to make it so complicated that there are  no  obvious  defi-
ciencies.                             - Charles Anthony Richard Hoare