[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Patching naming scheme



Le jeudi 30 mars 2006 à 22:55 +0200, Denis Barbier a écrit :
> On Wed, Mar 29, 2006 at 06:35:16PM -0500, Jeff Bailey wrote:

> > ## {local|branch|cvs|submitted}-{ARCH|all}-{"Description}"
> > ## Branch means commited in the upstream branch.
> > ## CVS means committed in the upstream trunk.
> > ## Submitted means we're hopeful that it will go in.
> > ## Local means that it's not going upstream.
> > ## The goal is to get all patches to "Branch".  Patches should be split
> > by
> > ## arch so that it's easy to send just the relevant bits to various
> > ## arch maintainers if needed.
> 
> When I first read your message weeks ago, I did not like it much, mostly
> because of the patch status being part of its filename.  I changed my
> mind, and believe now that your suggestion makes a lot of sense.

Cool.  =)

> > and here's a sampling of how the patches look once renamed:
> > ## Branch:
> > ## CVS:
> > cvs-all-tlscrash
> > cvs-all-strxfrm
> 
> The 2nd field could be {ARCH|all|any} instead of {ARCH|all}.  The
> rationale is that porters should have a look at {ARCH|any} whereas
> I am mostly interested in 'all' patches.  The difference between
> 'all' and 'any' could be that the latter touches several architecture
> dependent files, whereas the former only modifies architecture
> independent files.

Right.  We can see I had a couple in the list where I had put "mash".
any is a far better choice. =)

> What about putting patches in subdirectories?  For instance
>  ## {ARCH|all|any}/{local|branch|cvs|submitted}-"Description"

Separating by subdirectories probably makes sense for the number of
patches we carry.  I remember this was part of why you suggested quilt
before.

Breaking it down this way might make it clearer when giving write access
to porters, too.

Although when you're in the situation that RMS is telling you that
you're being too ideological about freedom, maybe, just maybe, it's
true.
- Matthew Wilcox

Attachment: signature.asc
Description: Ceci est une partie de message =?ISO-8859-1?Q?num=E9riquement?= =?ISO-8859-1?Q?_sign=E9e?=


Reply to: