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

Re: rpms/git/devel .cvsignore, 1.63, 1.64 git.spec, 1.70, 1.71 sources, 1.63, 1.64



Josh Boyer wrote:
> On Fri, Aug 29, 2008 at 07:41:08AM -0400, James Bowes wrote:
>>I did this because someone else complained about not using
>>gitexecdir first.

Someone will complain no matter when the change happens, I'm sure.
I'm convinced that a warning could be hand delivered by a beautiful
naked person of whatever gender the user prefers and many would still
scream when the change finally landed. :)

>>But it makes sense to me, since I have not yet done any due
>>diligence to see what might break when we do move the git-*
>>commands. So no worries, we will move the commands, but there's no
>>real rush, is there?

IMHO, making the change with the update to 1.6.0+ is the right time to
make it.  Otherwise it seems confusing if the commands are moved in
upstream git-1.6.0 and not until some later git-1.6.0.x update in
Fedora.

Basically, it seems to me that git-1.6.0 is the flag day for
moving the git-* command out of PATH and keeping that in sync between
upstream and Fedora seems less likely to cause confusion for git
users.

(Imagine folks that use git on a variety of distros finding that a
hand built git-1.6.0 doesn't have the git-* commands, Fedora does,
Debian doesn't, Mandriva does, etc.  Maddening -- and might happen
anyway if other distros decide to make gitexecdir == /usr/bin, but at
least Fedora could be consistent with the upstream defaults. :)

> Though an F-9 update will not.  You just don't yank out commands
> like that from a stable release.

Yeah, I fully agree.  And that said, is there a compelling reason to
push git-1.6.0 as an F9 update┬╣ (aside from "we always have to have the
very latest release" mantra that many users chant)?  F9 should be fine
with git-1.5.6.x, shouldn't it?

Users who simply *must* have git-1.6.0+ could rebuild the F10 package
and assume responsibility for handling the incompatible changes it
includes.

If git-1.6.0+ only gets pushed to rawhide and a release note is added,
then folks who update to F10 will have an opportunity to catch the
change and fix any scripts they still have which rely on the git-*
form.

┬╣ One minor reason I can think of is the reverting of git status being
  paged, but that's mostly just an annoyance.

-- 
Todd        OpenPGP -> KeyID: 0xBEAF0CE3 | URL: www.pobox.com/~tmz/pgp
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Some of the narrowest minds are found in the fattest heads.
    -- Anonymous

Attachment: pgpF6PT3FIlaT.pgp
Description: PGP signature


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