[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Gnu-arch-users] Re: Best way to handle « large » binary f iles
From: |
Miles Bader |
Subject: |
[Gnu-arch-users] Re: Best way to handle « large » binary f iles |
Date: |
08 Apr 2004 16:23:01 +0900 |
Matthieu Moy <address@hidden> writes:
> I would like to tell Arch « OK : Keep this file in the archive, but
> forget about revision patch-1, patch-2, patch-3, patch-5 and patch-6.
> Only base-0, patch-4 and patch-7 are important now ».
>
> I'm affraid I can't do that on a single file with Arch.
>
> Probably with a one-file archive, with a snapshot branch doing tags to
> the main branch from time to time, and fast archive cycling on the
> main branch, this would do it.
I'd think you'd want to use real merging, not tagging: do daily
development in a local archive (local so nobody makes any tags into it),
with period merging into the real archive. When the local daily archive
gets too big, just make sure it's been merged, and then delete it, and
start a new local archive (tagged from the real archive).
[Or actually to make the namespace nicer, don't actually create a new
local archive, just a new (arch) version of that branch (again tagged
from the real archive, not the old version), and physically delete the
old version tree from the archive.]
That's how I'd do it anyway.
-Miles
--
"1971 pickup truck; will trade for guns"