This is the mail archive of the
libstdc++@gcc.gnu.org
mailing list for the libstdc++ project.
Re: [patch] stephen's staging headers patch
- To: libstdc++ at gcc dot gnu dot org
- Subject: Re: [patch] stephen's staging headers patch
- From: Loren James Rittle <rittle at latour dot rsch dot comm dot mot dot com>
- Date: Thu, 5 Jul 2001 18:29:21 -0500 (CDT)
- Cc: bkoz at redhat dot com
- Organization: Networks and Infrastructure Lab (IL02/2240), Motorola Labs
- References: <20010702193318.A5839@disaster.jaj.com>
In article <Pine.SOL.3.91.1010702194935.5563C-100000@taarna.cygnus.com> you write:
>> I would also prefer soft links. Among other things, I can look at a file
>> in the build dir and /know/ where it came from instantly.
> Is there anyway I could get you or Loren to finish off this patch?
I can commit to fixing up the areas that I knew about problems in (and
any obvious ones that are seen in my environment) but not the whole
thing since there are many parts I don't fully understand yet.
For a patch of this size, we should consider using Mark's approach
(perhaps with a twist where we don't branch tag everything in the
repository):
Create a branch for only those few configuration files that are
changing with this massive patch. Request that no one change them on
mainline for a day or two. Phil (he volunteered to do some work on it
as well), Stephen (?it is his patch, right?), you and I can work on it
during that time without breaking other people using mainline. Then,
we roll it to mainline.
I know how to do all that work without trashing the CVS tree if you
want to attempt it that way...
Otherwise, I suppose I can apply your patch to my working tree, hack
on it and send it back. However, that isn't really using CVS the way
it works best.
Regards,
Loren