Re: [gentoo-dev] cvs+overlay script?

[George Shapovalov]

Well, I also use separate trees (and I know I am not the only one, as 
occasionally this kind of "bookkeeping" was coming up on irc, but apparently 
we are in minority for the most part :)), but when I process ebuilds or fix 
bugs I prefer to work in the "active", synced tree. When I am done with 
tinkering I simply copy stuff over into cvs tree and only perform 
updates/commits there.  
In case I terribly screw something up I don't even have to do any cleanup, as 
I have a habit to only touch my local cvs tree when I've finished processing 
and testing stuff, just emerge sync and carry over..
Provides some isolation and a safer feeling if you ask me :). 
Oh, and it seems to require less additional actions than what you described ;) 
(usually just a single cp command, well, in addition to all the repoman 
scan... stuff that we have to do anyway).

George

On Friday 20 February 2004 14:31, Grant Goodyear wrote:
> On my laptop I have both a local portage tree populated by rsync (which
> I update fairly frequently), and a cvs tree that I only update when I
> need to work on an ebuild that I maintain.  What I've been doing when I
> need to hack on an ebuild (such as dev-python/quixote, for example), is
> I temporarily connect my portage overlay to the appropriate part of my
> cvs tree via "mkdir -p /usr/local/portage/dev-python && ln -s
> /path/to/gentoo-x86/dev-python/quixote
> /usr/local/portage/dev-python/quixote && cd
> /usr/local/portage/dev-python/quixote && cvs up -dP".  After I get the
> new ebuild fixed and committed, I zap the symlink in my overlay.
>
> So, am I the only person working in this fashion?  If not, has anybody
> already written an elegant script to handle all of the directory
> creation and linking?  If so, should we get that script added to
> gentoolkit-dev?
>
> My not-so-bright thoughts for the day...,
> g2boojum


--
gentoo-dev@g.o mailing list