Re: Newbie GAR make build issue.

Just now Guy Hulbert made 15 LEDs in my apartment flash with this:
> Nate Riffe wrote:
> 
> >Try building it on Debian 3.0.  If you do not have access to a machine
> >running Debian 3.0, we have a Debian 3.0 chroot and instructions:
> >
> 
> Now there's an idea ... how much effort would it take to add 'make
> chroot' to GAR ?  Obviously, this would not apply to gargnome but I
> would like to be able to 'make chroot' for the singularity as well
> as bbc.

See meta/build-chroot.  It doesn't work yet, I kind of ran out of
steam on that one.  If you want it to work, submit a patch.

> This is what I originally meant by "orthogonal".  You have GAR --
> essentially gar.mk and gar.lib.mk and you have BBC -- which I think
> of as what is in meta/bbc/ and you have all the supported packages
> in the various other subdirectories.

I'm way ahead of you.  Do try to keep up.

> Although BBC nominally depends on GAR,

Sort of how your conception "nominally" depended on sex, I guess.

> one could have meta/woody (i.e.  debian3.0) which is completely
> independent of BBC. I don't think this is true for the singularity.
> If I understand things correctly the singularity is a subset of BBC.

GAR is already orthogonal to the BBC.  There are a variety of projects
that use it including LNX-BBC, GARNOME, and Gusto.  With some of the
more recent work, it's now possible to build a cross-compiling
toolchain and libc for GNU/Linux on any architecture, and clearly
that's useful outside of the LNX-BBC project.  There are many such
things that GAR *could* be used for, and if you have some particular
thing in mind, I suggest you get cracking on it rather than continuing
to espouse the fantasy that it *could* be done.  That said, GAR simply
isn't an appropriate solution to the problem of building Debian 3.0
systems.

> In order to support multiple systems you would need a way to include
> different versions of the same packages.

This is an incredibly simplistic view of the problem of ABI
dependency.

> This was one of the other "orthogonality" issues I mentioned.  I
> have not yet checked out the CVS tree so I don't know whether this
> is part of the GAR system or not.

You remind me of someone whose name I will not mention who has a very
annoying tendency to imagine problems that do not exist and then
imagine fantastical solutions to these made-up problems.  He does
absolutely no research into these problems whatsoever.  I suggest you
check out a copy of from cvs *now* rather than waiting until Christmas
and try to answer your own basic questions.  It would save us all a
lot of time.

-Nate

-- 
--< ((\))< >----< [email protected] >----< http://www.movealong.org/ >--
pub  1024D/05A058E0 2002-03-07 Nate Riffe (06-Mar-2002) <[email protected]>
     Key fingerprint = 0DAC F5CB D182 3165 D757  C466 CD42 12A8 05A0 58E0