[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Gm2] adding compiler to distributions
From: |
Gaius Mulley |
Subject: |
Re: [Gm2] adding compiler to distributions |
Date: |
06 Oct 2008 11:24:38 +0100 |
User-agent: |
Gnus/5.09 (Gnus v5.9.0) Emacs/21.4 |
Hi Andreas,
many thanks! I guess, initially, it makes sense to start with 4.1.2
Leopard (and build gm2 onto the vanilla gcc-4.1.2 + Leopard patches
from Apple).
regards,
Gaius
Andreas Fischlin <address@hidden> writes:
> I would be happy to do the Mac OS X part. I know how to do it at
> least. The only thing I don't know how to "extract" the gm2 from its
> environment. Thus I would need some help there. But preparing a dmg -
> the standard method these days of distributing software - is not a
> problem if the software works in the first place within a particular
> folder.
>
> Currently Xcode 3.0 with gcc 4.0.1 would be quite a flexible solution
> to use and is the most commonly used version of Xcode, since it works
> for MacOS X 10.4.x (Tiger) as well as 10.5.x (Leopard). The latest
> Xcode, i.e. v3.1.1, however, uses gcc 4.2. and is only available under
> the latest Mac OS X 10.5.x (Leopard).
>
> Andreas
>
>
> Gaius Mulley wrote:
> > Andreas Fischlin <address@hidden> writes:
> >
> >
> >> Dear all,
> >>
> >> This reminds me of the distribution issues, I have been discussing with
> >> Gaius previously (last year). IMHO gm2 would profit greatly, if it
> >> would be released separately from gcc. Moreover, once you have a
> >> binary, it should be possible to move that binary among machines of the
> >> same platform, which is AFAIK unfortunately not the case with current
> >> gm2 release scheme. How do others think about this?
> >>
> >> Please consider the following points: Most Unix systems of one or
> >> another form - BTW I'm coming from OS X (Darwin variant) - have
> >> typically already some gcc compiler installed. In the case of OS X the
> >> gcc preinstalled by default is often not the one gm2 uses. Why not
> >> following a release scheme where you get gm2 alone and that release is
> >> made for a particular gcc or a range of gcc. In case someone wishes to
> >> follow the same scheme gm2 uses now, this could still be done by any
> >> user by simply downloading an additional gcc component. And all those
> >> that already have a gcc installed, can simply add gm2 to the set of
> >> compilers without having to struggle with the coexistence of several
> >> gcc on the same machine. Anyone with me on this?
> >>
> >
> > Hi Andreas,
> >
> > I believe it can be released separately from gcc (I've just checked
> > the last deb package built and it contains a gm2 which should coexist
> > with a debian packaged gcc-4.1.2). Using a gm2 compiler built on
> > another platform might be possible - but if it is to be installed in a
> > different directory it might be a less than satisfactory experience
> > for the user. In theory it could work (I think) but the user would
> > have to always specify -B/path/to/cc1gm2. This could easily be tested
> > by building gm2 for debian and using 'alien' to convert it into an rpm
> > and installing it under Fedora. I did try this last year - it almost
> > worked.. there was a problem with crt0.o IIRC and I suspect this
> > could be fixed with a little time.
> >
> > Certainly I like the idea of building gm2 when grafted onto particular
> > gcc sources. This is probably the way to go - and hopefully would
> > allow cygwin MacOS ports to be much more stable. I guess the next
> > question, is "in what time frame?" .. I was aiming to finish the ISO
> > libraries and type COMPLEX before the next release (aiming at New
> > Year) - but is this too far in the future?
> >
> > Are there people on the list with Solaris, Gentoo, Fedora, Cygwin and
> > MacOS (any others) package building experience? I can build debian
> > packages from some scripts I've obtained - if anyone is interested I
> > can put them up on the web site with basic instructions? The debian
> > build process is reasonably automated, you just need a fast machine
> > or a lot of patience :-), as it uses the pdebuild chroot mechanism.
> >
> > regards,
> > Gaius
> >
>
> --
>
> ________________________________________________________________________
> Dr. Andreas Fischlin, Ph.D., Group Director
>
> Terrestrial Systems Ecology
> Institute of Integrative Biology: Ecology, Evolution, Infectious Disease
> Department of Environmental Sciences, ETH Zurich
>
> Address:
> ETH Zurich, CHN E21.1
> 8092 Zurich, Switzerland
>
> Phone: +41 44 633-6090 / Fax: +41 44 633-1136
> http://www.sysecol.ethz.ch/Staff/af/
> http://www.sysecol.ethz.ch/
>
> _/_/_/ _/_/_/ _/ _/
> _/ _/ _/ _/ Eidgenoessische Technische Hochschule Zuerich
> _/_/_/ _/ _/_/_/ Swiss Federal Institute of Technology Zurich
> _/ _/ _/ _/ Ecole polytechnique federale de Zurich
> _/_/_/ _/ _/ _/ Politecnico federale de Zurigo
>
> Make it as simple as possible, but distrust it!
> ________________________________________________________________________
>
> * *
Re: [Gm2] adding compiler to distributions, Gaius Mulley, 2008/10/06