On 26 May 2006 at 22:04, Frederik Schueler wrote:
| Hello,
|
| On Fri, May 26, 2006 at 01:36:09PM -0500, Dirk Eddelbuettel wrote:
| > Hm, built on all other platforms.
|
| It did too on amd64, when we built the old, unofficial archive (before
| gcc-4.0).
Ack. I think that was probably before I switched R to depends on gcc >= 4.*.
| > R itself is now built with gcc/gfortran
| > 4.0.* -- do you understand where all the 3.4 versions of gcc/g77 come from?
|
| g77 depends on cpp (>= 4:4.0.3-4), gcc (>= 4:4.0.3-4), g77-3.4 (>= 3.4.6-1)
All fine, but who depends on g77? For R, I make'em build-depend on
r-base-dev, and that now has
Package: r-base-dev
Architecture: all
Section: devel
Depends: r-base-core (= ${Source-Version}), refblas3-dev | atlas3-base-dev |
libblas-3.so, build-essential, gfortran (>= 4.0.2), gcc (>= 4:4.0), g++ (>=
4:4.0), libncurses5-dev, libreadline5-dev, libjpeg62-dev, libpcre3-dev,
libpng12-dev, zlib1g-dev, libbz2-dev
R itself then encode gcc et al in /etc/R/Makeconf as 'gcc' et al. I don't
have to force that to be gcc-4.0, gfortran-4.0, ... do I? (I did in the past.)
| Seems gcc-defaults has to be updated, and all r-cran packages have to be
| updated?
I don't think so. Look at my maintainer QA summary page at
http://qa.debian.org/developer.php?login=edd&comaint=yes
and some 40 or 50 R/CRAN packages build fine, include a maybe 5 or 10 with
Fortran. Not sure why this failed now ... Also note that R itself just build
fine this morning as well as last Friday. [ We're in pre-release for R 2.3.1
so I do weekly builds to help. ]
Dirk, puzzled
--
Hell, there are no rules here - we're trying to accomplish something.
-- Thomas A. Edison
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
