[R-sig-Debian] building from source after installing Debian packages

Dirk Eddelbuettel edd at debian.org
Wed May 11 03:22:44 CEST 2005


Ramon,

On 10 May 2005 at 13:54, Ramon Diaz-Uriarte wrote:
| I signed (rdiaz-guest) and was looking around.

Excellent -- I just added you to 'pkg-bioc' (which also covers CRAN, the
projects just happens to have been started by Matt with an initial BioC focus)

| I could not find the code for 
| building packages (it seems there are no files for the pkg-bioc project which 
| I can access from alioth); is this the Perl script you mention and Egon 
| refers to in his message today to the pkg-bioc-devel list? 

I think Egon covered that in his follow-up.

| In addition, a couple of questions: to prevent accidentally building packages 
| which might use Goto's BLAS I assume all package building should be done 
| against a non-Goto's version?. 

Dunno -- I never installed Goto's BLAS but stuck with Atlas. With Atlas, it
is actually much slicker, and easier, thanks to Camm's wizardry with the 
virtual packages. You can build with /any/ version of refblas, blas, atlas,
... and the correct dependencies are created. (There have of course been
bugs, but it is stable now).

Just try with Goto BLAS. If it creates a Depends: we cannot satisfy you will
surely hear from us :)

| Finally, I know no Perl at all (= even 
| surrounded by bioinformaticians I've resisted heroically all efforts to coax 
| me to learn Perl), but I assume that should be no problem, at least in terms 
| of checking that things work, etc.

I sort-of remembered that from talking to you in the past.  Well, most of the
work at the next stage will be testing, and adding / refining Depends: at the
Debian layer (where package names may be different from what is encoded at
CRAN). 

Thanks for joining the team! We're glad to have you.

Cheers, Dirk

-- 
An economist is an expert who will know tomorrow why the things he 
predicted yesterday didn't happen today.  --  Laurence J. Peter



More information about the R-SIG-Debian mailing list