Hi Dan,
On May 7, 2007, at 10:06 PM, Jean-François Mertens wrote:
>
> On 08 May 2007, at 04:53, Daniel Macks wrote:
>
>> On Mon, May 07, 2007 at 05:51:04PM -0500, Remi Mommsen wrote:
>>> I guess it wouldn't be too difficult to extend the variant syntax to
>>> the Shlibs field. Is there any show-stopper/stumbling block which
>>> I'm
>>> not aware of?
>>
>> Nope, just need to make sure that validator also enforces a
>> BuildDepends on a version of fink that knows about this new
>> syntax. Otherwise a package built by an older version of fink will
>> include that literal varianting syntax as part of the Shlibs field.
>
> Hoping I understand correctly: it is not only variants, but the full
> power of conditionals that should be available there.
> E.g., some pkgs name their libs in an architecture-dependent way:
> no reason to make any variant, but conditionals are needed..
>
> (cf eg the the info files for pwlib1 or openh323-1)
Thanks for implementing this feature. I gave it a try and it seems to
mostly work. There is an issue with the validation when running with
'fink -m' when it comes to validating the deb directories:
Validating .deb dir /sw/src/fink.build/root-root5-shlibs-5.14.00e-3...
Error: Shlibs field specifies /sw/lib/root/libEGPythia6.5.dylib, but
it does not exist!
Error: Shlibs field specifies /sw/lib/root/libG4root.5.dylib, but it
does not exist!
Error: Shlibs field specifies /sw/lib/root/libHbook.5.dylib, but it
does not exist!
Removing runtime build-lock...
Removing build-lock package...
These shlibs do indeed not exists and are protected by the
conditional. Here's a snippet of the shlibs field:
(%type_pkg[-cernlib]) %p/lib/root/libEGPythia6.5.dylib 5.0.0 %n
(>=5.02.00-1)
(%type_pkg[-geant4]) %p/lib/root/libG4root.5.dylib 5.0.0 %n
(>=5.14.00-1)
(%type_pkg[-qt]) %p/lib/root/libGQt.5.dylib 5.0.0 %n (>=5.15.04-2)
%p/lib/root/libGX11.5.dylib 5.0.0 %n (>=5.02.00-1)
%p/lib/root/libGX11TTF.5.dylib 5.0.0 %n (>=5.02.00-1)
(%type_pkg[-cernlib]) %p/lib/root/libHbook.5.dylib 5.0.0 %n
(>=5.02.00-1)
%p/lib/root/libHist.5.dylib 5.0.0 %n (>=5.02.00-1)
Interesting to note is that the validation does not complain about
the libraries which are only built by the '-qt' variant (and which
are missing, too). I guess, that is connected to the order of the
type field:
Type: -cernlib (boolean), -geant4 (boolean), -qt (boolean)
i.e. it looks like that only the last conditional is taken into account.
I copied the current version of the root5-devel info and patch file to:
http://www-clued0.fnal.gov/~mommsen/fink/
Remi
--
Computers are like air-conditioners, they stop working properly when
you open Windows. (Anonymous)
*********************************************************************
Remigius K. Mommsen e-mail: [EMAIL PROTECTED]
University of Manchester URL: http://cern.ch/mommsen
Fermilab, MS 357 voice: ++1 (630) 840-8321
P.O. Box 500 fax: ++1 (630) 840-2649
Batavia, Il 60510, US home: ++1 (630) 236-0932
*********************************************************************
-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
Fink-devel mailing list
Fink-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/fink-devel
