[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

RE : RE : RE : RE : Bug#679990: ITP: clipper -- object oriented development kit for crystallographic computing



> Yeah I know shame on me :)! I was just working on other problems first...
> Seriously I don't think the rules file is in a shape where it can be used yet. A
> lot of cctbx packages were added and others got thrown out since last year. Also
> there is a bug in fftw3tbx were also header files are missing. Problem is that
> they are providing their own sources for the libs and we need to patch that all
> out and repack.

yes a big task...

>  The whole upstream package is a construction site :D.
> As long this little issue aren't fixed I can't even say for sure how much libs
> are being build finally since scons just skips building without an error.
> Also we need to remove a few build-deps. For example libcbf and libann. They are
> extending the libcbf with some cpp for detectors and linking it statically to
> libcbf.a and python-ext. For libann they are doing self-includes so I guess it's probably best
> to link that statically too!

can you add these informations into the wiki page. so I will avoid to ask again and again.
I know myself ;). then it would be possible to point upstream to the wiki page.

so they are extending the libcbf library and did not try to integrate their changes.

> I decided to work on the shlib versioning first or do you have different
> opinion in what should be done first?

- ok, let's first produce shlib the right way.
- update the rules files to build it the upstream way.
- then deal with all library produced during this official build until we fix all identified issues.

all this to avoid building a build system on our own.

See you

Fred


Reply to: