[PD-dev] file releases on sourceforge
Frank Barknecht
fbar at footils.org
Sun Mar 2 10:28:29 CET 2003
Hallo,
guenter geiger hat gesagt: // guenter geiger wrote:
> Ok, so what we will have to do (I know we talked about this when doing
> the debian packages, but this release should be general for GNU/Linux.
>
> - pd-externals
> (all externals that do not depend on additional libraries
> except libc and libm)
>
> compiled for:
> * Linux
> * OSX
> * Windows
This sounds okay. Should we include the libraries there, maybe
stripped into single externals? This way, we would automatically get
rid of name-conflicts like abs~.
> - flext
flext can be build to depend on libsndobj and/or stk. How to deal with
that? STK might have a license problem (waveguide patents,...), so
maybe it cant't be included into distributions like Debian. SndObj
already is in Debian (Agnula). So I see two possibilities:
a) Build flext without SndObj and STK, but maybe have an explanation
in a package README, how to compile it with the two.
b) Build two packages, flext-pure and flext-synthesis (or flext-stk
and flext-sndobj), including both.
I could live with a), but b) would make the inclusion of my syncgrain~
external easier, which depends on flext-sndobj.
> - flext-externals
> compiled for:
> * Linux gcc2
> * Linux gcc3 (Do we need both versions ??)
> * OSX
> * Windows
gcc2 seems to be hard to get right with flext. As oth RedHat and
Debian (don't know about OS-X) now default to g++-3.x I wouldn't
bother with g++-2.x.
> Then there are some externals left that depend on libraries (like the
> ogg things). Should we compile them statically ?
I think, statically feels wrong for Linux, but it could be the right
thing to do on Windows.
> Ah, yes, ask everyone if it is ok to release the code as it is now.
>
> Document installation (or add installer scripts ?)
> Should we release .rpm and .debs directly ?
That's what I would prefer. Or include the debs in Debian unstable and
give the rpms to Planet CCRMA.
ciao
--
Frank Barknecht _ ______footils.org__
More information about the Pd-dev
mailing list