On Mon, 7 Feb 2005, Nicolas Mailhot wrote:
Le lundi 07 février 2005 à 22:12 +0100, Dag Wieers a écrit :
> On Mon, 7 Feb 2005, Anthony Green wrote:
>
> > On Mon, 2005-02-07 at 05:58 -0500, Build System wrote:
> > > beecrypt-4.1.2-1
> > > ----------------
> > > * Sat Feb 05 2005 Jeff Johnson <jbj(a)jbj.org> 4.1.2-1
> > > - upgrade to 4.1.2
> > > - put java components in sub-package.
> > > - check that /usr/lib64 is not used on alpha (#146583).
> >
> > You called this sub-package beecrypt-java. I suggest renaming this to
> > something like beecrypt-java-jni. It only contains the JNI C side of
> > the beecrypt java library. The library of java code for beecrypt is
> > widely known as "beecrypt-java" (google for
beecrypt-java-2.0.0.zip).
> > beecrypt-java hasn't been packaged yet, but it will surely require this
> > beecrypt-java-jni.
>
> I would suggest java-something as a standard name, just like
> python-something, perl-something and other subpackage policies.
Please no oh no.
Do you imagine the mess if ~ 1000 java-something hit rawhide ? Because
we have this number of java packages in jpackage.
Well, I'm happy that perl packages have a seperate namespace and I would
love python, mono and java packages pursue this further than they do (even
without a CPAN alike infrastructure).
Does Jpackage have 1000 java-class packages ? Because I'm not arguing to
have everything java-based to fit this scheme, only packages that extend
the java 'platform'. Just like perl-modules, python-classes or for that
matter xmms-plugins...
-- dag wieers, dag(a)wieers.com,
http://dag.wieers.com/ --
[all I want is a warm bed and a kind word and unlimited power]