Re: Experimental slink python and jpython packages

["Lorenzo M. Catucci" <lorenzo@argon.roma2.infn.it>]

On Tue, 13 Oct 1998, Gregor Hoffleit wrote:

> 
> 1.) Play safe. Stick with the current scheme for slink. I'll upload a
> new revision 1.5.1-5 with minor fixes only (all official patches
> applied).
>
This is how I think we should go...
> 
> 2.) Play risk. Adopt a new scheme similar to my proposal, e.g.
> 
> - Remove python-net, python-misc; new package python-lib. This would
>   break a few packages' dependencies; those had to be rebuilt.
>
This could be done too; python-lib could even provide python-net and
python-misc, therefore I think it could well be done too

I'm still unsure about the /var/cache idea: we all know Cpython can import
packages even though the corresponding py[co] doesn't exist -if with some
speed drawback- and I expect for Jpython it is just the same, but the
speed drawback is there, and (worse to me) /var/cache is meant to be a
local hierarchy, while /usr/{lib,share} can really be shared between
different machines...

As for the install time, I think it would be better to 

a. make sure pycompile doesn't end up recompiling un-needed stuff
b. put in an option to call it backgrounded and redirected to /dev/null

As for the py[co] removal while uninstalling python, wouldn't something
like a 

   find <whereever> -name '*py[co]' -exec rm {} \; 

or, better 

	for file in <package py-list>
	do
		for add_char in co
		do
				rm $file$add_char
		done
	done

inside any postrm do equally well (or even better?)

Ciao,

l.