Re: Semi-retiring: All (ok, some) packages must go!

[John Lapeyre <lapeyre@physics.arizona.edu>]

*Raul Miller wrote:
> On Sun, Oct 10, 1999 at 05:07:22PM -0700, John Lapeyre wrote:
> >     This is OK.  Except that Raul Miller took pdl already.  r-pdl is
> > somewhat obsolete.  I currently have it patched into the pdl source.
> > The pdl people will take it upstream, but I haven't had time to correct
> > a couple of things.(eg, there is an identifier collision).  I don't know if
> > he will continue to include r-pdl as a patch.
> 
> I was not planning on supporting r-pdl, as I have not ever used R, and
> I wouldn't know where to start.  However, I did notice that there are
> a number of files under Lib/R which are not in MANIFEST.  I'll probably
> leave them there until I run into a problem with them.
> 
> Is r-pdl important to anyone?  I'm concerned about the size of the debian
> .diff.gz for pdl, and would hate to waste a lot of time doing something
> that no one cares about (or, worse, breaking R support and not finding
> out about it).
   
  Well, I use r-pdl. I find it much more convenienent to have it folded into
the pdl source. I don't know how many other people use it. The upstream pdl
people (you should probably get on their list) expressed an interest in having
it included in the standard source, but as I said, I didn't have time to clean
up a few small issues ( a few hours).  If you don't keep it in the source, I
either have to take the time to send r-pdl properly upstream, (which I don't
want to do now) or put r-pdl back into unstable.
   r-pdl is fairly simple, it is just a set of functions related to 
probability distributions.  I think it is reasonable to expect something
like this in pdl.  It is built just like the other extensions that are a 
standard part of pdl. A C library is built and then a pdl interface is
built-- documentation is added, etc.  It is perhaps easier to understand
than the standard pdl slatec interface, because of the issues with fortran.
   On the other hand, an addition of that size does not really belong in a
patch; it should be moved to a separate package, or I should do the upstream
integration. But, if you could leave it in the patch for a while, it would buy
me some time.  
     Or maybe, I can remain the pdl maintainer at least long enough to
resolve the issue (maybe through potato).  I wouldn't mind hanging on
to it for a while. What do you think ?

   John    
-- 
John Lapeyre <lapeyre@physics.arizona.edu>,  lapeyre@debian.org
Tucson,AZ     http://www.physics.arizona.edu/~lapeyre