Unix vs Linux - the movie.

Michael Schmitz schmitz at biophys.uni-duesseldorf.de
Tue Jul 25 14:53:03 EST 2000


"John S. J. Anderson" wrote:
> Michael> I'm extremely pleased with EMBOSS using GPL though - GCG
> Michael> going closed source was a sad day for me, and there's been
> Michael> other unfortunate licensing decisions since. There will be
> Michael> Debian packages of EMBOSS and RNAfold, there won't be
> Michael> packages for other stuff.
> 
> I decided a while ago not to use _any_ closed source software for
> science if it was at all possible. If you think about it, true peer
> review isn't possible unless you can see the source. I'm just waiting
> for the inevitable report of the bug in microarray quantitation
> software (to pull an example out of the air) that will result in paper
> retractions.

Paper retractions ?? Who's retracting papers these days? But I agree on
the peer review part, even if it usually doesn't happen before someone
runs into really funny results. But if that happens, it's good to have
the source. If you don't have someone skilled in programming in your own
group, you can always find someone in your computing resource center, in
the CS department, etc. 

Try proving to some software vendor that there's a particular bug in
their software, without sending them the diff to prove it. The M$
attitude is far too common already. 
 
> I realize having the source doesn't _cure_ this problem, but it makes
> it a lot easier to detect, and might inspire the companies doing the
> coding to do a better job of it.

To make it easy to detect, they need to have done a pretty decent job
already, or it won't be much fun. But I doubt many companies are going
to open their source, partly for IP reasons, partly for shame :-). The
benefit of getting a few bugs spotted by others doesn't seem to count.  
I'm more concerned about people at universities and other publically
funded institutions going more and more closed source. I'm not sure why
that happens, support for some in house software given away to others
has always been an issue, and the number of people capable of solving
install problems and the like should rather rise with the abundance of
Linux. That doesn't mean I'd risk to dump all of my code on the general
public, but I'd not try to be perfect in terms of packaging, autoconfig
etc. before releasing something. I'm sure there are people out there
that are better at writing configure scripts or using autoconf than I
am. 
 
> *nodnod* The nice thing is that some (most?) of the experience
> transfers across the different Unices.

The Unix way of solving little problems with simple tools sure
transfers, and even though things aren't always in the same place on
Solaris, OSF, Irix and Linux it's manageable once you know what to look
for. 

I refrain from a witty Obi-Wan quote as the thread is already getting
out of hand :-)

	Michael







More information about the Bio-soft mailing list