Without rehashing all of the preceding arguments, it seems to me that the
root question here is whether "publicly funded molecular biology research"
(in the U.S. at least) can be organized as a single large entity for the
purpose of purchasing services and products, or whether each autonomous
unit (individual/lab/department/school) must act independently.
I think that you can make a pretty fair case that since a big chunk of the
money comes from the same source (Uncle Sam) that the single entity model
should be allowed. That being said, I think that it is essential to
contract out for as many of these services/products as possible. This holds
as much for DNA sequencers as it does for software. Does anybody doubt
that we could get a better deal on sequencers if we required companies to
bid on the contract for say, 1000 sequencers at a shot? Does anybody doubt
that this would be cheaper, and the quality better, than if we tried to set
up our own shop to produce these things?
This gets back to the NCBI matter - I have a minor problem with contracting
"database access software" to a single entity (ie, all the eggs in one
basket). However, I have big problems with having a government agency do
the work - when was the last time the government produced anything
efficiently? No discourtesy meant to the folks at NCBI, but since you
don't have competitition you _will not_ be efficient. Don't take me wrong,
I think that the BLAST server is the greatest thing since buttered toast
and is saving us wads of money, but I bet we could save even more if the
purveyor of that service had to bid for the right.
What I would really like to see is an environment that supports a pool of
Molecular Biology literate programmers who can be called upon to work on
one software project or another. I would certainly like to see the
software produced by these people purchased in bulk, but only after
demonstrations of competing packages by several groups. (I'm pretty sure
that public money spent on funding prototype development would be recouped
in "better software" at the end.) I'd also like these folks to be available
to clean up and distribute academic software - you know, with real manuals,
installation instructions, and support. Maybe we could even retarget
software development to support the majority of possible users! ( To rehash
an old thread :-) ).
Ok, that's a rough outline of the proposition - anybody want to delve into
this can of worms?
David Mathog
mathog at seqvax.caltech.edu
manager, sequence analysis facility, biology division, Caltech
