I agree wholeheartedly with Bill Pearson.
The fact that what the software manufacturers have concentrated on is
providing an interface to public domain algorithms is shameful, particularly
considering what the state of those interfaces is. Those interfaces have
encouraged researchers to think of their workstation as a cheap mini, and
continue to relate to them in an ASCII environment, while the rest of the
world has moved on to client-server environments, distributed computing, and
X-windows. In many cases, it is not even the most current versions of those
algorithms that they are supplying.
I personally think that is not the competition that Entrez and the interface
tools [Vibrant] present that is the threat. It is the fact that, with Entrez,
NCBI just raised the standards to where it should have been years ago, and
there are a lot of people waiting to applaud the effort. Had it been any
of the companies themselves that did it, they would have gotten the same
applause.
I would also say that interface research is *certainly* within the
purview of database research; I would point to the visual programming
paradigm [sometimes referred to as the data flow paradigm] I referred to in a
discussion not long ago.
I would add to the discussion that the being able to supply the primary
databases at no additional licensing fee is enough of a subsidy for the
software companies. Some, but not all, do provide utilities if you
maintain your own. On the other hand, I don't blame them for taking
advantage of it.
Ernie Retzel
University of Minnesota
ernest at lenti.med.umn.edu
