Don't be too high-tech!!!

Ernest Retzel 1535 49118 ernest at lenti.med.umn.edu
Sat May 8 14:52:08 EST 1993


> When you put a new service on the
> Internet you should judge its success not by how many people you reach, but by
> how many people your resource disenfranchises.

There aren't too many things that I disagree with Doug on, but I think
this may be one of them.

When you go about designing a solution to a problem, you go about it by
choosing the *best way you can think of* to do it.  The more novel
the problem, the more likely it will be that the solution will be
either non-traditonal or at least complex.  A long time ago, I was having
a discussion with a computer science collaborator, and my point was
similar to Doug's--we can't use those things [in that case, it had to
do with something that required a lot of disk space].  His comment was
"Disk space is infinite, RAM is infinite, CPU power is infinite--those are
implementation details!  How do we best solve The Problem?"  And he
was essentially right--back then, 1 gig was a lot of disk; 32 megs was a lot
of RAM; and *maybe* two processors on a million dollar machine.  Now you
can catalog-order a server with 20 processors, 1 gig of RAM and a terabyte
of disk.

The lowest common denominator philosophy is great, if that is what the
problem is.  But when the problem becomes more complex, so does the 
solution.  I think the goal of the community should be toward an upward
slope, not maintenance of the status quo.  I also think that an 
entity like NCBI is precisely in a position to lead the community up that 
slope, or at least show us where it is.

Ernie Retzel




More information about the Bio-soft mailing list