Future software directions
doelz at comp.bioz.unibas.ch
Fri Feb 19 02:46:07 EST 1993
Sorry, I can't resist...
In article <9302180457.AA20119 at net.bio.net>, KNECHT at UCONNVM.bitnet writes:
|> currently running MacVector on Macs and have GCG running on a VAX.
|> The question has come up as to the future directions. I see lots of
|> development going on in the PC world and from my experience with GCG,
|> I expect it to die a slow death that I will shed no tears over. I hate
|> having to work with a manual open on my lap.
As you might be aware of, the costs of biocomputing for PC's are not as cheap
as generally assumed. If you buy a PC, CD-ROM, and software, this is all
licensed to ONE SINGLE CPU, i.e., one person's desk. Depending on the size
of the department, a purchase of 20 PCs of this kind already equals a (very)
reasonable workstation. You might have missed that both the alpha VMS as well
as the UNIX boxes have come down to quite reasonable prices. If you happen
to have GCG on UNIX, there's already some (admittedly interim) solution for
Workstations out there, resembling a pull-down application finder. Without
pushing GCG, I have colleagues with all PC software available in the PD, and
even one or two commercial offers, and instead enjoying menu and clicking,
the poor folks steadily show up with questions like "How do I search
my sequence against the most recent database which isn't on CD yet?" , "How
do I get and send sequences into the PC from my eMail system?", "Why can't
I do the operation X on the software Y but must instead try to work around
by using method A in package B?". "Why didn't you tell me that I needed
to switch on option C in the pulldown D, otherwise option E might fail to
recognize the laser printer"?
|> Some have heard that development at Los Alamos in regard to the Human
|> Genome Project is targetted at Unix workstations (Sun Sparc) as opposed
|> to PC's. Any opinions or facts out there related to this information
|> and the general utility of Macs vs. Suns for all the things we will
|> want to do for protein and DNA analysis in the not to distant future?
Without knowing which detailed software you are talking about I would like
to emphasize that talking about Macs and Suns is shortsighted and leaves
no option for the future. If a company wants to make a product it usually
offers it for PCs and/or Workstations. If you have ever seen Solaris on
Suns you might be aware of the developer's pain to port a 'generic' SUN
application from Solaris 1.0 to Solaris 2.0. Therefore, code of commercial
suppliers is generally written in a way that enables to prepare another
operating system version with (modest) ease.
Unfortunately, academia does not think this way. I have captured a lot
of software in the past (actually, the UNIX software particularly is
concerned) and trashed most of it within short time. The 'life cycle' of
an academic software is diploma thesis, PhD, and if you're lucky a PostDoc.
No support, no guarantees. You get what you deserve. If you review the
grants you wrote in the past, and look at the cost for chemicals, would
you still apply for a restriction enzyme if you get it for free because
it was isolated from biological sources in a student's course? Free? Free?
And then look at the computers you purchased. Just the hardward and maybe
an OS upgrade. Application software is public domain, and if not you
even '''borrowed''' (SHAME!!!) it from your neighbor?
Without flaming myself, I know that the description is black and white
painting without getting into details, but (1) I want to get the idea
across and (2) I currently use a black and white terminal and can't find
the toolbox to make it pink :-)
| Dr. Reinhard Doelz | RFC doelz at urz.unibas.ch |
| Biocomputing | DECNET 20579::48130::doelz |
|Biozentrum der Universitaet | X25 022846211142036::doelz |
| Klingelbergstrasse 70 | FAX x41 61 261- 6760 or 267- 2078
| CH 4056 Basel | TEL x41 61 267- 2076 or 2247 |
+------------- bioftp.unibas.ch is the SWISS EMBnet node ----------------+
More information about the Bio-soft