JRAMON@ccuam1.uam.es ("J. R. Valverde") (02/14/91)
This is my personal impression on the subject.
As Dave said, portability is becoming an important subject, as well
as user friendlyness. Right now I'm writing software on a MicroVax II com-
puter since it is always available to me, but I'm ready to jump to an
ULTRIX workstation soon. The reasons: Unix computers are cheaper than Vaxes
and far more powerful than Macs (I don't even consider MS-DOS and it's 640K
barrier).
Computers are becoming more important every day in a Biologist's life.
But time is too. A mac is nice, easy and fast, but being a monotasking
mono-user machine, requires lots of lost time (e.g.: if you issue a search
you won't be able to use it for hours, if you do it on a mini and get the
results, you'll be stuck for a while with kermit). I can't envisage any
molecular biologist sat in front of its Mac for hours while playing Tetrix.
Now, you need a multitasking system that's as easy to use as the Mac.
DECWindows' OK but Vaxes are expensive and since the database grows so fast,
you'll have to expand it (more expensive yet). The same for any proprietary
computer/system. And the system!!! Lots of MB (MegaBytes) in your hard disk
(say 100 - 200).
Finally, you don't need to buy a new computer nor a workstation to
keep in pace with times. Just as now you can do with an 8080 or Z80 machine
and Kermit as well as with a Mac to run the GCG in a server, you'd like to
continue using your AT or Mac in the next years, this requires that the
windowing system employed in the software package be networkable.
Join all of this and ask for portability (I wouldn't like to port
again a program through 5 computers/systems). What you get is something
like X-windows and UNIX/VMS, or OpenLook, of OSF/Motif.
Now let's go to what really matters. Joe Scientist works in a Public
Research Center, in a small country and runs short of funds. Should he not
work or program his Great Problem Solver? He needs something cheap. Anything
cheaper than Public Domain? He can program with low costs, and buy a program
with low or no royalties, or even get it for free!
In brief, I think that X-Windows is a very good choice (as GCC, G++,
etc...) and I would stress that ALL involved people made the most to move
to these tools. We need quality software at low prices and, if we plan to
share our knowledge the ONLY acceptable solution is this.
Anyone knows of other solutions? As far as I know, we should pay
royalties if we made software with most common commercial packages in one way
or another. Or everyone should have his own set of compilers. BTW, C compiler
comes free with UNIX, but with NO other platform.
J. R. Valverde
Biomedical Research Institute
Madrid - SPAINwill@turbo.bio.net (Will Nelson) (02/14/91)
One aspect of X that Dave didn't mention is that the X protocol can be implemented using a wide variety of languages and operating systems. It has already been implemented in C and Lisp, under UNIX, VMS, DOS, AmigaDOS, and even Unicos on a Cray supercomputer. Many of us feel that the day will come when the underlying operating system will be invisible to the user. You should be able to hop from system to system and your X environment will be what you see and deal with. -- Will Nelson Internet: will@presto.ig.com Intelligenetics, Inc. Uucp: ames!ig.com!presto!will 700 East El Camino Real (415) 962-7363 Mountain View, CA 94040