Joe Smith (jes at presto.med.upenn.edu) wrote:
: Oh dear. Here we go again. I don't meant this to be advocacy and I
: refuse to get involved in a PC vs. Unix discussion (as much as I enjoy
: them ;-). It's just a different opinion.
: I agree: there's value in both. We use both in the lab to good
: advantage.
: But, while the PC systems have caught the workstation in performance,
: and sometime ago surpassed them in price/performance, they have also
: increased greatly in complexity. Around here at least, I find that
: Don's last statement no longer rings entirely true. People spend a
: lot of time fiddling around with the Macs and PCs, trying to get
: things to work. We also spend a lot of money trying to keep up with
: the latest (and hopefully compatible) versions of the OS and all the
: add-ons.
I cannot agree here, Wintel has a ways to go for price/performance.
Most PC's are cheaper though. Also about user interfaces (previous post)
This is the fault of developers, or system administrators. I used to
work for MBCC (Molecular Biology Computing Center) now Advanced Biological
Computing Center at the Univ. of MN. The amount of work that can actually
get done was radically more (and still is much more) than could be done on a PC.
The User interfaces we had made it easy to interface, I grant only text,
but it would be EASY to make interfaces for the tools in TK/TCL or Open GL.
And with SGI's (my new employer) the price of $6,000 for a Unix workstation
that can Run W95 or WNT software, its price/performance (for real work)
is much greater. (My opinion of course, not representing my employer).
.... Lots of good stuff deleted...
: OTOH, with the advent of Linux and other cheap/free Unixes, the
: price/performance gap is completely gone. There is no longer a
: premium to be paid to run Unix. The usability gap is also shrinking
: as more commercial software becomes available.
Yes, Linux will even improve the PC over Windows (iiick). The price
difference is not that much for similar systems (of course there is little
high level support for linux, perhaps Cygnus will.) Also the ability to
grow the system as your needs grow is limited on PC's. (as you say just
below).
: So, I see a value in the desktop Unix market, in that there's more
: stability and more room to grow smoothly.
Interfaces should be easy to write especially in Open GL, and TK/Tcl,
these langauges were not always available and X is not real pretty for
coding (^8...
Perhaps list the software that you find efficient and usable, then
a number of people (perhaps CompSci students) could write user interfaces
for these tools.
Thanks! (These opinions are my own, and yes, I would have posted the
same thing even if I did not work for SGI).
Mark
------
Mark Dalton CH3-S-CH2 H H O H
SGI/Cray Research | | | \ |
Los Alamos,NM 87544 CH2-C-COO //\ ---C--CH2-C-COO C-CH2-C-COO
mwd at cray.com | | || || | // |
NH3 \\/ \ / CH NH3 O NH3
NH
My home page: http://lenti.med.umn.edu/~mwd/mwd.html
BEAM Robotics: http://www.webconn.com/~mwd/beam.html
WWW Cell Bio. course: http://lenti.med.umn.edu/~mwd/cell.html
