On January 21, I posted the Following message:
"Hi everyone. I am trying to decide between the Unix and Windows NT
workstations for a new AVANCE system to be installed in a few months.
The spectrometer will be predominantly used for solids experiments. I have
read the responses to a similar question posted last June by Jim
Bloxsidge, but would like to hear from anyone who has new experiences to
report. Thanks!"
I received 11 responses, half of these from people trying to make the same
decision but who have no new information. The general feeling is similar to
what Jim Bloxsidge reported: while NT is probably the way of the future,
there are still bugs to be worked out.
One person reported:
"We went to Karlsruhe for demos, and an NT workstation was used for
the solids demo. It went very smoothly - the computer seemed to
be as fast as the SGI and did not crash. We had a meeting with Mike
Engelhardt and he told us that they were having far fewer complaints
from the NT customers but he attributed this to the fact that windows (and
therefore NT) users were used to their computers crashing and therefore
took it in stride!"
A solids applications engineer from Bruker Billerica who is testing WINNMR
on both platforms reported:
"Based on the fact that your lab is predominantly pc, I would most
heartily recommend the NT. Yes, there are some little jittery bugs on
the NT, but we are finding and identifying and fixing them via our Patches.
(And allow me
to emphasize: these are insignificant bugs that have more to do with
the graphics emulator than anything else.) As a fluent Unix user, I find
a certain comfort level in using the SGI. However, WINNMR, the Avance
software package, runs the same on both systems. The NT is far more
versatile in terms of copying and pasting spectra and other
results directly to pc programs. The only negative aspect of the NT
system applies to Unix-literate users: the file management system
is different. It requires a little extra effort to learn to navigate the
file directories, but as we have mapped these out, that shouldn't be a
problem."
Several people who currently use Unix-based systems recommend buying the
Unix version:
"I don't think that NT can beat the stability of the O2's, but if you
just use it to drive a spectrometer it should
be ok. Windows 2000 sounds like it will be better. Probably mostly
boils down to which one you want to use.
The unix systems and pc's can coexist on the network. Samba on the
sgi's allows the pc's to see the sgi disks
in the network neighborhood. Our sgi stopped talking to the
spectrometer today. The other nmr spectroscopist had to call me at home
because he hadn't seen it happen before. Happens about once a year."
"I don't have any experience of an Avance with an NT workstation, but
we recently took delivery of a new Unix equipped solids machine.
The standard SGI O2 currently being supplied is a lovely machine with a
300MHz
processor and a 19 inch monitor plus lots of memory, I think 128
MBytes, very fast indeed. I would go for unix every time. The only
slight question might be long term support, but operating systems seem to
be diversifying again at the moment so that might not be a problem. In any
case it is likely to be relatively inexpensive to replace the O2 with a
PC some time in the future, should it become necessary, and you might
as well have the benefit of a proper
computer while you can."
"I would advise against NT. I am currently in the process of
replacing the SGI INDY on my DMX (used 30 % for solids) and
thought that it might be time to convert to NT." I decided not to because
"bugs are currently far more numerous on NT than SGI. I will be
buying an SGI O2 within the next few weeks."
"If you want to develop an additional skill set that is very much in
demand, go unix and devote all your spare effort to becoming an expert
unix administrator. Even intermediate ones will always be able to put food
on the table for the foreseeable ten years or more."
Tom Walter
tom_walter@waters.com
This archive was generated by hypermail 2b29 : Sun Dec 31 2000 - 01:15:01 PST