----- Original Message -----
> From: "Jensen, Andrew J" <andrew_j_jensen_at_groton.pfizer.com>
> To: <ammrl_at_chemnmr.colorado.edu>
> Sent: Wednesday, November 28, 2001 5:58 AM
> Subject: PC version of XWINNMR
Andy,
> We are evaluating the PC version of Bruker's XWINNMR licensed software,
> primarily for off-line processing in our offices. We would like your view
> of this software, whether you like using it or not, what the limitations
> are, and the future of PC vs SGI. Any information would be useful.
I'm not sure if you mean the NT PC version or the Linux PC version. I have
used both.
I use the NT version on my office PC and like it very much. It is very stable
and works exactly like the SGI version. In fact, I like the NT version of
XwinPlot better than the SGI version as it integrates much better to your standard
Office apps like PowerPoint, Word, etc. The only complaint that I have relative
to the SGI version is some clunkyness (sp?) in the colour mapping. At least on my
machine (specific to my graphics hardware?), you MUST set the display to 8 bit
(256 colours) in order to get smooth 2D intensity scaling. If you set the display to
16 bit or True Colour (24 bit) you can tell The Exceed X server to use Pseudocolour
and everything will work. However, 2D intensity scaling is no longer "real time". If
you change the threshold you must wait until you release the mouse button to see the results.
There are other occasional strange colour mapping problems when working in this
mode as well. They are annoying enough that I switch my display to 256 colour before
using XwinNMR for 2D (1D isn't much of a problem). This only takes a few seconds,
though.
The Linux version works just like the SGI version, but does have some colour
mapping issues with the Xserver supplied with the Red Hat Linux. You
essentially have to use 8 bit colour. Bruker's web site discusses this in detail,
as does Rudi Nunlist on one of his sites. Unfortunately, my Linux system
was hacked and trashed after being on-line for only about two weeks - despite
having all of the latest Red Hat updates and secure passwords on everything :-( The
system was used to look for Sun RPC holes all up and down the West coast of the US,
and then trashed to the point where the only recovery will be a complete
re-install. I have purchased a book on Linux security and will re-install when I have
time - with a severe hosts.deny file this time!
There is one complaint that I have with all versions of XwinNMR. This is the
non-standard way in which the text entry boxes in dialogs like EDA and EDP
work. They do not work like Motif, Tk/Tcl or Windows text entry boxes, and they
can drive the occasional user of the software absolutely nuts! And why must these dialogs
beep at me every time you change a value? If there is an option to turn this off,
I haven't found it (Version 2.6). The confusing thing is, some dialogs, like gradient
shimming (based on Tk/Tcl?) and Shape Tool (Motif) DO behave in a standard way.
Please, please, please! Remove this UXNMR holdover and rebuild EDA, EDP, etc.
with standard Motif text widgets.
As far as the future goes, I don't think it matters. The computer is now a small component
of an NMR system price, and the interface is now just a network cable. Changing
to a different computer in the future shouldn't be a big problem as long as the vendors
are reasonable about things like license transfer.
Cheers
-Kirk
Kirk Marat, Ph.D. NMR Facility Manager
Dept. of Chemistry and Prairie Regional NMR Centre
University of Manitoba
Winnipeg, Manitoba, CANADA
R3T 2N2
Phone: (204) 474-6259 FAX: (204) 474-7608
Email: kirk_marat_at_umanitoba.ca
Received on Wed Nov 28 2001 - 13:49:19 MST