***  For details on how to be removed from this list visit the  ***
***          CCP4 home page http://www.ccp4.ac.uk         ***



It's very odd, but I was just having this exact conversation myself, and would be interested to see what people are doing. I just convinced my vice president to give me some money so that I can build a test computer before making the investment to replace 6 old dying SGI's. I too have old SGI's (Indigo II's, about $300 each on e-bay), and am wrestling with replacing them for structure solution (complete package, from molecular graphics to xplor/ccp4/etc...). I have a few additional requirements that I would also like to know about here:

1. I really want 3D capability, meaning NuVision/Crystaleyes. For this, one needs a high end graphics card, running around $2000 as far as I can tell. I'm willing to purchase this, but want to know what platform people find best first, as I don't want to set this up in Linux and then find that I should have been using OsX for example. G5's will run this card (instead of waiting for the new macs), but of course any pc will also work with this card (Nvidia Quadro FX 4500 is one that specifically works on macs, there are a few others that are similar and just as good for all I can tell, but I think they only work on pc/linux).

2. I'm also spoiled in that I would like some sort of dial box thing. I find them convenient to use, though if I could get the same result from switching to a space ball for example, I'd be willing to try.

I use Turbo-Frodo for modeling, I guess I could change, but right now I really don't want to. I know coot is out there, as is O and other programs, and I could learn to use those I guess if the reward is worth it. Right now I don't have to change so I simply haven't, and as I get older change is getting more difficult (dang, never thought I'd say that). Turbo uses dial boxes for input devices, and I'm not certain it will recognize other alternatives (like space balls as I mentioned above).

Along these lines, I already have much of the hardware from my present graphics lab (emitters/dial boxes/glasses), and thus I just want a new system that I can port this hardware over to (and thus not have to re-buy anything).

As you can read between the lines, I'm torn at throwing away these awesome $300 computers and replacing them with something that will be at minimum $3500 to do the same thing (video card for $2000 and computer for around $1500). It's true that I could give up stereo and dials for speed and convenience, but do I really have to? I think I should be able to buy a new computer that has the same qualities/capabilities as my old SGI.

Also, as mentioned by Sergei, is it worth waiting for this new Intel/Mac chip that will run all platforms and thus will be a "generic" computer. Will we have to wait for software, or will the software simply compile on it with no issues. I've waited this long, so a little longer won't hurt, but if waiting won't necessarily buy me anything, I'd like to get this started now.

I did put Linux on a computer this past summer, however, I will admit it was an older computer and so it is a bit slow. It wasn't fun to install, and in fact while I did just follow directions, it wasn't something that I enjoyed doing. If linux is the os of choice, what version of linux is best to buy/get??? James Vergis commented that Red Hat is the way to go, and I'm more than willing to do that if it is easier/nicer to use than Fedora (which is what I tried). Is that what most people find?

Thanks much.

David Roberts
[EMAIL PROTECTED]

At 07:34 AM 4/19/2006, Sergei Strelkov wrote:
***  For details on how to be removed from this list visit the  ***
***          CCP4 home page http://www.ccp4.ac.uk         ***


Dear All,

like many other groups, we are currently facing the need
to move from an aging SGI/Alpha cluster to a more modern
computing environment. The obvious alternatives to choose from
are Linux, OSX or Windows. We are aiming at solving interesting structures
rather than at developing software -- and yes, we only have a limited support from IT professionals.

1. MS Windows:
While many people would probably exclude Windows right away,
I am especially interested in collecting opinions about this option.

True, a Windows box can never be a full replacement for a Unix
workstation. However, many pieces of software (not least the CCP4/CCP4i)
are perfectly usable under Windows! The main advantage in our case,
and in most other places, probably, is that nearly all students use/own PCs
(desktops or laptops) already.

Do there happen to exist any web resources summarising the setup and
use of Windows for biocrystallography (something like the
awesome site by W. Scott on crystallography on OSX)?

It does seem that there are no Windows versions of several important
crystallographic programs/packages but I may be not up to date --
does anyone have a list of those that are available?

With Windows, there exist ways and tools that provide
modalities typical for unix environments -- how usable are
all these? How usable is cygwin?

2. OSX:
Clearly, there has been much attention to this platform
lately, and most Xtallographic programs have been ported.
Hardware is generally good. The setup/maintenance
is generally easier than for Linux.

One specific quiestion: is buying an Intel-based
Mac (and not a G5-based) is advisable at this stage
as much of the software is not yet optimised for Intels?

The well-known problem with Macs is a lesser choice
of general-purpose software. One particular difficulty
is that the OSX version of MS Office (especially the Powerpoint)
has never been optimised properly (IMHO).

3. Linux:
Many advantages and one big concern: the difficulty of setting up.

I would be very grateful for any feedback, please write either
to the BB or to me directly. I will post a summary of all replies.

Best wishes,
Sergei.

--
*NEW* address and e-mail:
Prof. Sergei V. Strelkov, Protein Crystallography Group Coordinator
Laboratory for Analytical Chemistry and Medicinal Physicochemistry
Department of Pharmaceutical Sciences, O&N2
Catholic University of Leuven
Herestraat 49 bus 822, B-3000 Leuven, Belgium
Phone(w): +32 16 33 08 45   Fax:    +32 16 32 34 69
Phone(p): +32 16 23 19 06   Mobile: +32 486 29 41 32
E-mail:   [EMAIL PROTECTED]


Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm


Reply via email to