School rumination

[Bev Vincent <brv@msc.com>]

On Mon, 7 Apr 2003, a.l.spek wrote:

> My feeling is that 1-2 days is rather too short for a computing school,
> and 3 days is really an absolute minimum.  I'm not sure that I would worry
> so much about access to computers...get people to bring laptops.
> Do we need to cater both for windows and linux though?
> Web access might be necessary.

I attended a Crystallographic Computing school back in 1986 in East
Germany where the only computer present was an Apple that Henk Schenk
brought behind the Iron Curtain with vague worries that he might not
be allowed to take it back out of the country at the end of the session.
He used it to demonstrate his SIMPEL program in action, but only as a
supplementary demo to the theoretical presentation.

Hands-on computer experience is certainly useful, but equally of interest
to me back then were the lectures on computing subjects that were done
without anything more than an overhead projector. I remember in particular
David Watkin's talk on least squares refinement. Having him walk us through
a pencil and paper refinement was very instructive. Also Bill Clegg's talk
on computer graphics -- the basic concepts of how you start with a set of
coordinates and end up with something on the screen taking things like
overlap, rotation, perspective into account. All done conceptually.

> Several program systems are mentioned...is there any intention to include
> these in the hands on session? Some are commercial - are limited licences
> available for these?  Perhaps we should stick to free software, as this is
> more about programming methods and access to a compiler is going to be
> more appropriate.  Under these conditions it may actually be better to get
> people to use their own compilers & computers -ie what they're used to.
>
> Interfacing with databases might also be included in the list of subjects.

In an earlier e-mail I saw my name suggested for macromolecular computing.
Probably not the best idea, since I have little programing experience and
no structural experience in that field. I've been mostly into the small
molecule end of things.

I'd suggest automation as a topic, including:

   * crystal mounting
   * automatic crystal centering
   * automating data collection algorithms
   * unattended processing
   * automatic structure solution and refinement

with all their attendant benfits and pitfalls.

Bev