VMD-L Mailing List

From: Axel Kohlmeyer (akohlmey_at_cmm.chem.upenn.edu)
Date: Wed Jan 23 2008 - 04:56:19 CST

michel,

[...]

> If you can get TCL scripting working for grant proposal writing,
> please let us know! :))

i suspect we first have to invent a new language for that.
GPL (Grant Proposal Language) would be a proper name. ;-)

e.g.:

if (num(postdoc) == 2) incr(paper(nature))
...

> > ...and then, of course, there is the problem, that (too?) many people don't
> > submit their scripts or script fragments to the VMD script library, so a lot
> > of useful stuff is lost when they move on to other projects.
>
> That is true. This could happen, in part, because we think that our
> lousy scripts are too simple or too specific and wouldn't be useful
> for other people. We also assume that there are other people who can
> do a better job at writing useful scripts, yet clean and elegant...

...but what about: if it is good enough for you, somebody else
could use and may even improve it. why should people have to
re-invent _everything_ when there is an option to just improve on
what is already mostly there. most people just need the ideas,
i.e. we can have a 'dump' for script fragments, for ideas, rather
than completed scripts. as you may have seen from the original
question, there is most likely no script that can do _exactly_ what
is needed, but there is a higher probability that somebody has
a framework, that could be adapted. most of the time this is sufficient
to get people started. but when there is nothing to look at.

... and don't forget. everybody has to start somewhere. nobody
is born to be an expert and how else can one learn to become
better than by getting feedback by others on the stuff that one
has written.

> > other than that, there is also the alternative of sponsoring somebody
> > with the proper expertise to write a custom script. since people spend
> > a lot of money on computers and (commercial) software, i am always
> > wondering why there is nobody offering 'incentives' to get their specific
> > problem solved (by and expert and quickly).
>
> this is because... Science is selfless and altruistic! :)

oh, i must have missed that one. ;-)

but seriously, i personally have found that helping other people
can be _very_ selfish. my 'drug' is solving (challenging) problems,
so if other people provide me with interesting challenges, i am
always inclined to give it a try (time permitting) and since in most
cases, people just need a few ideas or pointers, i can have the fun
of coming up with ideas, while the others have to do the legwork
of programming and in the end we're all happy (i have learned whether
my idea works without having to do it myself and the person asking
got his/her problem solved fast(er)).

also, since only few people want to do it, this is a "seller's market"
and over the time people have been very generous in many ways,
not to mention excellent job opportunities that start popping up.

> Seriously, I thought the same myself. For example, I consider myself a
> bad (really bad sometimes) programmer, so I usually find myself
> spending a couple of hours trying to write a simple script. This is a
> huge waste of time. Unfortunately, I'm not really sure if there are
> people out there who want to deal with these issues in a regular basis
> (even when getting incentives!). This might be in part because most of
> the people here want to focus on the science and get their work done
> (very valid thing to do). And the talented TCL-script writers (like
> you or John) are just too busy with many other things!

well, on one side, a lot might depend on the type of incentive,
but also, if people would realize, that there _was_ a market, this
might change a lot. i've recently some discussions with people
on this issue and more or less i came to the conclusion that science
needs to offer jobs (and train people for it) with a new job description:
something like scientific software support, much like the technical
support you need for running a spectrometer or alike.

with the ongoing invasion of computers and software into every
part of research, but the limited willingness to treat this in a professional
manner (how many clusters are managed these days by graduate
students or post docs who barely manage to keep them working
without making too much of a mess? how many clusters are run
in an very inefficient way, since the computer scientists running it
have no idea of what the workflow and requirements of scientific
computing are?) i am very concerned about what will happen in
the future. once the few idealists and 'dinosaurs', who were forced
to learn how computers and operating systems work inside out to
be able to use them at all, die out, we could run into a situation
where we have lots of facilities wasted, because there is nobody
around who knows how to use them (well).

cheers,
    axel.

>
> Cheers,
> Michel
>
> >
> > i'd be very interested to learn how people here on this mailing list
> > see the perspective of having their technical problems sort of
> > 'contracted out' so they can focus on the science...
> >
> > cheers,
> > axel.
> >
> >
> > >
> > > In doing that from prmtop/mdcrd (I am working with Amber in this moment) I
> > > guess that atoms names should be read from a pbd generated from prmtop/mdcrd,
> > > and given to the script in their precise format (I mean, for example, that atom
> > > names in ff99SB for the protein are uppercase, while those for the ligand in
> > > GAFF ff are lowercase).
> > >
> > > That applies also to Luis Gracia's RMSD Traj Tool that I am finding extremely
> > > useful and time saving.
> > >
> > > Thanks
> > > francesco pietra
> > >
> > >
> > > ____________________________________________________________________________________
> > > Looking for last minute shopping deals?
> > > Find them fast with Yahoo! Search. http://tools.search.yahoo.com/newsearch/category.php?category=shopping
> > >
> > >
> >
> >
> > --
> > =======================================================================
> > Axel Kohlmeyer akohlmey_at_cmm.chem.upenn.edu http://www.cmm.upenn.edu
> > Center for Molecular Modeling -- University of Pennsylvania
> > Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
> > tel: 1-215-898-1582, fax: 1-215-573-6233, office-tel: 1-215-898-5425
> > =======================================================================
> > If you make something idiot-proof, the universe creates a better idiot.
> >
>
> 

-- 
=======================================================================
Axel Kohlmeyer   akohlmey_at_cmm.chem.upenn.edu   http://www.cmm.upenn.edu
  Center for Molecular Modeling   --   University of Pennsylvania
Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
=======================================================================
If you make something idiot-proof, the universe creates a better idiot.