VMD-L Mailing List

From: Timm Essigke (essigke_at_chemie.fu-berlin.de)
Date: Tue Aug 15 2000 - 07:47:51 CDT

John Stone wrote:
>
> Tim,
> As it turns out, this is in fact a bug in VMD 1.5. We also
> discovered another problem with the RMSD code that has been plagueing
> VMD since VMD 1.3, but we had not been able to debug until now.
> We have already fixed the RMSD computation problem (sometimes crashed VMD,
> particularly on Linux, and particularly noticable in VMD 1.3), and are
> now working on fixing the GUI problem you have also discovered.

Yes, I had this problem very often with VMD 1.3, but in version 1.4 it
seemed
to be stable to my experience (I can not remember a crash...).

>
> There might be a way for us to give a Tcl script to existing VMD 1.5
> users that circumvents the GUI problem, but the sometimes-crash or
> malfunction problem was in the measture_fit C++ code in VMD, and requires
> updating to a new binary of VMD. We'll probably be releasing an "a1" version
> of VMD 1.6 very shortly, and the RMSD fit problems will hopefully be
> cured once and for all.

For the moment I can use VMD 1.4 for RMSD computations, but I am looking
forward
to VMD 1.6.

> How are things going aside from RMSD fit? Any other comments/feedback?
> Anyone? How was VMD 1.5 /wrt to other features, stability etc? Comments?

The rest of the VMD 1.5 code seems to be rather stable for my work. I
encountered
no serious problems yet. The new feature to define materials is very
nice and is
working fine with POVRAY. Maybe it would be nice to add more predefined
materials
as glass, metal, plastic etc. (as parameterized e.g. in POVRAY), because
it is rather
complicated to adjust it via the dialog.
An old weakness in VMD (not a special VMD 1.5 problem) is the limitation
to 16
definable colors. Recently I wanted to display active side residues each
in an
other color and the hydrogenous of these residues in the same but a much
lighter
color. There I reached this limit very fast...
Usually it is no problem, but in these few cases...
I don't know, how many changes has to be done for setting this limit
e.g. to 256
colors. Maybe its to much work for these rare cases...
The new GUI in VMD 1.5 (Linux) was the most obvious change. I had to get
used to it...
What I missed most, is the right-mouse-button menu, which was much
quicker and
convenient as the mouse dialog in this version. But it had a good
effect, too:
it forced me finally to learn the keyboard shortcuts, which are even
quicker than the
old menu ;-)
But something else about this menu: I would wish to get a request ("Are
you really sure?")
for the "Reset View" button. I hit it once accidentally when I wanted to
press the
"Save Config" button...
Maybe a "Undo"-system would be fine, too, but is much more work to
add...
Additional to the ability to move two molecules independently from each
other (Mode move)
interactively, it would be helpful if they could be rotated
independently, too.

Some other problems I already reported to this list are the the command
line (Unix prompt)
options which don't work in all cases and that CONECT records in PDB
files are not read.
I would be nice to have an additional option at the UNIX prompt to start
VMD scripts directly
after initialization (like a play in the VMD window or a -script in
rasmol).

I hope I gave you some more ideas what could be added or changed to make
VMD even more powerful
and convenient to use.

Thanks for this great program!

Timm 

-- 
Timm Essigke
Freie Universitaet Berlin * Fachbereich Biologie, Chemie, Pharmazie * 
Institut fuer Chemie (Kristallographie) * Takustr. 6 * D-14195 Berlin *
essigke_at_chemie.fu-berlin.de * http://www.chemie.fu-berlin.de/~essigke *
Tel.: +49-30-838-54627 * Fax.: +49-30-838-53464