VMD-L Mailing List

From: John Stone (johns_at_ks.uiuc.edu)
Date: Wed Mar 03 1999 - 12:23:29 CST

Hi,

> I have just tried 1.3b2 on my solaris box. It runs OK, but I don't
> notice any difference from 1.2. Are there any user-noticable
> developments?

Well, it was partially my goal to make no large user-noticable
changes, but there have been a few actually...

1.3b2 should be user-noticably different in these ways:
  a) new snapshot code that works on non-SGI machines, and writes PPM files.
     (docs are behind on this, but there're a change or two more coming here..)
  b) new environment variable for what default image viewer to use.
     (docs are behind on this, but more changes coming here too)
  c) uses less memory per atom (about 56 bytes less actually) than 1.2 did,
     so 1.3b2 uses about 10-15% less memory total for a good sized system,
     than 1.2 did. This is nice if you're loading a 200,000 atom system :)
  d) New renderer support and some fixes that affected all of the external
     renderers. (New export code for my parallel ray tracer "Tachyon" ...)
  e) No longer needs Det/Eigen/Inv from the LASSP package for its measurement
     features.. (used to have to distribute these external programs with VMD)
  f) Uses the latest versions of Tcl, Tk, TclX, Mesa etc.
  g) Added support for VRPN, although this isn't quite perfect yet, and isn't
     documented at all yet (although there's a link to VRPN from our
     related programs / page).
  h) There are a few more fixes going into VMD for 1.3b3 that will affect
     all of the OpenGL versions of VMD, bringing the OpenGL versions much
     closer to the behavior of the original GL-based VMD. There are other
     fixes too, but they are still in progress. We're mainly trying to
     cut down the memory usage of VMD and eliminate old/dead code in this
     release. The next release of VMD (1.4) will have many more radical
     changes which will have a lot to do with the Win32 port, but I can't
     really outline what they are yet, since they are still in-progress..

There've been lots of other less notable changes to VMD between 1.3b2 and
1.2, maybe even a few other notable ones that I'm forgetting..
But, I'm glad that it "feels the same" to you, because one of my goals
was to accomplish a lot of these changes/fixes without altering VMD's
fundamental behavior.

Let me know how the Solaris builds run for you. They are built using
the latest version of Sun's OpenGL, so they are pretty bleeding edge :-)
They work fine for us, but its always good to get feedback from the users too.

If we get our Sun C/C++ Workshop 5.x compilers in shortly, I'll hopefully
be releasing VMD 1.3 final with binaries built by the newest compilers,
which supposedly yield an average performance increase of 12-15% according
to Sun, and according to people on the net. That would be very nice for
us, since our Sun Ultra 60 Elite3Dm6 is basically the fastest VMD box we
have presently. Another 12-15% faster would be even better :)

Any other questions on the betas anyone??

  John Stone
  vmd_at_ks.uiuc.edu 

-- 
Theoretical Biophysics Group   Email: johns_at_ks.uiuc.edu
Beckman Institute              http://www.ks.uiuc.edu/~johns/
University of Illinois         Phone:  (217) 244-3349
405 N. Mathews  Ave              FAX:  (217) 244-6078 
Urbana, IL 61801, USA          Unix Is Good For You!!!