Gordon,
I fully agree with your assessment that PyMOL is currently unable to reach
its potential due to a number of factors, including poor documentation.
Nevertheless, I do have a defined plan for eliminating these deficits. It
involves securing funding to hire people under contract to document, support,
and develop PyMOL on an ongoing basis. While I am grateful for the continued
assistance of PyMOL users on numerous minor tasks, it is not reasonable to
expect anyone to prepare a complete well-written manual or to perform other
major tasks without some form of compensation. Also, PyMOL is still immature
enough in certain areas that a full manual really can't be constructed until
the 1.0 features are settled upon (soon hopefully!).
Understanding this situation, several large pharmaceutical companies and
many smaller entities have lent support to PyMOL development through donations
and license purchases. At present the flux is low, but we are getting to a
point where some real headway can be made. The forthcoming Incentive PyMOL
(iPyMOL) add-on will hopefully increase the flux by providing a specific
reward to those who fund the PyMOL development effort.
I know it may sound like PyMOL/iPyMOL is becoming commercial gray-ware, but
that's not it at all. I haven't devoted thousands of hours to this project
just to see it end up as yet another expensive, lousy, inflexible, and
closed-source commercial tool. PyMOL will remain a vital open-source project,
and iPyMOL will exist solely to accelerate PyMOL's growth and guarantee its
long-term health. Nurturing PyMOL and similar efforts to fruition on such
terms is central to my career vision -- it's what I stand for in this industry.
I believe that well-supported, well-documented, and well-maintained
open-source software will be a superior research product to what is currently
offered by the major scientific software vendors. My hope is that scientists
in industry and academia will eventually come to appreciate this, and then
choose to make creation of such software economically viable through funding of
balanced approaches such as the one I am taking with PyMOL.
Free software is great once it exists, but there are far too few developers
in this field to expect the kinds of dramatic successes achieved by the Linux,
GNU, and BSD communities. If open-source/free-software is going to have a
large impact on molecular modeling, it will need direct support from users,
universities, governments, and for-profit companies. If really good
open-source tools were available and able to meet the needs of research, would
you be willing to pay reasonable amounts of money for their development,
maintenance, and support? I hope so.
Thus, pharmaceutical companies like EMD/Lexigen/Merck can choose to
continue to spend millions of dollars on traditional closed-source packages
with no hope of a reprieve, or they can choose to funnel a small fraction of
that money into efforts like PyMOL. Over the long term, open-source packages
offer vendor-independence, cost-effectiveness, flexibility, opportunities for
innovation, and support for broad sharing of information and technology. These
are all key ingredients in scientific progress, both within a single
organization and across the industry.
Realistically, it will take a year or more before PyMOL can fulfill the
potential you described. How soon this happens will depend in part on whether
people like yourself are content to merely gripe about the status quo, or
whether you are prepared to commit resources towards improving it.
Cheers,
Warren L. DeLano, Ph.D.
PyMOL Creator
Founder, DeLano Scientific
-Original Message-
From: gwebs...@lexigenpharm.com [mailto:gwebs...@lexigenpharm.com]
Sent: Monday, October 28, 2002 10:33 AM
To: pymol-users@lists.sourceforge.net
Subject: [PyMOL] PyMol Manual
Fellow PyMol Users
Although I like using PyMol, I would like to submit a gripe about the severe
lack of a decent PyMol manual.
A deeper description of the program's features is required if people are
seriously going to use PyMol for tricky things beyond simple visualization
(simple visualization programs are ten-a-penny and PyMol could be so much more)
Particularly lacking are details of structural editing (mutation, bond
rotation, cut-and-paste etc.), interfaces (if any exist) to routines for enegy
minimization and a detailed description of the Python command line interface
with examples of scripting and how to extend PyMol's capabilities.
By neglecting to adequately document what may be a really useful program, the
PyMol user community might fail to attract the kind of user base and support
that it deserves and that it needs to ensure its success.
A good, clear and well-thought out manual really helps to make a case for using
a particular piece of software (and it might also lower the odds of PyMol
having to join the gray