VMD-L Mailing List

From: Axel Kohlmeyer (akohlmey_at_cmm.chem.upenn.edu)
Date: Mon May 25 2009 - 16:10:47 CDT

On Wed, 2009-05-20 at 18:58 +0200, MyLinkka wrote:
> Hi, Axel,

hi ting,

> Thank you for the very educational reply! I can fully understand the
> difficulty in developing and
> maintaining an open source project. Therefore, there is no any complaint
> but just some suggestions
> in my previous email. (I carefully added quotation marks for some words
> there as maybe you've noticed.)

yes, i understand them as suggestions, i am just providing my personal
opinion and i think this is just as valid. it may be a character flaw
of mine, but it always irks me personally when people wrap their
opinions or suggestions into "world domination plans".

> I believe any software author wants more and more people to use his/her
> masterpieces, no matter for

i don't agree with this statement. i program first and foremost to
solve problems, and because i like show off, if i have found a nice
solution to a problem. of course when working on a "package program"
like VMD, one has to consider how much the existing user base would
be irritated by changes that would make it more difficult to use it.
particularly "more people would use it" is not much of a motivation
if the software already solves the problem it was written for.
in my observation, "i did it because i needed it" is far more
often motivation than "i did it because it would attract more people".

however, if possible, i try to make my work generally useful, as it
mostly needs only little extra effort and i also see it as my way to
pay my respect to those people back that put in the effort to write
a software package in the first place. the same goes for helping
people to use it better.

> what kind of reason, no matter if it is commercial or free. If this is
> true, then how many users using it
> does count, unless the software is designed for some particular users. I
> think this is one of the
> impetus that one software can become better. Most open source

i disagree with this statement, too. having the majority does not
automatically mean that something is the best decision. particularly
(non-commerical) open source software is the place where one can
write software the way once is comfortable with and not to please
as many people as possible. after all there is little financial
incentive (i don't have any gains or losses if more or less people
are using VMD or the code i contributed).

> contributors are software specialists or enthusiasts,
> but they often overlook the experiences of end users,
> whereas many commercial software really have to take it serious. (This
> is not my words, the saying can be found in some review papers of open

i cannot agree with this statement either. particularly commercial
software is often written in a "we did it because we say so" way
and you have to either adapt or not use it. getting commercial software
changed as a user is next to impossible, unless you have a lot of
leverage. the development process is frequently different. on many
levels marketing and management determine how it should look like
and what it should be able to do. serving the needs of users only
matters if it fits into the "management plan" otherwise it is just
an expense that will cut into the profits.

> source community) I think many of the vmd users are working in
> (bio)chemistry or material sciences.

the major issue is not a problem of _wanting_, but rather a problem of
not being able to do so. there is _very_ little (accountable) credit to
be had for working on scientific software, thus the number of people
maintaining open source scientific software is ridiculously small.

this is why it is so important that more people get involved to share
the work. sadly the recent trend is that while the number of users is
growing, the average need for help and support seems to be growing
while at the same time the willingness to step in and help out is
declining, at least in relative terms. this leaves more work for less
people and demands are mounting.

> Vmd and pymol are often compared as far as I know. Although one may
> say a person in this field who cannot program is not qualified, I do
> know many people with very little knowledge of programming or even
> scripting with tcl or pyhon in this field but they are excellent. Many
> of them are not using this kind of software to do research directly,
> but for the publishing or for simple visualization purpose.
> For them I think I don't have to talk too much about the importance of
> slick user experience. Of course the software authors could decide

i think a good user interface has to keep simple things simple,
make common usage patters easy, but should not try to make complicated
things _look_ simple, or try to add gimmicks that don't provide any
functionality, just to please some people who don't want to invest
a little bit of extra effort to adjust a tool to their needs.

> their preference, more 'intuitive' or more 'professional'. But I
> don't think it is that incompatible like oil and water. You are right,
> to improve the software, people should ask more but also
> share/contribute more. But to report problems is the first step. ;)

since the "problems" that you reported were no real problems,
my challenge to you would be to get busy and have a look at
the VMD documentation and try to find a good spot where the
suggestions given to you should be added (and then perhaps some)
and explained so that your fellow users would not run into the
same problem. if you believe that attracting more users would
be a good thing, that you would hopefully agree with me that a
good documentation is particularly important for this purpose.
...and unlike writing software, you don't need to be a trained
expert to write good documentation. in fact, i strongly believe
that people that have fewer insights into the inner workings
of a code are better qualified to explain it to others,
particularly beginners.

cheers,
   axel.

> ting
>
> > 

-- 
=======================================================================
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.