New/Developing "Linux and Crystallography" Page at CCP14.
Lachlan Cranswick
There is a new (and hopefully growing) Linux and Crystallography
resources page at CCP14. Additional info and pages to link
to by the resident Linux enthusiasts are welcome. (Especially for
"out of the box" binaries and rpms that can be installed with the
minimum of hassle - in the same way that makes Windows/Macs so
Powerful)
http://www.ccp14.ac.uk/solution/linux/
At the moment, the main speel (given most PC users are on Windows -
myself include) is a (possibly incoherant) tutorial on installing and
configuring a Redhat 5.2 Linux/MS-Windows 95 dual boot system as a possible
crystallographic server/workstation - which is one of the most obvious
hurdles. Detailed tutorials on installing dual-boot Linux seem to be
lacking/not readily visible on the internet(?)
http://www.ccp14.ac.uk/solution/linux/linwin95a.html
Again, comments welcome as much of the Redhat Linux partitioning
and installation did not work as advertised.
(I won't be killing off Windows 95 on my office desktop PC for the
forseeable future - especially with the large variety of high
quality ready to use binaries that are available).
Topics at Present (some incomplete):
Linux Overview
Which OS on a Crystallographic Client System on PC -
Linux, Win95/98 or WinNT?
Installing Redhat Linux and Win95 - Dual Boot
Linux as a High Performance Crystallographic Server
Linux Resources
Crystallographic Applications and Resource for Linux
----
Please note that if Crystallographic authors/users of Windows software
are getting worried about what the commercial MS-Windows 2000 might be
like; there is a new project called ReactOS which is trying to do for
Windows what Linux did for UNIX. In this case a free Windows
complient operating system, possibly better written than the
commercial version if what occured with Linux is any guide.
Again, very new/embryonic, but worth keeping an eye on and perhaps
worth encouraging:
http://www.sid-dis.com/reactos/
New ReactOS Mirror at CCP14 - http://reactos.ccp14.ac.uk
Lachlan.
Lachlan M. D. Cranswick
Collaborative Computational Project No 14 (CCP14)
for Single Crystal and Powder Diffraction
Daresbury Laboratory, Warrington, WA4 4AD U.K
Tel: +44 (0)1925-603703 Fax: +44 (0)1925-603124 Room C14
E-mail: l.cra...@dl.ac.uk
CCP14 Webpage (Under heavy construction):
http://www.ccp14.ac.uk
Lachlan Cranswick
>Date: Tue, 15 Dec 1998 10:18:28 -0500
>From: "Paul D. Boyle" <bo...@laue.chem.ncsu.edu>
>Subject: Re: New/Developing "Linux and Crystallography" Page at CCP14.
>Newsgroups: sci.techniques.xtallography
>X-Newsreader: TIN [version 1.2 PL2]
>
>Hi Lachlan,
>
>In article <l.cranswick....@dl.ac.uk> you wrote:
>: There is a new (and hopefully growing) Linux and Crystallography
>: resources page at CCP14. Additional info and pages to link
>: to by the resident Linux enthusiasts are welcome. (Especially for
>: "out of the box" binaries and rpms that can be installed with the
>: minimum of hassle - in the same way that makes Windows/Macs so
>: Powerful)
>
>That is an interesting view. In my opinion, that is one of the
>primary weaknesses of those platforms. One of the growing problems in
>the crystallographic community (again, my opinion), is the increasing
>tendency to distribute binary executable only distributions. For example,
>SnB and DIRAX are the first programs which come to mind. I think this
>is something which needs to be discussed in the community, and I was
>wondering if you mind if I responded publically in s.t.x to your above
>statement in hopes of evoking responses from various crystallographers.
>Maybe get a little flame war going? :-). I'll wait to hear your response.
[stuff deleted]
This is obviously a multi-faceted issue that could make a few volumes -
thus the following is restricted to a few "rantlets".
Also, if considered relevant, I do not program or write source code.
----
Many "crystallographers/scientists" now use "Crystallography" as a research/
analysis "tool". These people have no real day to day need to be programmers
to get the "job done". They may have never touched a compiler in their life.
I would suggest that for these "mainstream" crystallographers,
using quick and easy to install "TRUSTED" binaries is far superior over
supplying "specialist" source code for compiling on local computers.
Getting a specialist scientific program to compile and execute can be a
major hurdle for some; getting it to compile, run, and "run correctly"
can be extra hurdle of a non-trivial nature.
For many scientific programs, probably the only people qualified to
compile "trusted" binaries are those intimate with the code - most likely
the authors or other very active with the code?
Supplying source code as a "quality assurance" feature
would be a nice if this was possible; but the present research/funding/
commercialisation environment can make this difficult for many people.
--
However, if there is a threat to crystallography as a progressing
science (as Paul's response might indicate), my opinion
is that it is the loss and destruction of the existing
(and future) "specialist skills base" (full time crystallographic,
fundamentals, mathematical, programming, etc) to research, develop,
distribute and maintain available quality crystallographic software
and methods.
New and existing crystallographic software authors are effectively discouraged to
develop software due to lack of funding and appropriate recognition for these
arduous research endeavours.
With much contract based employment and downsizings, job/career stability
is an issue. How can a researcher properly learn principles and
techniques to implement their ideas within a short term contract?
Often, programming is considered a form of "non-research" time wasting.
Combined with an emphasis on publishing "lots" of "research" papers,
looking elsewhere to a more financially rewarding "career" outside
scientific programming must look very attractive to competent, and often
quite brilliant scientific programmers.
Much of this was type of argument was submitted as a "Letter to the Editor -
On Software Availability" and published in the most recent British Crystallographic
Association newsletter (Dec 1998, pp 25-26) titled:
"The Importance of New and Readily Available Crystallographic Software and the
People Who Author It."
The emphasis of this article was based on "my opinions" concentrating on the
situation in Britain (though I believe this may not be restricted to the
British research review and funding system). I guess any rebuttal arguments
on these potentially misguided and ill-informed opinions will be published
in the next issue. A summary of sorts is:
New, "readily available" software and algorithms for academic research
is essential for the development of crystallography as a progressive
science; without this, mainstream crystallography grinds to a snail's
pace.
People who presently develop crystallographic software are treated
like "dirt" by the research assessment and funding bodies. The
arduous, intellectual research endeavours that is crystallographic
algorithm and program development; and which provides exponential benefits
to the wider crystallographic and scientific community is not recognised
as such. (this creates a "variety" of follow-on effects in response
to this situation).
It is argued that a resulting effect will be the slowdown in the development
of mainstream crystallographic research.
It is argued that a slowdown is already visibly occurring (despite highly exciting
publications in the literature implying the contrary should be happening)
in the field of Structure Solution from Powder Diffraction data;
due to the above situation and resulting lack of readily/freely
available software for academic research.
(as a contrast to how Shelx, Sir, Dirdif, PATSEE, et al for single crystal
diffraction are available)
"Is it also just coincidence that some of the most respected names in
crystallography also supply freely available "benchmark" software
for academics?"
It is argued that funding bodies would have to be "completely clueless"
not to recognise and respond in a positive manner to this situation.
(though possibly in a more tactful manner - I will await any rebuttal
articles or "other responses" for clarification of this)
The BCA newsletter web-page is at:
http://gordon.cryst.bbk.ac.uk/BCA/Cnews/Index.html
But this version (67 Dec 98) of the Hardcopy newsletter has not made it
electronic yet.
---
One positive action that in my opinion would be of assistance to this rather
grim situation is for the relevant research and funding councils to
appreciate just how important the scientists who develop new algorithms and
techniques for the crystallographic community.
Scientific programming is an arduous and major intellectual research
endeavour on the part of the programmer/scientist. It is fundamental to the
positive development crystallography and the ability of the users of this
software to routinely solve more complex and challenging crystallographic
problems. The present decision and funding system lacks any sound basis
for not recognising the intellectual research endeavours undertaken to
develop "useable" scientific code through grants and support.
Armel Le Bail
> Many "crystallographers/scientists" now use "Crystallography" as a research/
> analysis "tool". These people have no real day to day need to be programmers
> to get the "job done". They may have never touched a compiler in their life.
> I would suggest that for these "mainstream" crystallographers,
> using quick and easy to install "TRUSTED" binaries is far superior over
> supplying "specialist" source code for compiling on local computers.
> Getting a specialist scientific program to compile and execute can be a
> major hurdle for some; getting it to compile, run, and "run correctly"
> can be extra hurdle of a non-trivial nature.
> .......
I completely agree ;-). Although a quite not brilliant scientific programmer
stuck in FORTRAN code, I am fortunately able to do some little jobs and
know how to use some compilers.
Just a very trivial example. With the new Bruker D8 diffractometer,
it seems to be recommended to record 10 patterns at 2 sec/step
rather than 1 pattern at 20 sec/step. OK. The EVA software provides
only the possibility to realize a A+B sum. So that you should repeat
this 9 times in order to sum your 10 patterns !!!
Suggesting to the EVA (or XCH) developpers to modify the software
will take one year and I will probably not obtain the new version
for free. So that, the best is to do a small software myself (and
I will not call that "research" - but is melting elements and
heating "research" when the result is unknown , and even if it
was known ?? Many crystals are made in this way - I would
say that "research" is the contrary of "routine", so that
"routine research" should be impossible:).
If somebody has already done this program, contact me please
(preferably with source code) !
--
Armel Le Bail - Universite du Maine, Laboratoire des Fluorures,
CNRS ESA 6010, Av. O. Messiaen, 72085 Le Mans Cedex 9, France
http://www.cristal.org/
Graham A. Patterson
Cranswick) wrote:
>----
[...]
>For many scientific programs, probably the only people qualified to
>compile "trusted" binaries are those intimate with the code - most likely
>the authors or other very active with the code?
>Supplying source code as a "quality assurance" feature
>would be a nice if this was possible; but the present research/funding/
>commercialisation environment can make this difficult for many people.
Looking at this from the software development angle:
If there is no difficulty in releasing source code on a public domain
basis (and not all employees have full intellectual property rights),
what are the considerations?
Pro:
The program doesn't die if the author moves on to other things and the
originating laboratory does not maintain it.
You can examine the code to verify that it works as you expect.
You can make additions and extensions for your self. Though that may
bring in copyright, even under public domain releases.
If you compile it yourself, you do have some confidence that the
program you are running isn't a trojan.
Con:
The author loses control of extensions and modifications unless he or
she is prepared to co-ordinate and make modifications in a timely
manner. Finding time to write software is hard enough, convincing
'the management' that maintainance is desirable can be impossible.
People can examine your code to verify that it works (and review your
programming style!)
Not everyone will be successful in compiling the source. And that will
generate a lot of mail about systems the author may not know.
The ideal would be binaries, with source available on request. I
usually use pre-compiled binaries for GNU applications, but there are
times when the source is useful. And if I really wanted a package, I
might be prevailed upon to do the binary for a sytem myself.
>--
>New and existing crystallographic software authors are effectively discouraged to
>develop software due to lack of funding and appropriate recognition for these
>arduous research endeavours.
So you have the ludicrous result that months, if not years of work
developing a program is judged solely on a short paper describing the
core tenets.
>With much contract based employment and downsizings, job/career stability
>is an issue. How can a researcher properly learn principles and
>techniques to implement their ideas within a short term contract?
And that supposes that the researcher is also trained and skilled as a
programmer. Or has access to a collaborator who is so trained and
skilled.
--
Graham Patterson http://www.reading.ac.uk/~slspatte
Sedimentology: The hard study of soft rocks
*** Opinions expressed are mine unless otherwise stated ***
Paul D. Boyle
> >Date: Tue, 15 Dec 1998 10:18:28 -0500
> >From: "Paul D. Boyle" <bo...@laue.chem.ncsu.edu>
> >Subject: Re: New/Developing "Linux and Crystallography" Page at CCP14.
> >Newsgroups: sci.techniques.xtallography
> >X-Newsreader: TIN [version 1.2 PL2]
> >In article <l.cranswick....@dl.ac.uk> you wrote:
> >: There is a new (and hopefully growing) Linux and Crystallography
> >: resources page at CCP14. Additional info and pages to link
> >: to by the resident Linux enthusiasts are welcome. (Especially for
> >: "out of the box" binaries and rpms that can be installed with the
> >: minimum of hassle - in the same way that makes Windows/Macs so
> >: Powerful)
> >
> >That is an interesting view. In my opinion, that is one of the
> >primary weaknesses of those platforms. One of the growing problems in
> >the crystallographic community (again, my opinion), is the increasing
> >tendency to distribute binary executable only distributions. For example,
> >SnB and DIRAX are the first programs which come to mind. I think this
> >is something which needs to be discussed in the community, and I was
> >wondering if you mind if I responded publically in s.t.x to your above
> >statement in hopes of evoking responses from various crystallographers.
> >Maybe get a little flame war going? :-). I'll wait to hear your response.
<Lachlan>
I sent you private email because I did not intend to make what I wrote
above public. It is contrary to "netiquette" to post private email in
a public forum without the prior consent of the author.
</Lachlan>
<regrets>
To the authors of SnB and DIRAX, it was not my intent to slight any of
you or the work you have done. On the contrary, I hold both pieces
of software and the intellectual effort behind them in high esteem
(particularly DIRAX), and my disappointment around the unavailability
of the source codes comes from my desires to either understand what the
programs does and/or to port the software so it better integrates into
the computational environment in my laboratory.
</regrets>
With regard to providing source code, I think there are a number of
reasons why providing the source code is an important issue.
1) It is educational. Today, crystallography is a computationally
intensive field. Access the crystallographic source code allows younger
crystallographers to see the details of crystallographic calculations. It
allows them (us? I'm not *that* old), to gain a deeper of understanding
of how what we read in text books has been actually implemented.
Can we stand on the shoulders of our predeccessors and develop a
better implementation, or one better suited to our particular needs?
"Canned programs" inhibit this aspect of crystallographic education.
And it is showing.
One crystallographer I spoke with once about this had the opinion that one
can be an decent crystallographer without knowing how to program, but to
be a really good crystallographer, one needs to be able write programs
which implement crystallographic calculations. In my experience, the
little crystallographic programming I have done has definitely promoted
imtimacy with the subject matter. Ten years out of my Ph.D., my goal as
a crystallographer is to develop an ever deepening my understanding of the
field. I feel like I have only just begun this "real" in depth education.
Access to source codes, and writing programs is an important part of my
continuing education.
2) It allows for better, faster innovation in software. When others
have access to the source code they can contribute bug fixes, and adding
functionality to the software. One of the things I have learned in
using Linux and GNU software is that openness promotes better software.
In the Unix community, the various GNU utilities are generally recognized
as being superior to their conventional Unix counterparts. "Open Source"
software such as Linux and Apache have the reputation as solid pieces
of software largely because many people have volunteered their time
and intellectual effort to make them better, more robust, faster.
In contrast, my perception is that crystallographic software is written
by small number of people, and the rest of us just use it. I would love
to see a broadening of the effort. Granted, not everyone has the time
or inclination to write a complete crystallographic program package,
but I think we should be doing more to encourage people to contribute
to writing crystallographic software as their means and abilities allow.
Fix a bug, send a patch to the author(s) can be enough.
3) Having the source code and being able to do some programming allows
us freedom to do what we need to do so that we can do our jobs better.
The danger in relying on "canned" programs (especially for service
crystallographers (of whom I count myself among)) is that it reduces our
ability to act as independent scientists. Service crystallographers, in
particular, are vulnerable to being reduced to button pushing technicians
whose only response can be, "the software doesn't do that" when they
are asked to do some particular task. This can only further lower the
perceptions/opinions of our chemist colleagues that crystallographers are
not "real" scientists.
4) Source code availability allows us to port programs to work in our
own computing environent. For example, suppose an author only releases
DEC UNIX executables of a program he or she wrote. I have only an SGI
or Linux boxes. Without the source code, my choices are restricted to
either buying a DEC Alpha or not using the program. With the source codes
it is a non-issue; I port the code, send the author the patches of the
the modifications I made (if any) to get the program to compile and run.
Everyone benefits.
I am sure there are more reasons for "open sourcing" crystallographic
software. This problem is not new, nor is it an isolated concern.
The last ACA newsletter the "Crystallographic Data and Computing Committee
Annual Report" mentions the diconcerting trend in data processing
software in that authors are not releasing their source codes. Also,
back in 1988 Le Page, Gabe and White wrote a Letter to the Editor of
Journal of Applied Crystallography about some of the problems caused
by the lack of source code. For those of you who are interested, see
J. Appl. Cryst. (1998) 21, 993-994.
> This is obviously a multi-faceted issue that could make a few volumes -
> thus the following is restricted to a few "rantlets".
>
> Also, if considered relevant, I do not program or write source code.
>
> ----
>
> Many "crystallographers/scientists" now use "Crystallography" as a
> research/ analysis "tool". These people have no real day to day need
> to be programmers to get the "job done". They may have never touched
> a compiler in their life.
I think there are several shades of gray which are getting washed out
in this argument. One doesn't need to a programmer to compile and
install software. I don't think one even needs to be a programmer to
tweak source code to get it to compile.
> I would suggest that for these "mainstream"
> crystallographers, using quick and easy to install "TRUSTED" binaries
> is far superior over supplying "specialist" source code for compiling on
> local computers. Getting a specialist scientific program to compile and
> execute can be a major hurdle for some; getting it to compile, run, and
> "run correctly" can be extra hurdle of a non-trivial nature. For many
> scientific programs, probably the only people qualified to compile
> "trusted" binaries are those intimate with the code - most likely the
> authors or other very active with the code?
I guess I disagree here. If one is going to work in a field which
technical and computationally intensive, then one needs to know a
little something about how computers and programs work. With regard
to installing software this minimally includes knowing how to use
configuration scripts and make files.
If a program is difficult to install, one can either fix it, or write
to the author explaining for what platform one wants to compile the
program and what are problems. Or better, one could do both, and
everyone benefits.
Maybe this is related to what Lachlan wrote about getting the "job
done", but I have noticed a growing trend among graduate students
to be reluctant to do anything that is outside of a strictly defined
scope of their project. They have the focus on "getting the job done"
rather than the broader approach of learning what they need to learn in
order to get the job done. It is in the learning of what one needs to
learn that a good part of our education and job skills are developed.
Very often, that is where the struggle, challenge, and insight are the
richest.
> New, "readily available" software and algorithms for academic research is
> essential for the development of crystallography as a progressive science;
> without this, mainstream crystallography grinds to a snail's pace.
I agree with you completely. As I pointed out above, having authors to
distribute sources can do a lot to encourage others to contribute and
quicken the overall pace of development.
>
> People who presently develop crystallographic software are treated like
> "dirt" by the research assessment and funding bodies. The arduous,
> intellectual research endeavours that is crystallographic algorithm and
> program development; and which provides exponential benefits to the wider
> crystallographic and scientific community is not recognised as such.
> (this creates a "variety" of follow-on effects in response to this
> situation).
While funding would be nice, I don't think it is necessary. It hasn't
been a historical trend and I don't expect that to change in the future.
Nevertheless, much high quality software both inside and outside
the crystallographic community has been written without any funding.
What you do need is people willing and able to write programs. It will
go better if it can be done collaboratively; which in my mind means
distributing the sources.
later,
Paul
--
Paul D. Boyle | bo...@laue.chem.ncsu.edu
Director, X-ray Structural Facility | phone: (919) 515-7362
Department of Chemistry - Box 8204 | FAX: (919) 515-5079
North Carolina State University |
Raleigh, NC, 27695-8204
http://laue.chem.ncsu.edu/web/xray.welcome.html
kkol...@pop3.concentric.net
put into them).
I think the Problem is more-generalized than is the view you've
expressed. I'm also a programmer. I work-for-hire so that I can do
scientific explorations just for the Joy of it ("weird", I know :-)
but what I wish to point out is that there's a Ruthless propensity to
"Borrow", without giving Credit, that Transcends the delimited scope of
"you-and-I as programmers", but which never-the-less, reduces to the
stuff to which you point in your posts... folks "racing" (Trampling)
each other to garner the "rewards of having been 'first'".
I've experienced such first-hand... even though I'm an Amateur, the
Professionals have endeavored to "Borrow" Everything I've done, even
though I've done it out of Love for Science.
The Dynamics with respect to which you've posted Transcend the scope in
which you've discussed them, Lachlan.
When I Ponder such, it's flat-out Easy to See that, in such, Science is
Dying.
It's a Sorrow. ken collins
Lachlan Cranswick
First I should apologise to Paul and everyone for my being more
clueless than usual and misreading the headers on a "private" email;
which had a >Newsgroups: sci.techniques.xtallography line
which I had (mis)interpreted as being Cc'd/crossposted to the
newsgroup inviting a response.
I myself did not interpret a slight on any software author, probably
as this type of topic of "closed source"/"open source" is a
open discussion on other newsgroups and web pages,
especially amongst Linux related groups. In comparison, I
feel my response and opinions may have been been
offensive to people interested in maintaining and
increasing standards within the crystallographic community.
"Should a minimum standard amongst practising crystallographers (which I
include to be people "using" crystallographic software) be the ability to
routinely validate the crystallographic code and/or enhance the
source they are using on demand?"
"Should there be an insistence on crystallographic software being
"open source"?"
While I am in agreement with the above, it is not a practical
option if the underlying research system will not support this.
If this is the case, a goody amount of pressure for a positive
outcome should be directed on the root causes for doable solutions,
not the effects.
----
If terms of present trends.
Existing crystallographic software (especially small molecule
crystallographic software) is so sophisticated and well written
that it "is" routinely used as a tool without a fundamental
understanding of the mathematics and fundamental science
behind it. Many crystallographers may be using crystallographic
software as a means to an end; as one part (possibly quite
small) of an overall project.
(However, it is still possible for responsible "users" of crystallographic
software to validate there results using "independent" packages and
freely available separate "quality checking" software such
as Platon. Users who primarily use either Shelx "or" Crystals,
should be able to cross validate results. There are other possible
options available for cross validation depending on the structure problem.)
Should any crystallographer be treating their software as a "toolset "
in varying degrees of "black boxiness"?
Trends/pressures in publication, funding, tenure/very short term
contracting, how credit is proportioned, etc, imply "mainstream
evolution" could be favouring those who use crystallographic
software as a "tool"; and those who give the research/assessment
system what it wants to see. With many national borders being tighter,
can many scientists interested in general employability afford
to take a deeper interest in many of the methods they are using,
other than as "tools" to keep them employable? (food on the table
argument)
---
On the subject of releasing source code and collaboration.
There are trends, counter-trends, arguments and counter arguments
One of many elements being that institutions that are increasing
focussed on maximising external commercial funding; making freely
available release of source (and executables) increasingly difficult.
Releasing code and executables can hold no funding benefits for the
institutes. While a dodgy statement without knowing the sums, it
can be argued some of the most successful commercial crystallographic
programs are also ones that are also freely available to academics
in some form or another.
Much scientific/crystallographic code may show it's origin
as a "proof of concept" before being released. It may have a large
portion of the author(s) personality attached to the code and
could be a major undertaking to understand even if the reader
has a thorough understanding of crystallographic principles.
It is possible that only the original author may be able to modify this
code with any confidence such that it will still work as advertised.
Thus if modern implementations are required, it may make more sense
from a "quality" point of view to recode it again using modern
languages and algorithms. Would such an programming option
gain acceptance from a funding and research assessment level?
----
"Is does not mean ought". I hope people have not been gaining
the impression that because I have the opinion that things "are"
this way, they "ought" to be this way. As has been stated already,
releasing source code is a great quality checking device; would
allow the option of hacking through the code and creating new
software derivatives.
While there are obvious benefits to crystallography to have it all
"open source", why hasn't it happening? Again, my opinion is that the
present publishing/funding/evaluation system not only does not
encourage this, but provides an active disincentive. The same
goes for the use of "more depth" underlying the use of
crystallographic software.
This is very difficult to solve if the assessment/funding mechanisms are by
their actions saying "just get the job done". As effective crystallographic
programming requires "a major depth of knowledge" and appreciation
of the fundamentals, this does not bode well for future crystallographic
software development, and people who want to make use of this software.
Armel Le Bail
> While there are obvious benefits to crystallography to have it all
> "open source", why hasn't it happening? Again, my opinion is that the
> present publishing/funding/evaluation system not only does not
> encourage this, but provides an active disincentive. The same
> goes for the use of "more depth" underlying the use of
> crystallographic software.
>
> This is very difficult to solve if the assessment/funding mechanisms are by
> their actions saying "just get the job done". As effective crystallographic
> programming requires "a major depth of knowledge" and appreciation
> of the fundamentals, this does not bode well for future crystallographic
> software development, and people who want to make use of this software.
My opinion is that the behavior of sharing is an individual mind state.
I guess that academic researchers having lost the traditional
behavior of disclosing how science works are part of a new
clan of "for personal profit" researchers, expecting to become
a kind of Bill Gates of science. They expect both butter and the
money for butter, doing commercial jobs at academic places. This
is the new modern science : market science. This is perhaps
due to some lack of trusting in the reward system of science.
Those days, almost every crystallographer has a micro computer
on its desk, connected to networks, including Internet. Companies
providing licensed softwares imitate Bill Gates. One new
license per computer. Many laboratories have not enough money for that.
The conclusion is that there could be one expensive software installed
on one computer in common use. Possibly, this software is unused
because most researchers install a freeware doing almost the same
job on their own machine ;-). This is my case. When you need to
use this unique commercial software, there is already somebody
using it... Progress against piracy have rendered illegal copying
more and more difficult. CD-ROM are delivered with a master
diskette working the exact number of licenses you bought, no more.
Of course, I never practiced piracy, anyway.
Now, my contribution to free softwares of the day :
The small utilitary performing sums of powder patterns gathered
in a Bruker .uxd file is now available (Win95 executable +
Fortran source + manual + example) together with other softwares
at : http://www.cristal.org/also.html ( see UXDSUM)
This did not required "a major depth of knowledge and appreciation
of the fundamentals". Anyway, who can do the more can do the less.
Small softwares adapted to your needs make life easier, if you can
obtain them for free.