From jle@world.std.com Wed May 12 22:21:32 1993 Date: Thu, 13 May 1993 02:21:32 -0400 From: jle@world.std.com (Joe M Leonard) Message-Id: <199305130621.AA01442@world.std.com> To: chemistry@ccl.net Subject: Info re: AVS/Tools discussion Folks, I'd like to add my two cents worth in the tools discussion that crept into the AVS arena... First, one must separate the AVS Chemistry Viewer from AVS - one is a commercial software product (built on AVS) and the other is a general-purpose visualization environment, suitable for use in Chemistry or just about any other discipline with 2D/3D visual data. In many ways, the Chemistry Viewer is to AVS as Insight II is to GL/Phigs/PEX/datastructures. MSI utilized AVS as the basis for their Chemistry Viewer for several reasons: 1. Using AVS, one has 99+% source code compatibility between any supported workstation (which is just about every vendor, including PC Unix). Also, AVS provides scads of graphics tools in modular form, handles distributed networks with transparent data translation, permits parallel execution of multiple threads and includes a full software renderer permitting full 3D graphics on 8- bit color X devices. Also, one can mix AVS with X/Motif panels, allowing one to create a rather impressive GUI with reduced/minimal effort. This permitted the developer (i.e. singular) to create the first version of the Viewer in about 5 months, and permitted additional machines to be supported by an FTP and a remake... 2. I feel that systems such as AVS represent the next level of programming tools for application development. Why keep reinventing the wheel - I'm interested in the particular algorithms that solve my problems, but I'm NOT interested in how marching cubes renders the data, etc. Modular development systems greatly simplify my problem, so I can spend my time debugging the "interesting science" rather than the computer science. 3. The cost of AVS is not terribly high, as I expect that there are academic and site discounts (and probably developer discounts as well). Things like Developer's Kits must be purchased to do more than run other people's code, and they cost bucks too... The real issue with AVS pricing is that software developed using AVS requires that the end-user purchase some form of an AVS license. When this is on the order of hundred's of dollars (i.e. PC Software), with fun things like the ability to change networks, etc costing more, folks probably won't mind so much... For all I know, the pricing might be there already! 4. Commercial software will always have more "neat things", since folks are willing to pay for such things. If one wants flexibility and extensibility, the commercial packages are far less valuable than they are to folks who want "boring" things like stability, customer service, etc. Compared to commercial workstation prices for Chemical modeling packages, AVS is probably pretty cheap (boy, various salesman could skewer me on this one - but I've been away from that part of the business for over a year so be nice...). 5. Probably all of what I said regarding AVS is applicable to Explorer. The cost is lower, as I recall, but the ports to other machines are 1-2 years behind AVS. Still, given the market position of SGI in Chemistry, most folks will probably read this from an IRIS of some form (independent of what they number-crunch on). 6. If somebody wants the "content" of an AVS module w/o AVS, all that needs to be done is replace the graphics and communications calls with whatever mechanism is being used... Things are written in C and F77, so it's pretty easy to move from AVS to GL with the right primitive library or skill set. 7. Dan Severance (dan@omega.chem.yale.edu) made several comments about the tools philosophy... Except for purchase price, AVS has all of the requirements he stated. Again, Explorer probably does also... Long winded, aren't I? It's just that I feel it's important to make sure that folks know what's being talked about... If you have particular questions, feel free to email! Joe Leonard jle@world.std.com P.S. Please send all flames directly to me to reduce net overload... P.P.S. AVS comes with 10-20 Chemistry modules, many with full source in both C and F77 From jabs@chemie.uni-halle.dbp.de Thu May 13 10:52:50 1993 Date: Thu, 13 May 1993 08:52:50 +0200 From: jabs@chemie.uni-halle.dbp.de Message-Id: <930513085249*/S=jabs/OU=chemie/PRMD=UNI-HALLE/ADMD=DBP/C=DE/@MHS> To: chemistry@ccl.net Subject: MD Hi, has anybody experiences in the molecular dynamics calculations on bilayers or cluster of hydrogen bonded molecules? Thanks, Andreas From h.rzepa@ic.ac.uk Thu May 13 09:14:31 1993 Message-Id: <9305130815.AA03815@cscmgb.cc.ic.ac.uk> Date: Thu, 13 May 1993 09:14:31 +0000 To: CHEMISTRY@ccl.net From: h.rzepa@ic.ac.uk (Henry Rzepa) (Henry Rzepa) Subject: Re: Tools, tools, tools, and more tools. > The main sticking point I see with AVS is the cost. The original > poster was (I think) hoping for a cooperative effort to develop > public domain tools that everyone could use, regardless of budgetary > constraints. Then data and coordinates can be sent for point A to > point B with no concern over differing data formats, and electronic On this point we have an extensive set of chemistry "tools" for Explorer V2, written in C code, which we will be releasing soon. Much of the functionality of AVS/Chemistry viewer, and in some aspects, more, will be available. We are currently discussing how to make this available, but the intention is certainly to focus on LOW COST and source code availability!! Before someone asks, Explorer will be availble on a number of platforms very soon. Watch this space. Dr Henry Rzepa, Dept. Chemistry, Imperial College, LONDON SW7 2AY; rzepa@ic.ac.uk via Eudora 1.3.1, Tel:+44 71 225 8339, Fax:+44 71 589 3869. From h.rzepa@ic.ac.uk Thu May 13 09:08:55 1993 Message-Id: <9305130810.AA03317@cscmgb.cc.ic.ac.uk> Date: Thu, 13 May 1993 09:08:55 +0000 To: CHEMISTRY@ccl.net From: h.rzepa@ic.ac.uk (Henry Rzepa) (Henry Rzepa) Subject: Re: Net access to self funding databases & archives >Date: Thu, 13 May 1993 09:08:13 +0000 >To:jas@medinah.atc.ucarb.com (Jack Smith) >From:rzepa@ic.ac.uk (Henry Rzepa) >Subject:Re: Net access to self funding databases & archives > >Can I just raise the subject of Oxford Molecular. >They were set up to attract interesting >new code from young (or not so young) authors. The professionals >at OM would then "package the code" by debugging, etc etc, and resell >it to the world. Indeed they started with several codes originating fro >Ph.D. theses (dissertations in the USA) from Oxford students, >including Alchemy etc. A sort of semi-professional QCPE, where >the authors DO get royalties, but the end product is properly supported >and gone over by professional programmers. > >This has resulted (in the UK only) in the programs being made available >to academics at a relatively low price (600 pounds per site per annum, but >with a committment for 10 years ?). As I understand it, the industrial >price is MUCH larger! And I have no idea what the academic price is in >the USA (if indeed OM are marketing there). > >My worry is that programs which are either very specific, or are not >perceived as having any market potential, would not be supported. I >rather suspect that in its original aim of attracting and supporting new >groups, it is not doing too well (but if OM read this, perhaps they can >correct me!). Are people suggesting that QCPE become more like OM? > >Another site that has not been much discussed is MacSciTech, probably >because it is so machine specific. There are not many new Chemistry >programs submitted there, but it IS a free ftp site!! > Dr Henry Rzepa, Dept. Chemistry, Imperial College, LONDON SW7 2AY; rzepa@ic.ac.uk via Eudora 1.3.1, Tel:+44 71 225 8339, Fax:+44 71 589 3869. From kmoore@ncsc.org Thu May 13 04:25:57 1993 Date: Thu, 13 May 93 08:25:57 EDT From: Kevin Moore Message-Id: <9305131225.AA26941@duck.ncsc.org> To: avs@iris95.biosym.com Subject: Tools, tools, tools, and more tools. BTW, avs is not the only product that does all that you mentioned. There are many other commercial programs that do the same....including of course insightII from Biosym :-) While these programs do some of the things that the Chem Viewer does, they do not come close to interfacing gaussian and mopac to the extent and ease that the Chem Viewer does. The real power of AVS over these tools is the ability to add modules. While there maybe some scripting tools in InsightII and related programs to make it easier to manipulate data, the ability to add a different visualization method requires convincing the vendor that it is worthwhile. This is not true in AVS and though I have not used them I assume also for SGI's Explorer and IBM's Power Visualization tools. In these packages, the new method can be written using a standard data type that makes the programming effort fairly straightforward in most cases. The wealth of visualization modules that already exist provide a basis that probably does not require writing additional modules, but if it is necessary, one does not need access to the Chem Viewer source code to add new ones. These new modules will interact with the Chem Viewer just as easily as the Chem Viewer modules interact with each other. (Disclaimer: This is not a bash on InsightII, Sybyl or related. I use all of these packages regularly and they are *very* useful for the purpose they were derived. AVS has found a different niche.) ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ || Kevin Moore North Carolina Supercomputing Center || || Scientific Support Analyst 3021 Cornwallis Rd. || || (919) 248-1179 Research Triangle Park, NC 27709 || ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ From jtgolab@amoco.com Thu May 13 03:19:38 1993 Date: Thu, 13 May 1993 08:19:38 -0500 From: jtgolab@amoco.com Message-Id: <9305131319.AA16697@vacc47.nap.amoco.com> To: chemistry@ccl.net Subject: Vibrational Modes [Charles W. Ulmer writes . . .] > Date: Wed, 12 May 1993 09:44:43 -0400 (EDT) > From: culmer@stardent.chem.UTOLEDO.edu (Charles W. Ulmer II) > Subject: Tools, tools, tools, and more tools. > > [Partial message deleted] > > How many times have you > sifted through Gaussian Frequency output to determine if you had a transition > state? Why not animate the frequency and see it? The only draw-back is > that you must also have AVS. This is not really a draw-back though, since > AVS is so powerful. > > Charles W. Ulmer culmer@stardent.chem.utoledo.edu This is not completely true! If one looks in a recent QCPE bulletin, one will find a very nice program (tool? since it only does one thing) called VIBRATE. It is setup to animate vibrational modes from Gaussian on an SGI. Since one gets the source, it is very easy to modify for other output formats such as GAMESS, AMPAC, MOPAC, etc. It is very useful and doesn't cost much. :Joe jtgolab@amoco.com From DSMITH@uoft02.utoledo.edu Thu May 13 04:54:06 1993 Date: Thu, 13 May 1993 09:54:06 -0500 (EST) From: "DR. DOUGLAS A. SMITH, UNIVERSITY OF TOLEDO" Subject: Re: Tools, tools, tools, and more tools. To: lim@rani.chem.yale.edu Message-Id: <01GY4F5U5K16008RX8@UOFT02.UTOLEDO.EDU> File format conversion is currently being built into many larger packages if it is not already there. Gaussian has for many years supplied NEWZMAT, which is a stand alone program that converts many formats, but you must purchase a G9x license to get NEWZMAT. A few years ago we wrote a small stand alone program which converted between MacroModel, MM2, MM3, MOPAC/AMPAC, and Arkmod (a very old graphics package). We distributed this for a while, and are now in the process of building this into our AVS library of modules. Douglas A. Smith Assistant Professor of Chemistry The University of Toledo Toledo, OH 43606-3390 voice 419-537-2116 fax 419-537-4033 email dsmith@uoft02.utoledo.edu Doug From shenkin@still3.chem.columbia.edu Thu May 13 09:22:49 1993 Date: Thu, 13 May 93 13:22:49 -0400 From: shenkin@still3.chem.columbia.edu (Peter Shenkin) Message-Id: <9305131722.AA11760@still3.chem.columbia.edu> To: jle@world.std.com (Joe M Leonard), chemistry@ccl.net Subject: Re: Conf anal of cyclic ene's > When studying cyclic species with 1-2 double bonds (like > cyclooctene or larger), using the various searching methods out there, > are folks concerned about the cis-trans isomerization question? In > other words, do folks want to know about the min-E conformations of > cyclooctene OR the min-E conformations of cis-cyclooctene and/or > trans-cyclooctene? This arises not only in double-bonded cyclic systems, but also, for example, in cis/trans decalin. What one wants depends of course on.... what one wants! Cis/trans are configurational isomers, and are not likely to interchange under normal chemical conditions. Suppose, for example, you have natural product X that binds to known receptor R, and it is known that some linkage in X is cis. Suppose further that you wish to do a conformational search on X and later do binding studies between the low-energy conformations of X and R, in order to determine which conformer binds most strongly. In your conformational search you would not want to allow cis to isomerize to trans. On the other hand, if, based on non-structural criteria such as a QSAR search, you should propose that Y might be a potentially good binder to R, and if you then wish to do structural studies to confirm or refute this hypothesis, you might wish to allow cis/trans isomerism, since the results of the search followed by the binding studies might indicate which configurational isomer is likely to be more active. Finally, peptide bonds are an intermediate case, since they can isomerize slowly in water solution. -P. you would want to do a conformational search of X, without cis or the binding constant of the various conformations of prospective drug X with some receptor R, you ************************f*u*cn*rd*ths*u*cn*gt*a*gd*jb************************ Peter S. Shenkin, Box 768 Havemeyer Hall, Dept. of Chemistry, Columbia Univ., New York, NY 10027; shenkin@still3.chem.columbia.edu; (212) 854-5143 ************** The cost of flexibility is incomprehensibility. ************** From chm6@midway.uchicago.edu Thu May 13 03:57:01 1993 Date: Thu, 13 May 93 08:57:01 CDT From: "charles h martin" Message-Id: <9305131357.AA16738@midway.uchicago.edu> To: chemistry@ccl.net Subject: tools [...] >By the way, this brings up an issue that might be worth discussing >in general. Namely, is there a demand for chemistry "tools" rather >than chemistry "programs"? The difference is that a tool does only >one thing, perhaps with some options, whereas a program is more >monolithic. A tool is meant to be used in conjunction with other >tools and programs to accomplish tasks. On unix machines, tools >can be written using stdin and stdout so that they can be ganged >together in various ways and used in pipes in shell commands. In general, I would like to see more tools developed and shared. [...] >Ron Shepard >shepard@tcg.anl.gov I would like to answer this from the point of view from someone on the fringes of the quantum chemistry research world. >From what I can tell, there are many researches who simply use "... Errors are usually those of input, and these usually involve input geometry. Graphic interfaces are going a long way in removing this "traditional" source of error. ... " Clearly the Z-matrix <--> Cartesian Cooridinates routines are a step in the right direction. Indeed, I would love to use the simple input structure of something like Gaussian, but because it does not have an option to employ SIFS, I stay away because it is just to hard to interface with the highly specialized work we are doing. Having modular programs with robust data abstractions makes all the difference. To sum it up, packages (like COLUMBUS) which maintains a standard, robust data abstraction implemented in a modular fashion (like SIFS) facillitates faster and easier advances in electronic structure research. Within such a framework, the development of "tools" would (hopefully) naturally arise when other users develop their own software. Now I realize that this may just be a pipe dream as even in my own group it is difficult to convince people to all agree to program in a consistant fashion. In addition, being forced to program in Fortran-77 (as opposed to something like C++ or Fortran-90, for example) makes our work even more difficult. Never-the-less, the philosophy has still served us well and allowed us to make tremendous advances. Of course, I don't know about any general "tools" which our group has created yet, but in communicating with other COLUMBUS users, I have obtained some usefull stuff. (By the way, if there is anyone out there who uses COLUMBUS and SIFS for development of codes, I would like to open up a dialog so that we may indeed trade ideas about creating usefull "tools" for the package):) ================================================================== Charles H. Martin email: chm6@quads.uchicago.edu U.S. nail: c/o Freed Group The James Franck Institute and The Department of Chemistry The University of Chicago 5640 South Ellis Avenue Chicago, Illinois 60637 Work: (312) 702-3457 Fax: (312) 702-6853 ================================================================== From richard@TC.Cornell.EDU Thu May 13 06:45:56 1993 Date: Thu, 13 May 1993 10:45:56 -0400 From: Richard Gillilan Message-Id: <199305131445.AA25214@theory.TC.Cornell.EDU> To: chemistry@ccl.net Subject: Supercomputer Centers >3. As to budgets of Supercomputer Centers, they are not as large as some > may think (I am talking about publicly accessible supercomputer centers). > Of course, we will soon have no supercomputer centers, because workstations > are so terrific and cheap. I remember hearing this (and wishing it) when I was a graduate student in the mid 80's. The small supercomputing center (now extinct) that I worked with before the chemistry department purchased its first vax was notorious around campus as a hostile and awkward environment for getting science done. Working for one was the fartherest thing from my mind. We were in heaven when got got our first Sun workstations. It seemed to us that the writing was on the wall. Thankfully, the days of those narrowly focused cycle shops are mostly gone. Many fail to realize however, the extent to which today's supercomputer centers have diversified. If the ES/9000 here at Cornell were to vanish completely, for example, I doubt if it would make a serious impact. The center is heavily workstation and parallel-technology oriented with a research/educational institution-like atmosphere. The national centers (now MetaCenter) have positioned themselves well as general high technology centers of value to the nation's competitiveness. I doubt if they will go away soon. > Nobody however seems to realize that once you > have your workstation, you will have to buy/install/maintain all your > software on it and there will be nobody to ask around. Amen to that. It doesn't take many workstations to make for a full time system administration task. What a waste of graduate student and postdoc time! > And also, if you > do not believe that DIRECT MP4 can be easily done, you will have to buy > yourself a nice farm of big disks which will cost you 10 times more than > your CPU. > >In short, we better watch out so we do not throw out the baby with the bath >water. > >Jan Labanowski >Your Uncoordinator. >jkl@ccl.net From ssidner@unmc.edu Thu May 13 05:45:46 1993 Date: Thu, 13 May 93 10:45:46 CDT From: ssidner@unmc.edu (Steve Sidner) Message-Id: <9305131545.AA26408@molecular.unmc.edu> To: chemistry@ccl.net Subject: Re: QCPE, new guys, and bake sales. > From: mark-bower@theopolis > I am a new guy to the computational chemistry scene and > have been following this QCPE thread wondering what QCPE > stands for and also what it is/does. I decided to pick > on you since you made the comment about a bake sale which ... [I am forwarding part of my reply to the chemistry list in case other readers are not familiar with QCPE.] QCPE is the Quantum Chemistry Program Exchange. It was founded many years ago (10?) by Dr. Richard Counts of the University of Indiana, Bloomington, IN. The QCPE has every non-commercial computational chemistry program known to humanity. Richard is also a wealth of info on this subject. A 1 year membership is $40. This gets you the Bulletin 4 times per year, which lists new programs plus any information Richard and staff thinks is worth passing on to you. The Catalog is $3 (including postage!). This lists the complete holdings of the QCPE, and runs onto several hundred pages. QCPE Creative Arts Building 181 Indiana University Bloomington, IN 47405 USA Phone: (812) 855-4784 FAX: (812) 855-5539 Email: COUNTSR@IUBACS.bitnet Steve Sidner Eppley Institute, University of Nebraska Medical Center 600 S. 42nd St., Omaha, NE, 68198-6805, U.S.A. (402) 559-4921, FAX: (402) 559-4651 ssidner@unmc.edu ssidner@unmcvm.bitnet From scott@paganini.biocad.com Thu May 13 01:34:32 1993 Date: Thu, 13 May 93 08:34:32 -0700 From: scott@paganini.biocad.com (Scott Kahn) Message-Id: <9305131534.AA08913@paganini.biocad.com> To: CHEMISTRY@ccl.net Subject: RE: Conf anal of cyclic ene's In general, an individual knows whether a bond is cis or trans, in contrast to the stereochemistry on atoms (that may be known, unknown, or relative to other stereogenic centers). It would seem strange, therefore, to randomly sample olefin stereochemistry in a conformational searching algorithm. But what if this **was** desirable? When sampling diastereomers, surely it is meaningless to use energy as a measure of goodness. Diastereomers are different molecules that can have com- pletely different energy profiles. Unlike conformers, diastereomers do not normally interconvert under the types of conditions that most individuals are modeling - a boltzman distribution will not apply here. So, if diastereomers are to be sampled in a conf. analysis, it seems that the only prudent approach is to sample them all irrespective of the inter- diastereomer relative energies. I doubt that this type of sampling will be very useful in the cyclic diene case that prompted this response, but it **is** important in some drug discovery projects where atom stereochemistry is not completely known. ------ Scott D. Kahn, BioCAD Corporation (415) 903-3917 Email : scott@biocad.com From AHOLDER@VAX1.UMKC.EDU Mon May 13 06:00:36 1993 Date: 13 May 1993 11:00:36 -0500 (CDT) From: Andy Holder Subject: QCPE - The Official WORD To: CHEMISTRY@ccl.net Message-Id: <01GY4HETHHIQ9KP6G9@VAX1.UMKC.EDU> Hi Net Dudes and Dudettes! I just got off the phone with Richard Counts and he told me that all of our dreams will be a reality in the near future! QCPE is presently working the final bugs out of an ftp plan to allow its customers to get software directly from QCPE's computers. He said things should be up early this summer. It will basically work on the front end like it does now where you order and pay for the software. You will then be given an account and password to ftp into to grab the stuff. Sounds almost like what everyone wanted. Richard did ask me to post this, so it's the WORD. Andy Holder =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= DR. ANDREW HOLDER Assistant Professor of Computational/Organic Chemistry Department of Chemistry || BITNET Addr: AHOLDER@UMKCVAX1 University of Missouri - Kansas City || Internet Addr: aholder@vax1.umkc.edu Spencer Chemistry, Room 315 || Phone Number: (816) 235-2293 Kansas City, Missouri 64110 || FAX Number: (816) 235-1717 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= From HARPER_RICHARD_W@Lilly.com Mon May 13 07:15:10 1993 Date: 13 May 1993 12:15:10 -0500 (EST) From: Dick Harper 6-5990 Subject: Need to continue QCPE To: CHEMISTRY@ccl.net Message-Id: <01GY4JUNZS1E8WW0K5@INET.D48.LILLY.COM> Dear Netters, Without consuming inordinate quantities of bandwidth, I would like to agree with - >Don Boyd: >it could kill the goose that laid the golden egg to make QCPE programs >free. QCPE is a self-supporting organization and, I might add, has >done so very selflessly. If people have ideas on how to maintain >QCPE as a viable and reliable depository in this age of electronic >mail, they should offer them. I don't think QCPE has outlived its >usefulness. Code developers come and go, change jobs, etc., but some >institution needs to be in place to maintain a permanent repository. and - >Douglas Smith: >So, if it free and unlimited access were provided via internet, there >would soon be nothing to access. and - >1. Do we need centralized resources/archives. IMHO we need. > The QCPE is an example of a very successful one. The information you > get is timely. The manuals you get are often as good as the ones from > commercial sources. But most important, it is kept together, indexed, > and maintained, and rules about the minimal standard of submitted software > are enforced. The QCPE Bulletin gives news, and some publications there > are better than in leading journals. It is amazing how much you can do > with so little resources. > >In short, we better watch out so we do not throw out the baby with the bath >water. While life with QCPE has its inconveniences, it is hard to visualize a system that would more effectively serve the needs of the computational chemistry community. (That does not mean I we should stop thinking about how QCPE might be even more effective.) In reply to - >Jack Smith: >4) How about an un-scientific poll on how many would be willing to pay for >QCPE Internet access if a subscription mechanism could be worked out? If >so, how much? Should it be a flat fee or based on volume of service? >Should there be both academic and industrial rates? What should be the >redistribution rights of subscribers? Should program authors have any say? > And so on... I would personally have no problem with higher fees for access to QCPE materials. However, I don't pay for this service out of my own pocket, as I suspect some folks do. I think everyone's needs should be considered as the best compromise is reached. I think the program authors probably feel very comfortable with the existing distribution mechanism of distribution by QCPE, and with the integrity it assures their "products". IMHO, this coupled with the very reasonable prices charged for such distribution, and the informative Bulletin, account largely for the success of QCPE. Hurray for Richard Counts and QCPE! Let's think of ways to improve the effectiveness of QCPE rather than ways to supplant it. Dick Harper From: HARPER RICHARD W (MCVAX0::WALTZ) To: VMS MAIL ADDRESSEE (IN::"CHEMISTRY@ccl.net") From DSOUTH@uoft02.utoledo.edu Thu May 13 08:18:21 1993 Date: Thu, 13 May 1993 13:18:21 -0500 (EST) From: DSOUTH@uoft02.utoledo.edu Subject: chem libs. To: chemistry@ccl.net Message-Id: <01GY4L4L9DMU007C9J@UOFT02.UTOLEDO.EDU> From: IN%"lim@rani.chem.yale.edu" "Dongchul Lim" > It would be nice to put these tools (I mean source code) together as a library, > for example, "chemlib", as compared to "netlib" or "statlib". > We won't need QCPE soon, if we keep depositing our code to anon ftp sites > such as www.ccl.net, preferably in a library form, so that the code can > be reused without much modification. Along the same lines, why not develop a set of C++ classes for computational chemistry? Of course this only benifits the people that can program in C++, but the advantages would be an order of magnitude about the advantages gained by creating a chem.h. Perhaps Mark T. (of Argus) has something to add...I know he was trying to move Argus into the OOP universe. Then again, what AVS/Explorer/everyone-and-his/her-family are doing with their 4GLs (4th generation languages) amounts to this. Though it isn't obvious to everyone at first, working with AVS _is_ programming. With AVS available on platforms like MasPar (70K+ MIPS integer, AVS runs on the front end, not the DPU so there is little/no overhead for the pretty graphics, all for less than $1M) and an expanded Chem Developers Kit we would have a tool that allows the end users to simply _build_ whatever the situation required. What the AVS/Explorer/other-4-GLs do is take the unix concept (small, robust single purpose tools that can be combined to preform more complex functions) and transport it into a graphic enviroment. > One of most frequently asked questions seems to be how to convert between > different file formats. If we had a library for this one, we could simply > write the following pseudo code (in this case, using C) to convert gaussian > zmatrix to cartsian coordinates. With a sutiable 4GL, the user could just attach the required input module and the desired output module, along with whatever else he/she wanted to do to modify the data stream inbetween. Since AVS (and probably Explorer) already has a standard chemistry data type, the output of one module will be compatable with another module's input and vice-versa. > This way we won't have to repeat tedious works which someone might have done already. > The question is how much we are willing to share our source code... I agree 100%, I also think that the computational chem community should limit itself to JUST sharing source...making that neato-keen unix stdio tool into a AVS/4GL module allows others to use it without having to get their hands dirty. The only downside is that one has to be able to program in the AVS/4GL enviroment -- Not as hard as say Mac or X11 programming, but more difficult than standard Unix/DOS/VMS work. For those stuck in the 70's, AVS modules can be written in Fortran. :-) :-) [two smilies for the humor-impared] Dale University of Toledo From PCZCDJB@VAX.CCC.NOTTINGHAM.AC.UK Thu May 13 15:51:45 1993 Date: Thu, 13 MAY 93 20:32:31 BST From: PCZCDJB@VAX.CCC.NOTTINGHAM.AC.UK To: chemistry Subject: HyperChem, Gaussian and NT Message-Id: <2F802584_00075B48.0096C75631432FC0$10_1@UK.AC.NOTTINGHAM.CCC.VAX> Dear Netters, Does anyone know whether HyperChem and/or Gaussian '92 are likely to be ported to Windows NT when it finally ships? As this should allow versions to be produced both for cheapo PCs and for high performance Alpha-based workstations using the same source code it seems a fairly logical course for the vendors to take. By the way, I notice that PC magazine is claiming a fifteen fold increase in disk I/O performance under NT (the final beta version) with respect to DOS/Windows, even using 16 bit code. It would be interesting to see what effect this has on the performance of Gaussian in particular... ********************************************************************** Chris Boden Department of Chemistry, University of Nottingham, Nottingham NG7 2RD United Kingdom, Europe. Internet : pczcdjb@vax.nott.ac.uk From rsefeck@watson.ibm.com Thu May 13 10:27:04 1993 Message-Id: <199305131915.AA26975@oscsunb.ccl.net> Date: Thu, 13 May 93 14:27:04 EDT From: rsefeck@watson.ibm.com To: chemistry@ccl.net Subject: Visualization tools For those interested in visualization tools the following is a description of the IBM Visualization Data Explorer program, which like AVS and SGI Explorer is a general purpose visualization environment. As mentioned by another netter these types of tools provide more flexibility for researchers who are not just looking for a way of making molecular images but who are interested in real correlative analysis tools and ways of exploring their data visually. Like the others it is also not free, however if you are in an academic institution in the U.S. DX is available under the IBM HESC program (higher education software consortium). If your institution is in the 1H category you may be able to have an unlimited number of licenses for the cost of one. IBM has trial programs as well. DX also runs on non-IBM platforms Cornell University is also the Public repository (FTP sight) for user created modules and a set of chemistry specific modules are available there. The unified data model approach makes DX unique for importing data from just about any kind of program, modeling/simulation, experimental or observational data. IBM Visualization Data Explorer The IBM Visualization Data Explorer (DX) is a general-purpose software package for scientific data visualization. It employs a data-flow-driven client- server execution model and is currently available on five platforms: IBM POWER Visualization System (a medium-grain, coherent shared memory parallel supercomputer) and Unix workstations -- IBM RISC System/6000, Silicon Graphics Indigo and Crimson, Hewlett-Packard 700 and Sun Sparcstation 2. The client in this package is the graphics user interface. It utilizes XWindow and the Motif window manager, and always operates on a workstation. The server portion does all of the computation. It resides on a workstation or the PVS. In this sense, PVS provides significant scaling for larger data sets or more interactively or both. The server is controlled via a data flow executive, which determines what tasks need to be executed based upon user requests and schedules their execution. The executive accepts a well-defined protocol (a scripting language), which the user interface generates based upon input it receives. The executive can be operated independently of the user interface. It also supports a software cache. If a workstation has hardware support for three-dimensional graphics, then DX can utilize it. On IBM and Silicon Graphics workstations, gl is employed. On Hewlett-Packard workstations, Starbase is used. On Sun workstations, XGL is employed. The user has the ability to move back and forth between software and hardware rendering and progressive approximations in the same window(s) to tradeoff interactively vs. image quality, depending on the data and the workstation configuration. If a workstation only supports standard 8-bit XWindow, then 24-bit rendering is done in software, with the resultant images dithered to eight bits for display. Both hardware and software rendering support perspective and orthogonal cameras, multiple light sources and viewing, and arbitrary clipping. They both accept collections of opaque and translucent volumes, surfaces, lines and points simultaneously independent of color, grid size or type, opacity, etc. For hardware rendering, approximations (optionally under user control) are introduced for geometry not supported by the available hardware. The software renderer fully supports the integrated rendering with flat, Gouraud and specular shading. It has a two- pass design with near-linear speed-up for its parallel implementation. The first pass does calculations required per vertex. The second pass does scan conversion via an enhanced z-buffer (e.g., for irregular volumes, arbitrary clipping). In general, DX provides support for interactive visualization and analysis of both observed and simulated data. It is easy to learn and use. It supports interactions in a number of ways, including via a graphical user interface with direct (i.e., in images) and indirect (i.e., via Motif widgets) interactors, visual programming, a high-level scripting language and a programming API. This API provides data support, error handling, access to lower level tools, etc. for building modules. DX is very flexible. It is easily adaptable to new applications and data because it is built on a foundation of an integrated, discipline-independent data model. This data model describes and provides uniform access services for any data brought into, generated by, or exported from the software. DX supports a number of different classes of interesting scientific data, which can be described by its shape (size and number of dimensions), rank (e.g., scalar, vector, tensor), type (float, integer, byte, etc. or real, complex, quaternion), where the data are located in space (positions), how the locations are related to each other (connections), mesh dependency of data (i.e., node or cell center) or aggregation (e.g., hierarchies, series, composites, etc.). It also supports those entities required for graphics and imaging operations within the context of DX. Hence, regular and irregular, structured and unstructured data are supported as well as the notion of missing data (i.e., regions of invalidity). At the user level the details of the data model and its interface are hidden. This data model is fully supported by an external disk-based representation (format) for import and export. Data may also be imported through public domain structures like Unidata's netCDF, user-created custom filters and a general array handler. The general array importer provides a mechanism for the format or layout of data stored on disk, and then converts the data directly to the data model representation internal to DX. In addition, various image formats are supported for import (e.g., RGB) and export (e.g., Postscript, TIFF). An important consequence of this unified data handling approach is that operations in DX (modules) are polymorphic, interoperable and appear typeless to the user. In general, these operations follow a strict data flow protocol through pure function semantics. DX supports a number of realization techniques for generating rendering geometry from data. These include color and opacity mapping (e.g., for surface and volume rendering), contours and isosurfaces, histograms, two- dimensional and three-dimensional plotting, surface deformation, etc. for scalar data. For vector data, arrow plots, streamlines, streaklines, etc. are provided. Realizations may be annotated with ribbons, tubes, axes, glyphs, text and display of data locations, meshes and boundaries. Data probing, arbitrary surface and volume sampling, and arbitrary cutting/mapping planes are supported. DX supports a number of non-graphical functions such as point-wise mathematical expressions (e.g., arithmetic, transcendental, boolean, type conversion, etc.), univariate statistics and image processing (e.g., transformation, filter, warp, edge detection, convolution, equalization, blending, etc.). Field/vector operations such as divergence, gradient and curl, dot and cross products, etc. are provided. Tools for data manipulation such as removal of data points, subsetting by position, sub/supersampling, grid construction, mapping, interpolation, transposition, etc. are provided. There are also a number of relevant articles in the academic literature (e.g., conference proceedings, journals). These include: Haber et al, "A Data Model for Scientific Visualization with Provisions for Regular and Irregular Grids", Proceedings IEEE Visualization '91 Conference, pp. 298-305, October 1991 B. Lucas et al, "An Architecture for a Scientific Visualization System". Proceedings IEEE Visualization '92, pp. 107-113, October 1992B. Lucas et al, "A Scientific Visualization Renderer". Proceedings IEEE Visualization '92, pp. 227-233, October 1992 W. Ribarsky et al, "Object-Oriented, Dataflow Visualization Systems - A Paradigm Shift?", Proceedings IEEE Visualization '92, pp. 384-387, October 1992. L. Treinish, "Correlative Data Analysis in the Earth Sciences", IEEE Computer Graphics and Applications, p. 10-12, May 1992. L. Treinish, "Unifying Principles of Data Management for Scientific Visualization", Proceedings of the British Computer Society Conference on Animation and Scientific Visualization, December 1992. In addition, please see the December 1992 issue of GIS World and the April 3, 1992 issue of Science for some examples of DX applications. From topper@haydn.chm.uri.edu Thu May 13 13:48:59 1993 Date: Thu, 13 May 93 17:48:59 -0400 From: topper@haydn.chm.uri.edu (Robert Q. Topper) Message-Id: <9305132148.AA05707@haydn.chm.uri.edu> To: chemistry@ccl.net Subject: QCPE > > QCPE is the Quantum Chemistry Program Exchange.... (stuff deleted) > A 1 year membership is $40. This gets you the Bulletin 4 times per > year, which lists new programs plus any information Richard and staff > thinks is worth passing on to you. The Catalog is $3 (including > postage!). This lists the complete holdings of the QCPE, and runs onto > several hundred pages. See, this is a prime description of why the QCPE Catalog should be accessible online via Gopher. This way, both members and potential members of QCPE could search by keywords to see if there are programs available which will serve their needs. This would be much more efficient than ordering a catalog and receiving periodic hardcopy updates to the catalog. Moreover, I believe that more people would order more programs this way. Just my (last) two cents, Robert ******************************** * Robert Q. Topper, Ph.D. * * Department of Chemistry * * University of Rhode Island * * Kingston, RI 02881 USA * ******************************** * topper@haydn.chm.uri.edu * * (401) 792-2597 [office] * * (401) 792-5072 [FAX] * ******************************** "...what is truth?" -Johnny Cash