From eslone@mason1.gmu.edu Sun Feb 13 00:22:00 1994 Received: from mason1.gmu.edu for eslone@mason1.gmu.edu by www.ccl.net (8.6.4/930601.1506) id XAA07754; Sat, 12 Feb 1994 23:31:32 -0500 Received: by mason1.gmu.edu (5.65/DEC-Ultrix/4.3/GMUv3) id AA26319; Sat, 12 Feb 1994 23:31:31 -0500 Message-Id: <9402130431.AA26319@mason1.gmu.edu> Subject: Chemistry TrueType & Adobe Fonts To: chemistry@ccl.net Date: Sat, 12 Feb 1994 23:31:31 -0500 (EST) From: "J. Eric Slone" X-Mailer: ELM [version 2.4 PL23] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1009 Dear Netters: I have uploaded the most recent versions of my chemistry font to the ccl.net archive. The font and macro program are for Windows PCs, and the font is provided in Adobe and TrueType formats. The file is currently in the /incoming directory, named Chemistry_Font.zip. Please disregard the files named chemfont.* (I forgot to upload them as binary and they are unreadable). Comments and suggestions are welcome; Eric ________________________________________________________________________________ J. Eric Slone Scientific Consulting Services Serving Government & Industry Since 1982 Internet: eslone@mason1.gmu.edu Compuserve: 73757,2776 "True science teaches, above all, to Fax: (703) 751-6639 doubt, and to be ignorant." Voice: (703) 461-7078 Miguel do Unamuno ________________________________________________________________________________ From eslone@mason1.gmu.edu Sun Feb 13 14:22:06 1994 Received: from mason1.gmu.edu for eslone@mason1.gmu.edu by www.ccl.net (8.6.4/930601.1506) id NAA11101; Sun, 13 Feb 1994 13:50:35 -0500 Received: by mason1.gmu.edu (5.65/DEC-Ultrix/4.3/GMUv3) id AA19630; Sun, 13 Feb 1994 13:50:36 -0500 Message-Id: <9402131850.AA19630@mason1.gmu.edu> Subject: Chemistry Font To: chemistry@ccl.net Date: Sun, 13 Feb 1994 13:50:36 -0500 (EST) From: "J. Eric Slone" X-Mailer: ELM [version 2.4 PL23] Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1279 Dear Netters: The chemistry font which I announced previously has been moved to its new home. The files are now available on anonymous ftp/gopher on www.ccl.net in /pub/chemistry/text-processing/chemfont directory. File chemfont.zip contains all files in the distribution, though you can also retrieve them individualy. If you need unzip program for PC it is available in pub/chemistry/software as unz50p1.exe. You can also retrieve files by e-mail (binary files will be uuencoded) by sending a message: select chemistry limit 250kb ls cd software/chemfont get README get chemfont.zip quit to MAILSERV@ccl.net. As always, comments or flames are welcome! Eric ________________________________________________________________________________ J. Eric Slone Scientific Consulting Services Serving Government & Industry Since 1982 Internet: eslone@mason1.gmu.edu Compuserve: 73757,2776 "True science teaches, above all, to Fax: (703) 751-6639 doubt, and to be ignorant." Voice: (703) 461-7078 Miguel do Unamuno ________________________________________________________________________________ From news@nntp-server.caltech.edu Sun Feb 13 17:22:10 1994 Received: from punisher.caltech.edu for news@nntp-server.caltech.edu by www.ccl.net (8.6.4/930601.1506) id QAA12201; Sun, 13 Feb 1994 16:55:39 -0500 Received: from gap.cco.caltech.edu by punisher.caltech.edu with SMTP (8.6.4/DEI:4.41) id NAA21639; Sun, 13 Feb 1994 13:55:39 -0800 Received: by gap.cco.caltech.edu (4.1/DEI:4.41) id AA12703; Sun, 13 Feb 94 13:55:33 PST To: mlist-chemistry@nntp-server.caltech.edu Path: terry From: terry@wag.caltech.edu (Terry R. Coley) Newsgroups: mlist.chemistry Subject: OOP in Comp. Chem. and other ideas Date: 13 Feb 1994 21:55:28 GMT Organization: California Institute of Technology, Pasadena, CA Lines: 146 Message-Id: <2jm7kg$cct@gap.cco.caltech.edu> Nntp-Posting-Host: sgi1.wag.caltech.edu For the past two years various members of the Goddard research group at Caltech have been using advanced software techniques for computational chemistry. We will be talking about some of the results at the ACS meeting in San Diego this March. There are several areas of discussion when software approaches are considered: 1) how to make programs flexible and accessible to users 2) what language to use 3) what design techniques to use (OOD, etc.) 4) how to make code reusable In short, our approach has been to use whatever works. This somewhat lengthy post presents some of the approaches that have been working for us in each of these areas: 1) Accessibility/Flexibility. This is perhaps the most important point of all and is related to previous CCL discussions of "tools" for computational chemistry. Much code falls into two categories: a) monolithic programs b) simple, single purpose programs Type a) becomes difficult/impossible to maintain/extend without an in-depth knowledge of the entire program. Data structures tend to be so inextricably mixed with each other that there is essentially no way to fully test a new change for correctness and lack of side-effects. Type b) result in the chemist spending lots of time trying to make input and output file formats compatible. This is the case because research naturally tends to want new results combined in ways not previously considered. Our programs are all run using a Tcl (Tool Command Language, see comp.lang.tcl) embedded interpreter. Traditional programs read input, setup and perform a calculation, and output results. In contrast, we try to *enforce as little overall structure on the flow* of events as possible. Our objects, even the ones that perform calculations, are *interpreter-driven*, and are written as generally as possible. Major algorithms are implemented by simple Tcl commands with operate on internally defined objects. The course of a calculation is dictated by the Tcl script written by the user. This allows usrs to then easily recombine objects to perform different calculations as necessary. Most major data structures (e.g., Matrices, Molecules, Atoms) and their corresponding member functions are available to the Tcl script. What does this mean in a practical sense to the user? Suppose you have one of our programs and you need to output a final basis set overlap matrix in a particular form to be analyzed by another program. Without changing the executable, it is possible to write and test a simple Tcl script to output data from the program in whatever form is needed. All this can be done interactively... if there is an error in the script, just edit it and try again - no need to rerun the calculation that generated the result. Example: Suppose you wish to run a calculation using a specially generated set of one electron integrals. Without modifying the program, all that is necessary is to write an interpreted script to input the desired matrix (or perform a transformation on the current one). No change to the program is needed. It is hard to overstate the usefulness of this concept. A program designed in this way may be reconfigured to perform calculations not forseen by the author. Furthermore, all our Tcl-based programs can communicate with each other via TCP/IP. A single Tcl script can coordinate the flow of input and results from several different programs at once. 2) Language. Our code is primarily based on C++, but we make use of Fortran where it is the most appropriate tool. C++ can represent complex data structures more naturally; some numerical operations are done in Fortran or using reliable libraries. Example: Our C++ matrix classes are implemented with member functions that use the Fortran LAPACK routines for fast, parallel matrix operations. 3) Object oriented design. We have made use of object oriented design where it is most natural. In our current focus on quantum code, the data representing things like atoms, molecules, and basis sets are collected together with the code that implements operations on them in C++ objects. Essentially, what we are designing are libraries, but ones where the code/data structure closely reflects the organization of the ideas represented. Encapsulation of the data with code affecting our objects is very important for generating software that does not have arbitrary (and difficult to debug) interdependencies on other code. Features of the language, such as operator overloading, make the code easier to read, and assist in usability and reusability. We realize some of this is a bit philosophical, but we do think that it is pragmatically important. When software does not clearly parallel the way we conceptualize things and algorithms, confusion often results, especially when multiple programmers are involved. In our experience C++ (and probably other modern languages) leads naturally to data structures that mirror chemical ideas. 4) Reusability By reusability, we really mean two things: o Generality of coding o Maintainability of coding To a large extent good general code can be incorporated without change into new code. Maintainable code also leads to reuse by being approachable and extensible by a programmer other than the author. What are things that don't lead to reusability? Odd subroutine interfaces, confusion about what a routine actually does, silly output (or weird formats), overzealous speed optimization. Our use of C++ allows us to design so that the input and output are much more intuitive; the code is hopefully easy to read, so that it is more easily understood what a routine does. Unnecessary speed optimization is a frequent culprit in the production of unreadable, unmaintainable code. Our approach is to write code in the most straightforward (some might say verbose) way first. After correctness is verified, profiling is done. Only code that requires significant CPU-time need be rewritten. We have also been putting C++ templates to use by making certain container classes as general and reusable as possible. We hope to have this code ready for free public consumption via ftp around the time of the March ACS meeting. Although we have found the ideas useful in our work, the code is still young. We would like others to experiment using Tcl and TCP/IP for creating programs which can communicate with each other. Terry R. Coley, Ph.D Erik P. Bierwagen Goddard Research Group, Caltech Goddard Research Group, Caltech terry@wag.caltech.edu epb@wag.caltech.edu From Bohdan.Szymanik@vuw.ac.nz Sun Feb 13 21:22:13 1994 Received: from kauri.vuw.ac.nz for Bohdan.Szymanik@vuw.ac.nz by www.ccl.net (8.6.4/930601.1506) id VAA13505; Sun, 13 Feb 1994 21:06:19 -0500 Received: by kauri.vuw.ac.nz id AA08419 (5.65c/IDA-1.4.4 for chemistry@ccl.net); Mon, 14 Feb 1994 15:06:05 +1300 Date: Mon, 14 Feb 1994 15:06:05 +1300 From: B Szymanik Message-Id: <199402140206.AA08419@kauri.vuw.ac.nz> To: chemistry@ccl.net Subject: RIETAN Hi Netters, don't know if this is the wrong forum for this question but I'll try neway. Does neone know how I can get hold of the FAT-RIETAN package and in particular the program PREMOS that does Rietveld refinement on modulated structures. It'd be just perfect for a problem cell I'm looking at at the moment. I tried writing to the author F. Izumi but haven't received a reply so maybe I got the address wrong. Any clues appreciated, Thanks, Bohdan Szymanik Victoria University/Industrial research Ltd b.szymanik@irl.cri.nz From Jie.Yuan@UC.Edu Sun Feb 13 23:22:14 1994 Received: from ROLL.SAN.UC.EDU for Jie.Yuan@UC.Edu by www.ccl.net (8.6.4/930601.1506) id WAA13813; Sun, 13 Feb 1994 22:26:32 -0500 Received: from UCBEH.SAN.UC.EDU by UCBEH.SAN.UC.EDU (PMDF V4.2-14 #4918) id <01H8UPPLAWLC922VNM@UCBEH.SAN.UC.EDU>; Sun, 13 Feb 1994 22:25:28 EST Date: Sun, 13 Feb 1994 22:25:28 -0500 (EST) From: "JIE YUAN, CHEMISTRY, U. CINCINNATI" Subject: IUPAC.sit To: chemistry@ccl.net Message-id: <01H8UPPLBZ6A922VNM@UCBEH.SAN.UC.EDU> X-VMS-To: IN%"chemistry@ccl.net" X-VMS-Cc: YUANJ MIME-version: 1.0 Content-type: TEXT/PLAIN; CHARSET=US-ASCII Content-transfer-encoding: 7BIT Someone posted an article that he has put the nomenclature-to-graph program, IUPAC, on the osc archive. Can you also send it to info-mac archive? I have checked info-mac, and it was not in the chemistry software directory. The info-mac archive is the most popular site for Mac users. Cheers! ===== Jie Yuan === Chemistry === U. Cincinnati === Jie.Yuan@UC.EDU =====