From chemistry-request@ccl.net Wed Sep 11 02:45:17 1991 Date: Wed, 11 Sep 91 01:37:03 CDT From: nicholas@mulliken.mcs.anl.gov (John Nicholas) To: chemistry@ccl.net Subject: MD neighbor lists Status: R Hi. I am currently working on parallel molecular dynamics simulations of molecular systems. I have tried various methods of forming the neighbors list, such as linked-cells and the standard Verlet neighbors. Does anyone have a way of making the neighbor list that they feel is particularly good for parallelism that they would like to share? I haven't found any scheme that seems to be significantly better than the others. Also, the neighbor list has to be made with the exclusion of atoms that are 1-2 or 1-3 (maybe 1-4) bonded. Thanks in advance for any suggestions. Please mail me directly because "reply" is not working well at our site. John Nicholas nicholas@tcg.anl.gov From chemistry-request@ccl.net Wed Sep 11 14:53:50 1991 Date: Wed, 11 Sep 91 20:29:37 +0200 From: ron.shepard@itc.univie.ac.at (Ron Shepard (anl)) To: chemistry@ccl.net Subject: Request for Survey Information Status: R In order to make some decisions regarding portability and programming practices for future distributions of the COLUMBUS codes, I would like to gather some information about various machines that are being used these days and some of the habits of the chemists who use them. The answers to any or all of the following questions by anyone wishing to respond (only 1 response/person please) would be most useful. Please reply to me directly in order to avoid cluttering the BBoard, and I will post a summary of the responses. Thanks in advance, Ron Shepard shepard@tcg.anl.gov ********************************************************************** General Computing Environment Questions: ---------------------------------------- (1.1) Are you primarily a user of "canned" packages, or a programmer-developer of your own codes? (e.g. 80%user/20%developer) (1.2) How many types of machines and how many machines of each type do you currently use? (1.3) Which machines do you share with other researchers, and which do you use essentially by yourself? (1.4) On which machines are you responsible for system software maintenence? (installing OS releases, compiler updates, etc.) (1.5) On which machines are you responsible for applications software installation and maintenence? OS Related Questions: --------------------- (2.1) What kind of operating system(s) do you use regularly in your day-to-day work? (MVS, VMS, UNIX, OS/2, Mac OS, etc.) (2.2) Which of these OS are POSIX compliant at the C interface level (or will be in the near future)? (2.3) Which of these OS are POSIX compliant at the user interface (i.e. shell) level (or will be in the near future)? Language Related Questions: --------------------------- (3.1) How much of your time (hours/week) do you spend programming in assembler? (3.2) How much of your time do you spend programming in C? (3.3) How much of your time do you spend programming in Fortran? (3.4) How much of your time do you spend programming an OS shell (C shell scripts, DCL command procedures, IBM JCL, etc.) in order to link applications together? (3.5) How much of your time do you spend programming in other Languages? (Lisp, Mathematica, ADA, BASIC, C++, WINGZ Spreadsheets, HyperCard, etc.) (3.6) How much of your time do you spend writing interfaces to libraries? (DISPLAY, Phigs, X, Dore, PEX, Mac Toolbox, Windows, OpenLook, OpenWindows, Motif, etc. ) Fortran Related Questions: -------------------------- (4.1) Which Fortran compilers do you use that DO or DO NOT allow variables longer than six characters? What are the limits? (8 characters, 31 characters, etc.) (4.2) Which compilers DO or DO NOT have NAMELIST input? (4.3) Which compilers DO or DO NOT allow "REAL*8" declarations. (4.4) Which compilers DO or DO NOT support the following programming convention? subroutine sub( a ) common /block/ inum(1000) equivalence( len, inum(293) ) dimension a(len) return end (4.5) Which compilers DO or DO NOT support the 1978 MIL-STD-1753 extensions (DO/ENDDO, DO WHILE, INCLUDE, IMPLICIT NONE, the operators IOR, IAND, NOT, IEOR, ISHFT, ISHFTC, IBITS, MVBITS, BTEST, IBSET, and IBCLR, and octal and hex constants using the notation O'nnn', Z'nnn'). (4.6) Which compilers support the Fortran 90 ALLOCATE statement? (4.7) Which compilers support POINTER declarations? (4.8) Which compilers support %VAL() and %LOC() operators? (4.9) Which compilers support Fortran 90 array operations? C Related Questions: -------------------- (5.1) Which C compilers DO or DO NOT support 1989 ANSI Standard C? (5.2) Which C compilers support #elif ? (5.3) Which C compilers convert all floating point function arguments to the "double" data type. (5.4) Which C compilers use ## for concatenation? (5.5) Which C compilers replace parameters inside strings? (5.6) Which C compilers treat as significant LESS THAN 31 characters in a name? (5.7) Which C compilers allow the keywords, void, const, volatile, and signed? (5.8) Which C compilers support function prototypes. From jkl@ccl.net Wed Sep 11 15:09:52 1991 To: chemistry@ccl.net Subject: PostScipt files produced by Molecular Modeling Systems Date: Wed, 11 Sep 91 15:06:48 EDT From: jkl@ccl.net Status: R I produced PostScript Files made by plot commands of Insight, Quanta and Sybyl (note, that I listed them alphabetically) from Biosym, Polygen and Tripos respectively. All were latest versions for Iris 4D70. I wanted to include them in the TeX paper I am writing. TeX (and all other programs which import PostScript pictures into documents) require the EPS file (Encapsulated PostScript File). None of the packages mentioned produce the EPS file. It is usually easy to produce the EPS file from any PostScript file, assuming that the PostScript files are written with basic adherence to generic PostScript rules (yes, I mean it). There are several basis rules: 1. Use Adobe-2.0 standard 2. Use Header comments (e.g. %%BoundingBox: llx llu urx ury) 3. Never ever use initxxxx commands (i.e., initgraphics, initclip, initmatrix) 4. Do not use anything from statusdict 5. Do not leave anything on the stack My preliminary analysis of PostScript files produced by all the 3 packages is that their Postscript may not make them that much proud. The Biosym postscript %%BoundingBox dimensions are just dimensions of the whole page instead of the actual picture. I have to print the picture, take the ruler and find the bounding box myself. I have to measure the binding box myself in inches and then multiply them by 72. page top ---------------- | | | | | | | urx | |<------> | | | | llx *** ^ | |<--->*** | | | *** | | | ^ ury| | |lly | | | | | | |______v_____v_| page cottom Then I have to edit the Biosym PostScript file by adding %%Boundingbox header comment and deleting initgraphics command at the top of PostScipt file. Then I have to change the beginning of the file from: %!PS-Adobe-1.0 to %!PS-Adobe-2.0 EPSF-1.2 And it works then. I think that Biosym files are the ones which you can convert to EPS file format at a resonable expense of your time. HOWEVER !!! You cannot legally use them. They contain a record %%Creator: RRT. Copyright (C) 1989 BIOSYM at the top, which says that you cannot reproduce pictures made by Biosym Software (do not laugh --- this is exactly what it says). I am not sure if BIOSYM is serious about it and waits until we publish enough pictures and then they will sue us all, or maybe they have this big red rubber stamp and they put "Copyrighted" on anything. The Postscript files produced by Sybyl by Tripos Assoc. are, to say mildly, strange. They include the header: %!PS-Adobe-2.0 EPSF-1.2 Which means: "Hi, we are Encapsulated Postscript", and the first sight shows they are not, since they: 1. do not include %%BoundingBox header comment 2. they include: initclip and initmatrix. Deleting these commands ruins your picture. Keeping them in, prevents you from scaling and positioning the picture in your document They make nice pictures, but the only way you can include them in your document is by using scissors. Moreover, they are not copyrighted. The QUANTA does not plot what is on the screen, but gives you a limitted menu of what you can print (i.e., no maps, no fancy stuff, just simple molecular images). For this reason it is of limitted utility to me. However, they are easily converted to EPS files: 1. Change first line to: %!PS-Adobe-2.0 EPSF-1.2 2. find a BoundingBox by measuring a picture as described above and put as the second line a header comment, e.g.: %%BoundingBox: 0 0 612 792 3. Delete initgraphics command at the top of the file. After doing this the pictures scale and can be moved on the document page. DOES ANYBODY KNOWS HOW TO MAKE THIS EASIER ? CAN WE GET SOME COMMITMENT FROM SOFTWARE VENDORS THAT THEY WILL MAKE OUR LIFE EASIER IN A NEAR FUTURE BY FIXING THEIR ACT WITH POSTSCRIPT ? Jan Labanowski Ohio Supercomputer Center 1224 Kinnear Rd Columbus, OH 43212-1163 From chemistry-request@ccl.net Wed Sep 11 15:17:35 1991 Date: Wed, 11 Sep 91 14:58 EDT From: "Scott Le Grand" Subject: Re: MD neighbor lists To: nicholas@mulliken.mcs.anl.gov Status: R Hi, are you going for a massively parallel setp such as on a connection machine, or just vectorizing as on a Cray? One obvious scheme for 1-2 exclusions and such is to store a bit string on each processor for each of the atoms it handles which tells it whether to include or exclude such interactions. Since you will have at most 12 such exclusions unless you go to highly coordinated systems, you could also just store the specific interactions which are that way and either check for them or do them twice, subtracting them from the sum on the second try which would have the effect of not having done them in the first place. I just use boxes for neighbor lists myself. But communicating this data could be troublesome in a massively parallel environment.... Scott From chemistry-request@ccl.net Wed Sep 11 17:00:28 1991 Date: Wed, 11 Sep 1991 13:34:43 PDT From: "W. Todd Wipke" To: chemistry@ccl.net Subject: re: PostScript files ... Status: R Jan, your analysis of what each vendor is doing wrong is an excellent first step in bringing the vendors into line. Here is a case where there is a clearly public standard, EPS, the rules are clear, yet vendors are violating the standard. By pointing this out you have done a good service to the users and to the vendors. Good job! -Todd