\def\Ang{\thinspace\AA} \def\Angsq{\Ang$^2$} \def\percent{\thinspace\%} % \magnification = \magstep1 \font\bfbigten=cmbx10 scaled\magstep1 \font\bfbignine=cmbx9 scaled\magstep1 \font\itbigten=cmbxti10 scaled\magstep1 \font\scbig=cmbx7 scaled\magstep1 \font\sc=cmr7 \font\scl=cmr7 scaled\magstep1 \hoffset .46 true cm \voffset .46 true cm \hsize 16.0 true cm \vsize 23.0 true cm % % \baselineskip=12pt \normalbaselineskip=12pt % \outer\def\section#1\par{\vskip0pt plus.3\vsize\penalty-250 \vskip0pt plus-.3\vsize\bigskip\vskip\parskip \message{#1}\centerline{\bf#1}\nobreak\smallskip} % \outer\def\subsection #1 #2\par{\medskip\centerline{#1 \it#2}\nobreak\smallskip} % \outer\def\subsubsection #1 #2\par{\medskip\leftline{#1 \it#2}\nobreak\smallskip} %\raggedright \centerline{\bf MINUTES OF A MEETING OF WORKING GROUP 2 OF CCP4} \centerline{\bf WHICH WAS HELD AT BIRKBECK ON 10th January 1990} \bigskip \noindent Present: \medskip $$\vcenter{ \halign{#\hfil&\qquad#\hfil\cr David Rice (Chairman)&Sheffield University\cr Don Akrigg &Leeds University\cr Peter Brick &Imperial College\cr Peter Daly &Daresbury laboratory\cr Eleanor Dodson &York University\cr Phil Evans &MRC, Cambridge\cr Paul Mallinson &University of Glasgow\cr David Moss &Birkbeck College\cr Jim Raftery &University of Manchester\cr Paul Taylor &University of Edinborough\cr Francesca Young &University of Glasgow\cr Kim Henrick (Secretary)&Daresbury Laboratory\cr }}$$ \bigskip (Apologises from Ian Tickle and Simon Phillips) \bigskip \bigskip {\it 1. The minutes of the meeting of 4th October 1989 were agreed.} \medskip {\it 2. Matters Arising} \bigskip {\it 3. Report on the 1990 Study Weekend.} \medskip David Moss presented to the meeting the final timetable for the Study weekend. David then explained to the meeting that there had been an objection to David Stout's name being on the programme. An argument had been presented to the Study weekend organisers that David Stout had not realised the error in the structure of Azotobacter ferredoxin, until the re-determination of this structure by the Stout and Jensen group in Seattle. The members of the meeting however felt that as David Stout had made the error and was willing to present his results and correction to an audience aware of the background, then he was the correct choice in speaker. The second objection was to the lack of a speaker from a spectroscopic field, as it was said to the organisers that EXAFS had provided unambigous evidence that the Fe--S cluster arrangement proposed by D. Stout could not be correct. Phil Evans indicated that as the coordinates for D. Stout's structure determination had been deposited and that crystallographers had displayed these coordinates and were well aware that the structure had at best ``unusual geometry''. David Moss and Ian Tickle were sympathetic to the view that structure work could be balanced with the results in spectroscophy, particularly for metalloproteins, and had invited a speaker from the Seattle group. Professor Ellinor Adman had agreed to come to the meeting and give a presentation on ``A Tale of Four Iron--Sulphur Protein Structures: Sequence Errors and Other matters''. This had resulted in some modifications to the time table. The secretary then informed the meeting that based on the number of student participants from previous years study weekends, that CCP4 funds should be available to subsidise both the fee and accommodation for post graduate delegates. The secretary also noted that of the seventeen invited speakers, only seven had at this stage had confirmed their invitation. A question was also raised about the low quality of the close circuit television sound, for the coffee lounge area, and asked if this could be improved. [ NOTE: This matter was raised with the appropriate staff at Daresbury, and the secretary was informed that they were well aware of the problem and an improvement plan had been submitted, but they had not as yet received approval, and therefore there would be no change with the T.V. as to the 1989 Study Weekend.] \bigskip {\it 4. Report on the current finances of CCP4} \medskip The secretary then indicated to the meeting that CCP4 has funds of around 30,000 pounds, spread over three project numbers, with about 4,500 pounds in a suspense account held by S. Lowndes, 4,000 pounds in the income account (which can be carried forward to the next financial year without difficulty) and about 22,000 pounds in the CCP4 running costs project number. The latter includes funds carried over from the 1988--1989 financial year. The secretary also informed the meeting that the 10,000 pounds to be used to subsidise the 1990 Study Weekend, as agreed at the 4th October 1989 working group 2 meeting, would be paid immediately into S. Lowndes suspense account. This would serve two purposes, (i) to show that the normal 8,000 pounds travel funds had been spent and (ii) as a safety measure if there was problems in carrying forward the income monies from previous years. The secreatary also asked the meeting for approval in buying a stock of DC600XTD Quarter Inch Cartridge tapes (QIC-150), as these items (ca 33 ponds each) are the near universal unix tape format and were not a stock item at Daresbury, unlike the normal 1/2 inch reel to reel tapes. For the distribution of the Unix CCP4 implementation there is an increasing call for this type of media. For non-UK distribution these tapes would be charged for, including a handling charge. There followed a discussion on what CCP4 could spend its income on that would be of benefit to the UK users. Phil Evans suggested that as there was to be a joint ESF/CCP4 project to re-code the CCP4 programs to the new style LCF routines, to be held in Cambridge (possible date: 26-February to 2-March), that CCP4 funds could be used for the travel and accommodation costs for the 6 UK members at least and possibly for one or more of the ESF contributors. The Chairman agreed that this would be a reasonable request and could be met from CCP4 funds. The secretary suggested to the meeting that CCP4 could contribute to the costs of software tools to aid in making the fortran programs more portable. As discussed below in item 8, there were unexpected difficulties with code running on MIP's machines, even after the efforts to use Toolpack. Two commercial packages are available and have been recommended to Daresbury are SPAG and FORCHECK, from Polyhedron Software. A price of around 1600 pounds was suggested to us from David Wild. These software aids are basically used to try to get a handle on the awful code that ``someone else'' has produced. They together re-structure fortran, removing depricated Fortran code as per the preliminary Fortran 8X standard, and in addition indicate how to change all non standard fortran 77 into portable source code. Daresbury is currently investigating a range of CASE tools for fortran, and the secretary has indicated that CCP4 may contribute to the costs of SPAG and FORCHECK if these were made available on the Convex to UK users. Daresbury is actually looking into packages for the formal design of packages, however CCP4 may have little interest in this area. There is also a package available on the Daresbury Sun's that uses an interactive X-windows system to design a window environment for an applications program and ``writes'' the X-lib code for the final designed X-window system. In relation to having standard code, Fran Young, also indicated that she was having difficulties porting the existing Unix ccp4 suite onto an IBM system. Other suggestions were that future Study Weekends could afford to allow for more overseas speakers, or perhaps sponser a lecture tour. Support for an accessible standardisation package came from many members of the meeting and the Chairman expressed the view that a statement was required from working group 1 to clarify on what exactly the secretary to CCP4 could spend money on. [Post Meeting Note: Peter Daly has since contacted Polyhedron Software and found that the prices for a small mini (e.g. Sun 4, MicroVAX II) are 2390 pounds - SPAG and 2390 pounds for FORCHECK, while for a large mini ( VAX 8xxx) these prices rise to 3990 and 3990 pounds respectively. A Convex version is available for 6990 pounds (SPAG) and 4990 pounds (FORCHECK), and there is a 40\% discount for educational institutes. Peter has also received an IBM PC test disk which he is investigating. The programs themselves, SPAG and FORCHECK have now been ordered for the Convex (from central funds) and will be available to all convex users within a months time.] \bigskip {\it 5. CCP4 Secretary Resignation and Replacement. } \medskip The secretary informed the meeting of his resignation from the CCP4 post as of the 1-March-1990. There followed a discussion on the type of individual required to fill the post. The meeting asked that in the job description that the development, maintenance and distribution of the CCP4 programme suite should be prominent comments. The committee also suggested that a Working Group 2 representative could be present on the interview panel. The secretary has already agreed to Colin Nave's (px project team leader at Daresbury) request that he also be a member of the interview panel. Members of the working group were asked to make information about this permanent post in protein crystallography available within their respective groups. \bigskip {\it 6. Report on work directed towards a portable graphics package.} \medskip Peter Daly presented to the meeting the plan to rewrite PLOT84 into standard Fortran 77 allowing for GKS lookalike subroutine calls, such that any program using PLOT90, could then be linked with either a GKS (Uniras-GKS or VMS-GKS or Sun-GKS) or use Xgks (an X-windows set of routines written in C that mimic GKS), or with PLOT90. Any site choosing to use PLOT90 would then use PLOT90's in-house meta file, or use the drivers built into PLOT90. The actual form of the meta file was as yet undecided, with either a 2 point file including a header or a 3 point file without a header with the header information included as a type of dummy data record. There followed a discussion about the merits of GKS (with both Leeds, Imperial and Edinborough currently using GKS in some form). Peter Brick introduced the idea of using the ``standard'' CGM meta file format as a possible file format for the proposed PLOT90. Much of the discussion appeared to misunderstand what Peter Daly was proposing, not to standardise on GKS itself, but on a PLOT84 made to look like GKS, with the addition of using existing higher level routines which would also contain GKS like calls. The new PLOT90 was also planned to incorporate the work of Steve Kinder at Daresbury in designing a portable X-windows image processing package. Steve Kinders work was aimed at a general package that could be incorporated into MOSFLM, Laue software and both Fast Images display and in Image plate work. Steve's work is freely available and had been passed onto David Wild (EMBL - Heidelberg) and James Pflugrath (Cold Spring Harbor). Phil Evans suggested FIGS-3D as used in the new FRODO program called ``O''. The Chairman concluded that Peter Daly had carte blanche to produce a replacement for PLOT84 as a portable system would get used. \bigskip {\it 7. Report on the ESF - EACBM meeting of 10th Dec 1989.} \medskip Phil Evans gave a brief summary of the Hamburg meeting and indicated that the new LCF routines were available (with the exception of multi-record LCF routines) and had been tested at Hamburg, York (with VMS) and Daresbury (under Unix). This work was almost completely that of Sandra McLaughlan and Howard Terry at Hamburg. The system was based on Jan Zelinka's header system. The header system uses a set of Key\_Words that will be standardised on the same keywords for standard input and the parser then can be switched from either the header or to an input stream. The system allows for any number of future items to be added to the header(s). Peter Brick stressed the need for a set of complete documentation as the existing LCF system did not clearly define some aspects, for example how S or FOM were stored and how a column could be flaged if there was no data for that column. A complete description of the column types and a standard set of column program labels was also required. Phil Evans then expanded on the work-in to take place around the end of February in Cambridge for the rewriting of the programs to switch over to the new LCF. [The secretary is currently arranging with Hamburg a check list to be carried out for each program and a list of the programs to be converted.] \bigskip {\it 8. Report on the replacement of DLVD by a multiuser Unix system.} \medskip The secretary briefly informed the meeting on the tender exercise carried out Daresbury on the replacement of DLVD (vms) by a unix machine. The planned replacement of DLVD would still give CCP4 access to VMS via DLVA's replacement, a MicroVAX 3800, which has arrived in Daresbury, and disc space would be made available on this machine. Non-Daresbury staff however may not have access to this machine, although this has not been fully decided as DLVA was a general purpose machine. Due to the interest expressed in this exercise a summary of the Tender, the Tender evaluation, and the time tests have been included as an appendix to these minutes. \bigskip {\it 9. Report on the Joint CCP4/ESF Newsletter.} \medskip The secretary briefly indicated to the meeting that the newsletter had been ready for printing on time (15th December) however Hamburg required a logo attached and in the mean time more articles had arrived. The newsletter is currently now printing. \bigskip {\it 10. Group Reports.} \medskip Eleanor Dodson reported on the Multiwave Anomalous Dispersion workshop held in Hamburg (on the 11th December) and expressed her delight in Wayne Hendrickson's talk and following this had obtained copys of his scaling programs and was in the process of incorporating these into CCP4. David Rice reported on a disc that was proving highly successful in Sheffield. This was a HP 680Mb (formatted) disc with an ESDI controller that can be attached to a MicroVAX Q bus as a RA device (DUA). These devices cost around the 2,000 pound mark. Fran Young reported that IBM now have a set of hardware FFT routines that they would like to incorporate into an IBM implimentation for CCP4. Peter Brick indicated that unless the transforms were done in memory such a set of routines may still be slow with the current amount of disc traffic required for CCP4's fft routines. Paul Taylor briefly summarised an Edinborough proposal to the SERC which was seeking CCP4 support. This was for a central resource for dynamics calculations using a large Meiko transputer machine. Don Akrigg also indicated that this was hoped to be coupled to the new proposed ``super Janet''. This was not greeted with much enthusiasm from the committee members. \bigskip {\it 11. Any other business.} \medskip Phil Evans suggested reviving the practice of having an ESF representative at CCP4 working Group 2 meetings. The Chairman agreed that this was highly appropriate and there was a discussion on who could be invited to attend but no conclusion was reached. \bigskip {\it 12. Date of the next meeting} \bigskip {\scl A date for the next meeting was set for Wednesday the 25th of April 1990, at York University Department of Chemistry for 11-30 am.} \bigskip \medskip \centerline {K. Henrick} \centerline{(secretary)} \bye