IUCr 1996 Triennial Report - Report of Committee on Electronic Publishing, Dissemination and Storage of Information

Members: F. H. Allen, A. Authier, I. D. Brown, C. E. Bugg, P. W. Codding, E. Dodson, Y. Epelboin, H. D. Flack, A. M. Glazer, J. P. Glusker, S. R. Hall, S. S. Hasnain, J. R. Helliwell, H. Kamitsubo, E. N. Maslen (Chairman), B. McMahon, P. R. Strickland, M. Teeter.

The Committee has advised the IUCr Executive Committee on electronic submission of information, on data delivery and on other networked services. It generally agreed on principles, but conducted a lively Internet debate on how the objectives could best be achieved. The Committee met in Chester (1994) and in Montreal (1995). The Executive Committee approved the appointment of a consultant to advise on electronic publishing, and the report is being studied closely. A summary of the Committee's deliberations follows.

Electronic submission of information

Submission is linked so closely to utilization and storage that it proved desirable to consider all aspects together. The extension of CIF from small molecules to powders to proteins made it appropriate to allocate each task to specialist groups that work and report separately. The crystallographic community has been well served by their response to these demanding tasks.

During the triennium, Acta Crystallographica Section C standardized on electronic submission for manuscript submission. To reduce publication times and achieve economies, manuscripts should be submitted in word processor format to other IUCr journals and transmitted to and from referees electronically. It is impractical to implement total electronic submission to all journals simultaneously. Changes are to be made on a timescale that optimizes the benefits.

Electronic submission is favoured by the emergence of SGML as a de facto publishing industry standard. This does not define publishing procedures fully, because SGML permits adaptation by users. While the broad thrust of future developments is clear, aspects such as the range of manuscripts for which CIF to SGML translation will be preferred have yet to be determined.

Delivery

The IUCr's pilot electronic delivery of the World Directory of Crystallographers provided a network-accessible service, along with valuable experience. Originally, the Directory was to be based on STAR tools, no alternatives being available at that time. Considerable effort was expended in developing those tools before the World-Wide Web (WWW) protocol emerged. Since WWW links were installed, the electronic Directory has been used mainly in that mode. Most users prefer standard protocols even when technical performance favours alternatives.

Further developments in electronic delivery are planned. Scientific importance would assign top priority to Acta Crystallographica Section D, but there are technical and financial counter-arguments. Protein crystallography depends on high-quality diagrams, for which the low information density of raster graphics is inefficient. For many applications, raster graphics' generality is offset by difficulty in synthesizing information - at which current computing technology competes unfavourably with more versatile human observers. High loads imposed by inefficient graphics protocols are tolerable on lightly used networks served by powerful hardware, but performance deteriorates rapidly as network loadings increase.

Most information seeking on networks is interactive. An acceptable interactive service within a single land mass spanned by a strong network is readily provided, but one cannot guarantee satisfactory response that depends on inter-continental communication. Ideally, the IUCr should provide a comparable service to all of its members, but the service at many locations is limited by communications. Lack of information security on networks is a further but less acute problem. One must ensure that IUCr publications accessible on-line are not copied and re-sold.

Services that the IUCr could provide electronically include some ancillaries to journals. Currently (i) titles, (ii) titles and abstracts, and (iii) full articles on request (in appropriate formats) could be provided with at worst modest delays that would be tolerable for a free or low-cost service. There is no danger in making option (i) available as publicly accessible information. Options (ii) and (iii) could complement normal journal subscriptions. Option (iii) could be provided on a charge per page basis. It is appropriate to provide these for Acta Crystallographica Section C initially, extending this to other journals if the Acta Crystallographica Section C trial proves successful. The WWW could provide cost-effective communication between IUCr members and its publishing activities. Publishing the IUCr Newsletter on the WWW could expedite delivery and cut production costs. That would be relatively straightforward as neither full cost recovery nor high-quality graphical information is needed. Timing of full implementation requires care, because of the need to avoid adverse impact on scholarly publications by the Chester office.

Other networked services

B. McMahon's World-Wide-Web-based information server, an information interface on the e-mail network for the crystallographic community, is suitable both for structured documents and for informal publications. It has a Gopher option, allows transmission of CIFs and provides access to the World Directory of Crystallographers. It has the potential to provide information on journal papers.

It costs approximately GBP 50 per compound to check structural information prior to publication in Union journals. The amount of incorrect structural information presented by those without crystallographic training is significant. Problems are compounded by the number of crystallographic studies now submitted by research workers with limited training.

The high cost of checking makes it difficult for crystallographic databases to provide an affordable service. The CIF submissions to Acta Crystallographica Section C indicate that automatically generating crystallographic information via CIFs using the major crystallographic computing packages will reduce errors dramatically. Such CIFs are also well suited to standardized checking.

The IUCr is the largest single publisher of structural data, but the total published by others is far greater. Other journals' norms are well below IUCr standards, and it is difficult for those journals to achieve higher reliability in their crystallographic publications. Automatic checking of CIFs accessible through the WWW information server is an important service that does not impact strongly on the IUCr's finances. It has potential to reduce the cost of the crystallographic databases, benefitting all crystallographers. From a narrow perspective, it might seem opportune to make this self-supporting, but that is complicated by the limited success rate for automatic checking. A small but significant fraction of the errors in structural data that are not identified readily by automatic procedures inhibits such charging. It is appropriately provided free, because so many crystallographers benefit. Eventually, all structure solving packages should check structures automatically, so that fewer errors survive.

Electronic publishing of reference data, affecting mainly the International Tables for Crystallography, requires careful consideration. The Tables are widely used but the need for detailed and up-to-date information is not large enough to make crystallographic reference data attractive for electronic publishing. However, viewing electronic delivery in isolation from the printed publications over-simplifies the situation. An up-to-date archive that is accessible electronically could simplify the preparation of new editions of International Tables. A steady increase in the use of electronically accessible tables could be expected.

CD ROM back-up of Union journals was addressed by the consultant. A possible scenario is for abstracts to be delivered electronically to subscribers, who can optionally extract the desired full papers and/or deposited material from a central archive. Back-up copies of journal articles would be supplied on a CD ROM annually. This assumes that the subscriber is equipped with the software needed to perform text typesetting, and has access to a CD reader.

26 February 1996E. N. MASLEN, Chairman

Report by Managing Editor

SGML implementation in IUCr journals

The Editorial Office is currently undertaking a project to implement standard generalized mark-up language (SGML) for the production of IUCr journals.

An SGML document-type definition (DTD) for journal articles is being developed in conjunction with an outside consultant. The DTD will conform to publishing industry standards and to ISO 12083, the standard for the publication of scientific papers.

Templates allowing authors to submit papers using standard word-processing packages (e.g. LaTeX) are to be written, as are programs to translate from these formats to SGML. A number of SGML editing and translation tools are currently being assessed for the in-house editing of papers.

The implementation of SGML will be linked to the introduction of a new production database and database management system. It is envisaged that this will improve office efficiency, facilitate communications with editors, authors and referees, and lead to shorter publication times.

The SGML files will not only allow increased automation in the preparation of the conventional printed issues of the journals, but will also allow the production of electronic versions of the journals.

The existing CIF environment for structural papers is being maintained, and we are investigating the best way to integrate both production strands.

21 May 1996 P. R. STRICKLAND, Managing Editor

[Next section] [Previous section] [Index] in IUCr 1999 Triennial Report
IUCr Triennial Report: [1993] Report of Director of Archiving and Crystallographic Information. [1999] Report of Committee on Electronic Publishing, Dissemination and Storage of Information.

Copyright © 1997 International Union of Crystallography