[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Minutes of 1999 business meeting at Glasgow

Dear Colleagues

A Happy New Year to you all.

I apologise for the delay in circulating the formal report of the COMCIFS
business meetings at the IUCr Congress in Glasgow last August. Many of
you will be aware that the IUCr launched its own electronic journals at
that time, and developing the technical aspects of that has left me short of
time for other business recently! Many of the actions noted in the minutes
have already been undertaken, though others still require attention. I propose
to follow this email with a brief series summarising how far each action has
progressed, to my knowledge. These messages can act as placeholders in the
list archive for specific discussions.

Best wishes
Brian

==============================================================================

Minutes of a business meeting of COMCIFS, IUCr Congress, 5 and 9
August 1999, Scottish Exhibition and Convention Centre, Glasgow

Present:
I. D. Brown (Chair)
B. McMahon (secretary)
S. R. Hall
H. M. Berman (part)
J. D. Westbrook
H. J. Bernstein
P. E. Bourne
T. Proffen
P. R. Edgington
L. Pytlik

1. Merging dictionaries

The report of the technical working group on dictionary maintenance was
formally presented by Brian McMahon. No technical objections were
raised. Syd Hall stressed the importance of developing new dictionaries in a 
manner that required no changes to existing publicly distributed
dictionaries.

Action: The report to be circulated by email in final form for COMCIFS
approval.


2. Intellectual property rights

Issues of the Union's safeguarding its intellectual property rights to
CIF through patent and copyright claims had been raised on the PDB mailing
list, where it was suggested that such claims tended to deter software
developers who wished to operate within an open software development and
exchange model. Other anecdotal evidence suggested that this remains a
fairly widespread concern. Herbert Bernstein, Brian McMahon, Syd Hall and
the IUCr Executive Secretary Mike Dacombe had undertaken a review of
existing policy and had proposed modifications. The proposals amounted to
a rewording of the published policy that attempted to clarify the Union's
desire to encourage the dissemination of CIF-supporting software tools;
and the registration of CIF, mmCIF, DDL and other acronyms as trade or
service marks to identify and certify standard-compliant files.

The registration of trade marks was widely acclaimed as a positive
step. However, it was generally felt that the tone of the proposed revisions 
to the policy statement was still too aggressive. Herbert Bernstein
emphasised the vulnerability of software systems and products that were not
properly protected by legal instrument if they were acquired by predatory
large software businesses. Nevertheless, the consensus was in favour of the
Union's adopting far less emphasis on protective measures, and far more on
messages calculated to win the support of utility developers.

Actions: (1) The Executive Secretary to proceed with registration of
appropriate trade and service marks under suitable jurisdictions to
establish a legally defensible framework for identification and
certification of CIF-compliant entities.

(2) Further review of the official policy of the Union with respect to
intellectual property rights. Herbert Bernstein and Phil Bourne agreed
to lead a discussion including the wider community of users on the most
appropriate form for expressing the IUCr's policy on property rights.

3. CIF browser-editor

There was continuing demand for a user-friendly CIF editor and/or browser
to facilitate the use of CIFs by non-specialists. The need was especially
pressing in the field of small molecules, for submission to Acta Cryst. and
other journals and databases. Some tools already exist, such as the CCDC
graphical HICCuP validator/editor and the NDB mmCIF input tool; but they
fail to meet the requirements of portable software that is easily installed
across multiple platforms, and that matches the user interface anticipated
by authors used to modern word-processing software. 

There was some discussion as to the proper role of COMCIFS in commissioning
specific software products. It was felt that this was in certain cases a
legitimate role, for which COMCIFS could usefully bid for central IUCr
funding. In other respects, however, software development was a
responsibility of the wider community, and this was also an area where the
role of COMCIFS needed to be developed.

Action: John Westbrook and Paul Edgington to collate experiences of NDB and
CCDC work on CIF editors and establish any basis for onward development of
these or derivative tools under the aegis of COMCIFS.


4. Encoding and dissemination

Not discussed.


5. CIF and non-crystallographic data projects

Helen Berman reported on the existence of other non-crystallographic
database projects that could usefully interchange data with CIF. A protein
structure models database under the aegis of Glaxo and the Swiss
Bioinformatics Institute used a data dictionary in mmCIF format; and a
molecular recognition database of binding constants with spectroscopic and
calorimetric data was being developed. However, the nmrIF project
(originating at BioMagResBank) used an extended STAR format that was not
immediately compatible with mmCIF, and a database of NMR RNA secondary
structure had chosen XML as a data representation format.

While it was clearly desirable that non-crystallographic data projects
should be encouraged to use a data representation that was as fully
CIF-compatible as possible, the issues raised were how to ensure such
compatibility; and more widely, how to encourage CIF as a data mechanism of
choice within other communities.


6. Relationship to other file structures

At a time when there is a proliferation of data exchange mechanisms
operating at all levels from network transport (e.g. ASN.1) through 
semantic content tagging (e.g. XML, CML) it is clear that (i) there is
unlikely to be a single such mechanism serving all purposes; (ii) the
mechanisms that win widest support are those for which a wide range of
reliable software tools exists; (iii) alternative mechanisms may survive so
long as they can be cleanly and reliably translated into other formats.

CIF is particularly strong in terms of its content; but rather weak in terms
of its supporting software infrastructure. It was suggested that effort
should be expended in promoting CIF as a concrete exemplar for RDF metadata
groups (RDF, the resource description framework, is an initiative for
describing metadata of the WorldWide Web Consortium W3C); and, more
generally, in establishing a presence within relevant data representation
forums. John Westbrook has already taken the initiative here.

Related proposals included a technical specification of an interface with
XML, the emerging standard for web-based information transfer; and an
investigation of the process required to make CIF an international standard
as supported by the ISO.

Action: Herbert Bernstein and Syd Hall to investigate the process of ISO
standard development.


7. The future of DDL

Structural problems remain in maintaining CIF dictionaries under different
DDL formalisms. During the CIF open meeting, Syd Hall demonstrated a
dictionary extension language (DEL) developed with a modified DDL formalism
("DDL3") to codify and evaluate methods of inter-relating data items. The
demonstration was impressive, but raised the question of how best to take
forward such developments without fragmentation of the existing dictionaries 
and data files. Syd and co-workers in Perth are rapidly developing and
prototyping DEL applications.

Syd and John Westbrook were charged with the task of coordinating
developments in DDL and DEL languages.


8. Other business

Throughout the discussion of items on the formal agenda, there was debate on
the nature, role and organization of COMCIFS itself. It was felt that there
was a real danger of a loss of contact between the committee and the wider
community; and that, indeed, there existed a range of communities with whom
COMCIFS needed to engage.

Arising from the discussions came proposals for further reorganisation. A
number of subcommittees were proposed, with specific charges for technical and 
non-technical matters. 

(1) A DDL Development subcommittee, responsible for the orderly development
    of dictionary definition language formalisms upwardly compatible with 
    existing DDL1 and DDL2 implementations.

(2) A publicity and outreach subcommittee, responsible for raising the
    visibility of CIF as a data exchange mechanism, both within the
    community of crystallographers as a whole, and within the community
    of software developers.

    Among projects that could be undertaken by such a committee would be
    production and distribution of a newsletter, establishment of
    special-interest mailing lists, promotion of CIF software workshops
    alongside regional crystallographic meetings or computing schools,
    representation on other data standards committees.

    Thomas Proffen offered to look into organising a web workshop for new
    users of CIF.

(3) A software development subcommittee, responsible for commissioning
    and coordinating the development of software applications and libraries
    necessary for increasing the use of CIF.

(4) A dictionary review subcommittee, responsible for the content review
    of dictionaries submitted to COMCIFS for approval. This would be the
    body to which dictionary management groups would submit modified
    dictionaries for detailed review. 

Action: The Chair of COMCIFS to establish terms of reference for these
subcommittees and to identify and invite participation from the current
COMCIFS membership.

It was suggested that the scope of COMCIFS activity was becoming sufficiently
wide that the IUCr might in future considering reforming it as a full
commission, either as a new Information representation Commission, or as a
reinstatement of the Crystallographic Data Commission.

Brian McMahon
15 January 2000
==============================================================================

Reply to: [list | sender only]