[Date Prev][Date Next][Date Index]
(70) Future directions for COMCIFS: a discussion paper
- To: COMCIFS@iucr.ac.uk
- Subject: (70) Future directions for COMCIFS: a discussion paper
- From: bm
- Date: Mon, 9 Jun 1997 14:47:04 +0100
Dear Colleagues
We have been involved for some time in the detailed technical evaluation of
the powder and mmCIF dictionaries, and there have been a number of other
subjects that have drifted to the sidelines. Among these I should note:
revision of the core dictionary to align it with the new datanames adopted
by Acta C in the 1997 Notes for Authors; progress on the modulated
structures, electron density and symmetry proposals; progress with the
imageNCIF binary format proposal; a proposed diffuse scattering CIF. I
intend to address these items in the near future. However, as I mentioned
last time, the conclusion of this phase of our activities marks a good time
to take stock and review the nature and role of our committee. I therefore
circulate the discussion paper David Brown prepared (a little while ago),
in the hope that it will promote lively but considered debate, and perhaps
some definite proposals for our future evolution, that can be placed before
the Executive Committee during their summer meeting this year.
Regards
Brian
D70.1 Review of COMCIFS
-----------------------
To Members of Comcifs.
Now that we are approaching the end of the first phase of Comcifs'
operation, I have been asked to present a formal report of our work to the
Executive. This gives us an opportunity to assess how well Comcifs has
worked in the past and to see what changes are needed in our membership
and mode of operation to meet the changed requirements of the future.
Below I have prepared a position paper that points out what these changes
may be and asks questions about the best way we can prepare to meet them.
I hope it will stimulate a discussion that will lead to a consensus which
I can report to the Executive. Please send your comments to Brian McMahon
as usual.
David Brown
Comcifs Chair
----------------------------------------------------------------------
THE FUTURE ROLE OF COMCIFS
A DISCUSSION PAPER
by DAVID BROWN
1997.3.12
1. INTRODUCTION
The Crystallographic Information File (cif) uses the STAR file
structure (Hall, 1991, J. Chem. Inf. Comp. Sci 31, 326-333). STAR is
patented by the International Union of Crystallography (IUCr) and the cif
dictionaries that implement the cif standard are copyright by the IUCr.
The purpose of the patent and copyright is to ensure that ownership and
the rights to maintain the standard are retained by the IUCr. The
Executive Committee of the IUCr has appointed the sub-committee COMCIFS to
approve and maintain the cif dictionaries.
The first cif dictionary, the core version 1, cifdic.c91, was
prepared by an ad hoc committee and was approved by the Executive in 1990
(Acta Cryst. (1991) A47, 655-685). Subsequently further ad hoc committees
were established to prepare dictionaries for crystallographic powder data
and macromolecular structures.
In August 1992 the Executive of the IUCr decided to appoint a sub-
committee to be known as COMCIFS (Committee for the Maintenance of the CIF
Standard) to take on the responsibility for maintaining the cif standard.
The first committee was appointed early in 1993 and Comcifs held its first
and only meeting at the IUCr Congress in Beijing in August of that year.
By the time that Comcifs was appointed, a number of important
decisions has already been taken. Cifdic.c91 (core version 1.0) had been
approved, work was well underway on the powder and macromolecular
dictionaries, and the concept of storing the cif dictionaries themselves
as STAR files had been accepted. This required the construction of a
dictionary definition language (DDL) but the maintenance of the DDL
dictionary was not part of Comcifs responsibility. There are currently
two DDLs. Most cif dictionaries use DDL1 but the macromolecular
dictionary uses DDL2 which requires the cifs to be more structured
resulting in less flexibility.
2. TERMS OF REFERENCE
The terms of reference of Comcifs as stated in the original letters
of appointment were:
"1. To receive and examine all proposed amendments and extensions to
the cif standard. In particular to ratify proposed data names,
ensure that data structures conform to the basic cif specifications,
and to ensure that the definitions of data items are clear and
unambiguous.
"2. To ensure that adequate current documentation of the cif
standard is available to potential users in both printed and
machine-readable forms.
"3. To ensure that all accepted amendments and extensions to the cif
standard are announced in an appropriate publication or news item in
an IUCr publication.
"Further, Comcifs is to be a stand alone committee of the IUCr. It is not
associated directly with any of the Commissions or Working Groups that
were involved in the origination of cif. The Executive requested that
membership of Comcifs be kept as small as possible and be restricted to
those with acknowledged technical expertise in the areas of 1-3 above.
Initial membership was by invitation of the chairman (at that time Frank
Allen), with further co-option and replacement being the responsibility of
the Committee itself with IUCr approval. Comcifs reports directly to the
IUCr Executive Committee through the Executive Secretary."
I take it that this gives Comcifs considerable responsibility for
its own management, though its approval of new dictionaries should be
reported to the Executive of the IUCr for information and its membership
nominations are to be submitted to the Executive for approval. In the
past we have not exercised this authority; we have requested the Executive
to approve new dictionaries and looked to the Executive to appoint new
members.
3. HISTORY
Since its founding, Comcifs has been working to bring three
dictionaries to approval. In December 1996 the Core version 2.0
(cifdic.c96) was approved and the first versions of the powder dictionary
(cifdic.p97) and the macromolecular dictionary (cifdic.m96) are in the
final stages of approval. In addition work is in progress for
dictionaries covering electron densities, modulated structures and
symmetry, and technical discussions are underway on how best to represent
images and how to combine binary data with cifs. Because three major
dictionaries are now essentially complete, the nature of Comcifs business
will change. It is therefore appropriate to review our structure and mode
of operation.
4. CURRENT MEMBERSHIP
The current membership can conveniently be classified into four
categories. The present membership is:
Full Members
------------
I. D. Brown (Chairman), McMaster University (idbrown@mcmaster.ca)
S. R. Hall, Univ. Western Australia (syd@crystal.uwa.edu.au)
P. R. Edgington, CCDC (edgington@chemcrys.cam.ac.uk)
P. M. D. Fitzgerald, Merck Research Laboratories
(paula_fitzgerald@merck.com)
B. H. Toby, Reactor Radiation Division, NIST (Brian.Toby@NIST.GOV)
G. Madariaga, Univ. Pais Vasco, Bilbao (wmpmameg@lg.ehu.es)
M. Spackman, Univ. New England (mspackman@metz.une.edu.au)
B. McMahon (Coordinating Secretary), IUCr (bm@iucr.ac.uk)
Observers
---------
E.N.Baker (ebaker@cb-burns.massey.ac.nz) IUCr President
H.Berman (berman@dnarna.rutgers.edu) Chair, Database Committee
A. Authier (authier@lmcp.jussieu.fr) (Commission on Crystallographic
Nomenclature)
H.Flack <Howard.Flack@cryst.unige.ch> Chair of EPC
Consultants
-----------
G. M. Sheldrick (gsheldr@ibm.gwdg.de)
P. Murray-Rust (pmr1716@ggr.co.uk)
P. E. Bourne (bourne@sdsc.edu)
E. Ulrich (elu@nmrfam.wisc.edu) (NMR Databank)
O. Ritter (o.ritter@bnl.gov) (Protein Data Bank)
Auditors
--------
N.Spadaccini (nick@cs.uwa.edu.au) STAR/DDL1
J.Westbrook (jwest@rutchem.rutgers.edu) DDL2
A.Hammersley (hammersl@esrf.fr) Binary image format
T.Proffen (proffen@rsc.anu.edu.au) Diffuse scattering CIF
H.Bernstein (yaya@bernstein-plus-sons.com)
Full members, observers, consultants and auditors all receive the
mailings that constitute the business of the Comcifs (see below). Full
members are required to vote on matters put to a vote, and must therefore
accept responsibility for the decisions made by Comcifs. This implies a
commitment of time. Observers monitor the activities of Comcifs and their
relevance to the particular bodies they represent. Consultants usually
represent a particular constituency and are expected to take part in
discussions relevant to that constituency. They may also comment on other
matters, but are free to ignore items that lie outside their constituency.
Auditors, who are not strictly members of Comcifs, have no responsibilities
but may contribute to any of the discussions. Anyone may ask to become
an auditor.
5. CURRENT MODE OF OPERATION
Comcifs does not hold formal meetings. All business is conducted by
email. Comments are sent to the Secretary (Brian McMahon) who includes
them in the circulars that constitute the record of Comcifs' business.
These circulars, which are sent to full members, observers, consultants and
auditors, raise matters for discussion and include any comments received by
the Secretary. Usually matters are resolved by consensus, but where this
cannot be achieved, a formal vote is called. Only the full members have
voting privileges. About 15 circulars are issued each year.
Comcifs is not directly involved in the preparation of new
dictionaries. This is carried out by working parties, of whom at least
one member is usually a full member of Comcifs. The working parties are
encouraged to consult widely in the field which can be done inter alia
through electronic discussion groups, workshops, conferences or by posting
the current draft of the dictionary on the web. When a working group has
completed a draft it is referred to Comcifs who may return it with
comments for further work, or approve it provisionally for more widespread
circulation and comment. The final draft is then presented to Comcifs for
final approval. Only full members may vote and all full members are
required to vote on the adoption of new dictionaries. In practice, the
provisional approval step does not appear to be necessary. If the working
group has done its job properly, the final draft should already be
acceptable to the people working in the field.
6. FUTURE ROLE OF COMCIFS
In its first four years, Comcifs has been concerned with
establishing general policies for cifs in response to the questions raised
by the working groups. Much of its energies have been devoted to getting
the first versions of the powder and macromolecular dictionaries approved.
In addition it has given advice to working groups that are starting the
process of drafting other dictionaries.
In the future, Comcifs work is likely to change. More effort will
be spent in maintaining existing dictionaries in response to a much
increased user community. As more software is written to read, write and
manipulate cifs we can expect an increasing number of requests for changes
and additions to the dictionaries and we will need to find ways of
addressing these requests in a timely manner. Changes to the dictionaries
will likely be made in a piecemeal fashion, and Comcifs will have to
ensure that these changes conform to existing rules and practice.
The core dictionary was developed with the idea that cif would
provide means for the whole crystallographic community to transfer and
archive information, but its application to the publication of Acta
Crystallographica has given Comcifs experience in working with a client
whose operations are dependent on the cif standard. When the powder and
macromolecular dictionaries are approved, Comcifs will need to work with
other major clients such as the International Center for Diffraction Data
(ICDD), the Protein Data Base (PDB) and the Nucleic Acid Data Bank (NDB),
each of whom are using the cif structure as a primary archiving medium.
It will be necessary to ensure that Comcifs can respond quickly to
concerns that affect the day to day operations of these clients.
As the user community expands, individuals and writers of software
will also have concerns that Comcifs will need to address. About half the
Commissions of the IUCr have responsibilities for fields which overlap
with Comcifs. Some form of formal liaison with these Commissions may be
desirable. Such links already exist with the Commission on
Crystallographic Nomenclature and informal links have been made with other
Commissions. Other organisations such as the crystal structure databases
and equipment manufacturers may also wish to have a voice in Comcifs
discussions. We need to find the most effective way to involve these
groups in Comcifs deliberations.
7. FUTURE MEMBERSHIP
We should consider the best form that the membership should take and
how members should be chosen. I raise the following points for
discussion.
1. Is the present arrangement of full members, observers, consultants
and auditors adequate? Should there be other categories, or more or fewer
categories of membership?
2. Should Comcifs only seek members and consultants from among
individuals that it feels could give a broad and balanced coverage of the
field, or should organisations, particularly the major clients, be invited
to nominate their own representatives?
3. Should all the members of working groups be members or
consultants of Comcifs?
4. Should the Commissions of the IUCr be invited to nominate members
or consultants?
5. How should voting members be chosen? According to our terms of
reference, the voting membership should be kept small but it should also
be representative of the areas for which cif dictionaries have been or are
being developed.
6. Should the voting members be chosen from among all those
nominated above, or should they be a separate group chosen because they
represent no particular constituency?
7. Should the representatives of the major clients be included or
excluded from voting membership, or does it matter?
8. How many voting members are needed to provide the necessary
breadth?
9. Should each working group include a voting member?
While there are reasons for keeping the number of voting members
small, there is no particular reason to limit the non-voting membership
since all our business is carried out by email.
8. FUTURE MODE OF OPERATION
1. What changes are needed in our present mode of working?
2. Should the working groups be put on a more formal basis?
3. Do we need to retain working groups on the powder and
macromolecular dictionaries now that these have been approved?
4. Would our work be helped by opening a comcifs discussion group?
Some of the working groups already use discussion groups.
5. Draft dictionaries are already available on the web for those who
wish to consult them. Can we make better use of the web?
6. Do we need to change our voting procedure to ensure that frequent
incremental changes can be made in existing dictionaries?
========================
- Prev by Date: (69) mmCIF dictionary approved
- Next by Date: (71) CIF Core 2.0 and SHELXL; Review of COMCIFS
- Index(es):