COMCIFS adoption of DDL1

David Brown idbrown at mcmaster.ca
Fri Oct 14 16:14:12 BST 2005


The CIF dictionaries are written using a Dictionary Definition Language 
(DDL) which is described in a DDL dictionary.  Both the DDL- and 
CIF-dictionaries use the same DDL and syntax so that is should be 
possible to read any of the dictionaries using the same software..

Currently there are two versions of DDL in use.  DDL1 is used for the 
coreCIF and related dictionaries while DDL2 is used for the mmCIF and 
related dictionaries.  When these various dictionaries were originally 
designed it was anticipated that a program designed to validate a CIF 
could also be used to validate CIF- and DDL-dictionaries.  However, 
since most of the programming effort has been directed towards the 
manipulation of CIFs, it is only quite recently that anyone has taken up 
the challenge of writing software to validate CIF- and DDL-dictionaries 
against the appropriate version of DDL.  In the course of this work it 
became apparent that DDL1 contains some errors that make such validation 
impossible.  The errors are easy to corrected, but there is no 
individual or organization that has the responsibility for managing DDL1 
or the authority to approve any changes.  DDL1 is an orphan even though 
it underpins the whole CIF project.

DDL1 was written by Hall and Cook (J. Chem. Info. Comp. Sci. (1995) 35, 
819-825), but they have moved on and are not interested in managing the 
earlier versions of DDL.  At the COMCIFS closed meeting in Florence 
there was general agreement that, because an authoritative definition of 
DDL1 is essential to the functioning of many of the CIF dictionaries, 
COMCIFS should take responsibility for its maintenance.  That decision 
will need to be formally ratified by the voting members of COMCIFS, but 
before I ask for that vote, I am circulating the proposal to members of 
the COMCIFS Discussion List for their comment.

The Florence meeting of COMCIFS also endorsed the idea of moving towards 
a new (third) version of DDL (star_DDL) that would be compatible with 
CIF archives written in either DDL1 or DDL2, but that would in addition 
benefit from what we have learned over the past fifteen years and 
consequently would result in a much more powerful version of CIF.

One might ask why it is necessary to patch up DDL1 if it is likely to be 
superceded by star_DDL.  There are a couple of reasons.  Firstly, there 
is much work to be done before star_DDL is ready for general use, and 
dictionaries written using earlier  DDLs will likely continue in use for 
many years in parallel with star_DDL.  Secondly, the experience gained 
by developing advanced software for DDL1 will be useful in developing 
software for star_DDL.

If you have any comments on this proposal, can you please post them on 
the COMCIFS Discussion List by Replying to this message during the next 
four weeks.  After that time I will ask the voting members of COMCIFS to 
formally adopt the proposal that COMCIFS take responsibility for the 
management of DDL1.

David Brown
Chair of COMCIFS
-------------- next part --------------
A non-text attachment was scrubbed...
Name: idbrown.vcf
Type: text/x-vcard
Size: 298 bytes
Desc: not available
Url : http://scripts.iucr.org/pipermail/comcifs/attachments/20051014/c82cac73/idbrown.vcf


More information about the comcifs mailing list