Discussion List Archives

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

Modus operandi

Dear COMCIFS members

I append below a digest of the responses received so far (one) to David
Brown's suggestions on procedural matters. Paula raises some points on
this which are very relevant to the best way of handling our business.
Because we are a technical committee, the bulk of our effort should go to
working on technical matters, rather than refining points of procedure,
but an early consensus on the most efficient way of doing business would
let us get on with the job more rapidly - and I think it would be only fair
to Paula, in particular, to be getting on with technical matters (especially
with the rapid approach of next month's New York CIFtools workshop).

So I again invite you to respond to David's and Paula's comments at an
early juncture.

Regards
Brian

----------------------------
 From bean@com.merck Thu Sep 16 20:36:40 1993
Subject: Reply to David's message
To: bm@uk.ac.international-union-crystallography (Brian McMahon)

Fellow committee members and consultants:

I have read the message that we all received from David Brown earlier 
this week, and have some thoughts.  David's comments really address two
separate issues (mode of dictionary approval/upgrade and mode of 
communication amongst committee members) and I will deal with them 
separately.

o Mode of dictionary approval

I have no fundamental problem with the 6 stages that David has outlined
(although I have to admit I find it depressing to think that after so 
much effort the macromolecular dictionary is still at stage 2).

Having spent three years in the midst of this process, however, I 
realize that the process of dictionary development is not quite that cut 
and dried.  I think that these steps underestimate the complications 
arising from two sources:  1) the learning curve associated with 
learning how to "write good CIF" and 2) the difficulties that come from
building dictionaries as extensions to a core that is itself dynamic.

1)  An enormous amount of time and effort would have been wasted with 
mmCIF if we had not had Syd to interact as we proceeded.  Once you know 
CIF well, it becomes easy to "write good CIF".  Until you know it, you 
do a lot of things clumsily, and some of the things that you propose are 
just plain wrong.  We have to have a mechanism for providing a 
subcommittee council and guidance often during the stage 2 from a "CIF 
guru".  Whether that role is played by a single individual (the role 
that Syd has played for me and my subcommittee) or whether that role is 
better played by the ComCIFS as a whole, is open for discussion.  
Perhaps this is what David meant by 2c) consulting with other members of 
comcif as may be required, but I fear that that underestimates the 
importance of this kind of advice.  Someone with real CIF savvy (and a 
thorough knowledge of all of the dictionaries) has to be on call.

2)  My mail of last week to the whole committee raised the issue of the
relationship of a extension dictionary to the core.  The particular 
issue was the *_intro items, but the issue is more general.  The core 
does evolve, and it has to be prepared to evolve in real time with the
development of each extension dictionary.  A number of items originally
proposed for the mmCIF extension have already been co-opted into the 
core, at which point the subcommittee lost control over the definition,
examples, enumeration of that data item.  There has to be someone in 
charge of the core, working closely with the subcommittees.  Of course 
that "someone" could, and perhaps should, be the committee as a whole, 
but if we decide to handle it that way we must realize that this will 
mean being a very active and responsive body during the development of a 
new extension dictionary.

I realize that I see these problems from the perspective of my own 
experiences and current situation.  It may well be that the mmCIF 
extension dictionary is the most complex extension dictionary that 
anyone will ever have to deal with, and so these issues are local to 
this stage of the development of the CIF standard and will not be 
practical problems once we have a formally accepted mmCIF dictionary.  
But I have heard a good bit of talk about trying to extend the CIF 
paradigm to the macromolecular structures determined by NMR techniques, 
and I can assure you that these will be very real concerns if someone 
does undertake that task.

o Communication amongst committee members

I must admit that I rail a bit against all input going to Brian, and 
only being made available to the committee as a whole after been 
digested and summarized by Brian and/or David.  I can assure you that 
given the volume of email that I receive, the volume increase resulting 
from messages broadcast to the entire committee will be negligible.  
Still, I am willing to give the mechanism a try.  However, I reserve the 
right to ask that we reopen the issue of direct broadcast to the whole
committee, if I feel that this moderated mechanism is causing my message 
to be lost or diluted, or causing delays in information dissemination.

Again, my attitude about issue this may well be local to my current need 
to finalize the mmCIF extensions.  I really do need help from you all, 
and in a timely fashion.

Paula