Discussion List Archives

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

Re: Core CIF: Fourth round of proposed changes

  • To: Multiple recipients of list <coredmg@iucr.org>
  • Subject: Re: Core CIF: Fourth round of proposed changes
  • From: Sydney R Hall <syd@crystal.uwa.edu.au>
  • Date: Mon, 18 Nov 2002 02:38:42 GMT
David: I had a quick look at this on the weekend and here are some
preliminary remarks.

I agree that the current definition is inadequate, but be very cautious
in replacing it (and any other primary data definitions). The key to
knowledge representation is precise enumeration as this permits auto-
matic validation. One must look to the future therefore and define
descriptive items with "states" that can be searched and validated.
I've attached how colour is handled in our DDL3 core draft. You can
see that _colour can remain the composite descriptor but this maps
precisely onto _appearance sub-descriptors such as... e.g.
and each of these can be enumerated. (The tags in the DDL3 Core
draft are different to the above... its shown simply as an example).

One can also adopt a more quantitative RGB approach to colour (in
the attachment) which can be coupled to automated colour analysers
but I don't advocate this for the DDL1 core.

_exptl_crystal_density_meas_gt   _lt
I don't understand what thresholds "gt" and "lt" have to do with
density measurements. Do you simply mean the upper and lower limits
of a range of density measurements e.g. "high" and "low", or is it
something else?

_exptl_crystal_density_meas_temp_gt   _lt
I really don't know what these temperatures mean!

Would prefer the tag  _exptl_crystal_recrystallization_method!

_refine_ls_extinction_ etc.
Please be careful with what you are advocating here... in my view it's
moving the definitions away from the representation approach that will be
most useful in the future. As above, enumeration is the key and what
would be much better is to carefully enumerate 
to identify the widely used approaches. This simplifies both searching 
validation... and future expansion for new methods.

By the way, the reason for _refine_ls_extinction_expression is because 
were a number of different "Zach" expressions in use by various packages 
one stage... and it may well be still the case!

Thats my lot for today.



On Tuesday, November 12, 2002, at 01:01  AM, I. David Brown wrote:

> Dear Colleagues,
>      Having had no response by the Nov 11 deadline to the last set of
> proposed changes to the CIF core dictionary, I assume that everyone is 
> in
> favour and these items have now been marked as approved.
>      I enclose below the final set of trivial (or almost trivial) 
> changes.
> Once these are approved we can start on the more contentious changes.
> However, the changes below are not entirely trivial and I need some 
> guidance
> on the _refine_ls_extinction items in answer to a number of queries 
> following
> each of the definitions.  PLEASE can I have some feed back.  If there 
> are
> potential problems you will be the people who have to live with them if 
> you do
> not respond.
>      The deadline for reply is 16 December, though this can be extended 
> if a
> longer period of discussion is needed.
>      I look forward to hearing from you all.
>                     David
  Head@biomedchem.uwa.edu.au        ,-_|\   Professor Sydney R. Hall
  Syd.Hall@bcs.uwa.edu.au          /     \  Head, School Biomed & Chem 
  Fx:  61(8)9380 7330          --> *_,-._/  Faculty of Life & Physical 
  Ph:  61(8)9380 2725                   v   University of Western 
  www.biomedchem.uwa.edu.au                 Crawley, 6009  AUSTRALIA.

[Send comment to list secretary]
[Reply to list (subscribers only)]