Discussion List Archives

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

Some clarifications are needed

David, your observation that Herbert and my responses "does little to
address the point raised by Brian [T]"... as well as your explanation 
involving the phrase "it should be compliant and able to read CIFs 
using both DDL1 and DDL2 dictionaries." REALLY does highlight, in my
view, the imprecision (verging on confusion) that can arise in these 
discussions. My last input was an attempt, admittedly a primitive one,
to pin down what reading a CIF implies, and to distinguish the 
different concept levels that the STAR File support involves....so
that when we talk about the CIF requirements we are not confusing 
them with the DDL requirements, and so on.

So back to the beginning. Brian T said....

> from any file conformant to the CIF syntax." Someday I would like to
> create some pdCIF read filters. I doubt that my code would tolerate DDL2
> files. Would I then be prohibited using CIF to describe these filter

The meaning of Brian's comment "would tolerate DDL2 files." is what 
triggered my response. Was he referring to whether his CIF parser could
read data files containing names that were exclusive to the DDL2 (mmCIF)
dictionary? Or, was he implying that his CIF reader WOULD involve the 
pdCIF dictionary written in DDL1? That is, what are "DDL2 files", and 
what has this to do with reading a file conformant to the CIF syntax?

My interpretation of the conformant conditions in the draft policy
is that "reading a CIF data file" requires knowledge of the CIF syntax
but absolutely none about reading a CIF dictionary. Of course one has 
to know WHICH data items (i.e. names) have to read, but whether they 
exist in a dictionary not material. That is why my "layered" table
differentiated between a "CIF parser" and a "DDL parser"... the latter
needs the former but not the reverse.

OK, that is why David's comments are also confusing... and clearly if
that misunderstanding exists among this select band, the policy state-
ment has to be more precise and we need to get our terms straight!

My hope is that this document be directed at the "CIF parser" level,
because the requirements here are static and bound inexorably to the 
CIF syntax. Any level beyond that is more difficult and will change as 
the dictionaries and the DDL's evolve.

Cheers, Syd.

------
 syd@crystal.uwa.edu.au         ,-_|\       Professor Sydney R. Hall 
                               /     \      Director, Crystallography Centre
 Fx:  61(8)9380 1118       --> *_,-._/      Deputy Executive Dean of Science
 Ph:  61(8)9380 2725                v       University of Western Australia
 www.crystal.uwa.edu.au                     Nedlands 6907, AUSTRALIA.