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

[ddlm-group] Simplifying string handling in CIF2

Dear DDLm group,

This email is in response to Herbert's suggestion of reverting wholesale to CIF1.1 string handling conventions, and including the Grazulis elide proposal to cover the strings that would be otherwise impossible to represent.

As I see it, we are faced with four alternatives:
(1) Revert completely to CIF1.1 string handling and add the Grazulis protocol;
 - this means no triple-quoted strings, but we could reserve the prefix for future expansion
 - we would be arbitrarily reverting our previous decisions on string syntax
(2) Add the Grazulis proposal to CIF2 as per John Bollinger's draft, and maintain the adjustments to string syntax that we've agreed to, but drop triple-quoted strings;
 - again we could reserve the triple-quoted prefix for future expansion
(3) Ignore the Grazulis proposal and continue to search for an elide solution based on triple-quoted strings;
(4) Adopt the Grazulis proposal and continue to search for an additional elide solution based on triple-quoted strings;

I do not think that we will reach a mutually agreeable solution in finite time for a triple-quoted string solution.  Both Herbert and John W. seem quite certain that rich built-in semantics for triple-quoted strings is appropriate.  I and others disagree, and I think we have explored the arguments pretty thoroughly.  Therefore, the practical option would be to drop triple-quoted strings altogether or at least to defer them to a future minor enhancement (by reserving the prefix). 

I think the way forward therefore lies with either option (1) or (2).  As John B. points out, there are no new reasons to revert to CIF1.1 string handling, and this group has already spilt much ink on discussing changes to string syntax, so option (2) seems the most reasonable.

Could everyone please give their thoughts on the above analysis.

James.
--
T +61 (02) 9717 9907
F +61 (02) 9717 3145
M +61 (04) 0249 4148
_______________________________________________
ddlm-group mailing list
ddlm-group@iucr.org
http://scripts.iucr.org/mailman/listinfo/ddlm-group

Reply to: [list | sender only]