[cctbxbb] Troublesome Hall Symbols in CIF

[Luc Bourhis]

Hi,

Consider the following two Hall symbols:

(1)  -C 2y (x,y,-x+z)
(2)  -I 2y

They represent the same setting of the same spacegroup, so in a perfect world the fact that the sgtbx comes with (1) from some list of operators should not matter. Unfortunately, (1) is rejected by CheckCIF right now. Reminder: every single small molecule publication in Acta Cryst. must pass CheckCIF. On the contrary, CheckCIF accepts (2).

Yes, I know, this is a bug in CheckCIF. But the IUCR is very slow at fixing bug in that kerberos of a program and we do therefore put our Olex2 users in trouble when they use the smtbx to refine their structure, as opposed to ShelXL, which does output (2). Besides, (2) is clearly more pleasing to the human mind than (1). Thus I reckon there is definitively a case to make a change in the cctbx here. The question is at which level?

I don’t think we want to touch the sgtbx. I am of the opinion of patching iotbx.cif. Would that cause any trouble in Phenix and consort? Any remark or suggestion?

Best wishes,

Luc