Should data names deprecated in DDL1 be deprecated in DDLm?

[vaitkus]

The DDL1 version of the CIF_CORE dictionary contains many deprecated data items. There were two main types of deprecation:
A) Deprecation due to semantics (_atom_site_refinement_flags, _symmetry_cell_setting).
B) Deprecation to discourage the use of certain potentially misleading terms (e.g. _refine_ls_goodness_of_fit_obs was replaced with _refine_ls_goodness_of_fit_gt).

While data items affected by type A deprecation seem to be properly represented in the DDLm version of the dictionary, data names affected by type B deprecation were converted to regular aliases without any notice that their usage should be avoided. Note, that the DDLm reference dictionary does have the _alias.deprecation_date attribute to denote just such alias, but currently it is no widely used. Should it?

[jamesrhester]

Yes, I think we should deprecate type B items as well.