Request: bring the oVar/@type in line with the independently existing mechanism for referencing grammatical features.
When looking at http://www.tei-c.org/release/doc/tei-p5-doc/en/html/DI.html#DIHW , it seems awkward to me to use the @type on oVar for exclusively and purely grammatical information ("1] pt(past tense) ; 2] pp(past participle) ; 3] prp(present participle) ; 4] f(feminine) ; 5] pl(plural)").
We look here at a microsystem (indeed a tumor, I would say) of taxonomy with a corresponding set of arbitrary (even if common) abbreviations. I say "tumor" and "arbitrary", because in this same chapter, we introduce the gram and gramGrp, which can be given externally-anchored interpretations thanks to the DCR attributes. Using @type for this, and listing the values of @type as completely independent and dissociated from the gram(Grp) system, feels wro-o-ong.
This is related to the revamping proposal of the o/p-Ref/Var series. The @type attribute should be left open by just assigning the new oRef/pRef couple to att.typed.
seems reasonable
Council 2015-05-28: Piotr, would these actions answer the need?
Remove the override of @type so that it no longer has recommended values (which would leave existing documents valid);
Replace the recommendation to use @type for grammatical categorization with a recommendation to use the *datcat attributes instead, in the chapter prose and in the elementSpec.
It is a step forward, but remember the ticket is obsolete, if you can take a good decision on: https://sourceforge.net/p/tei/feature-requests/545/