Roles are handled in a source_citation for the event in the following way by GEDCOM:
1 SOUR @XREF:SOUR@
2 PAGE "WHERE_WITHIN_SOURCE"
2 EVEN "EVENT_TYPE_CITED_FROM" [ ADOP | BIRT | BAPM | ...]
3 ROLE "ROLE_IN_EVENT" [ CHIL | HUSB | WIFE | MOTH | FATH | SPOU | (ROLE_DESCRIPTOR)]
1 SOUR @S1@
2 PAGE p. 57
2 EVEN BAPM
3 ROLE (godfather)
I've read and evaluated "bettergedcom" and I am NOT a fan. They have a good mission (and I endorse the idea of a new GEDCOM), however I think that they are missguided. The GEDCOM standard (as with all real standards) has a governing body. The GEDCOM standard governing body is overseen by LDS and unless they (LDS and the keepers of GEDCOM) give up control of the standard nothing "bettergedcom" produces can become offically a new version of the "standard". They can make recomendation just like anyone else, but to my knowledge LDS has internalized the standard and moved beyond their need to augment and revise it.
Therefore we are stuck with what we have. However, IMHO all is not lost if software vendors would only:
1) follow the standard by reading/writing/using all published tags of version 5.5
2) adopt and consider "published" via handshake version 5.5.1 of the standard
3) remove all custom tags that have actual gedcom standard tags (both published v5.5 and draft v5.5.1)
4) remove all custom tags that can be written using gedcom standard EVEN and FACT tags. These GEDCOM tags are more descriptive and can be better shared than the 4 letter custom tags most software uses.
5) Learn to share. Most genealogy software vendors want to "lock you in" and never let you go. This is the main reason the GEDCOM standard will never move forward or if it ever does most software vendors will never adopt points 1-4.
Most issues with exchange using GEDCOM will be solved if vendors adopted these ideas. The few issue that are still nnot solved, need review. ASSOciations at the event and attibute level is one case. Historical place names (introducing a 0 level place) is another.
What we don't need is the continuation of tags like:
_DNA, DNA Markers
_INIT, Initiatory (LDS)
_MDCL, Medical Condition
_MILT, Military Service Desc
_MILTID, Military Serial Number
_MISN, Mission (LDS)
Many could use already defined tags.
_WEIG and _HEIG = DSCR (Yes, you must enter the data as "1 DSCR Hair Brown, Eyes Brown, Height 5 ft 8 in"
1 FACT "marker data"
2 TYPE dna marker