I like this proposal as amended by Jonathan.  I've always hated the original idea of having different names for the same gene classes in different species -- a 3-letter coding space is just not big enough.

I would suggest that point four be amended to make it absolutely clear that allele numbers (and variants such as polymorphisms, rearrangements, transgenes, etc) are to be unique, not only within species, but also across species.  That is, if Jonathan uses e2381 for a C briggsae dumpy, it can never be used for anything else, and dpy(e2381) would refer to a C briggsae gene, which might someday become Cb-dpy-3.  If syIs1701 is used for a transgene in C briggsae, it can never be used again for anything else, and particularly not for a C elegans transgene.  This is implied by Paul's remark that the Cb in Cb-syIs1701 would be redundant, but it can't hurt to be explicit.

--
Leon Avery                                        (214) 648-4931 (voice)
Department of Molecular Biology                            -1488 (fax)
University of Texas Southwestern Medical Center
6000 Harry Hines Blvd                            leon@eatworms.swmed.edu
Dallas, TX  75390-9148                  http://eatworms.swmed.edu/~leon/