When somebody finally proves that homosexuality is caused by some sort of cascading series of events that are triggered by exposure to pathogen X... that started in SHEEP... Greg Cochran is going to get a medal. Well, at least a coupon for half off the slushy of his choice at AM/PM. Cola or Swirl?

Realistically he'll probably need to do a Salman Rushdie until things cool off.

unfortunately, nothing about methodology. Hard to evaluate jsut the abstract, it asks you to accept their conclusions at face value. Interwsting though.

Hard to evaluate jsut the abstract

The article itself is only 5 pages -- probably 4 without the abstract & refs. However, the supplementary notes, figures, etc. are FIFTY-NINE (59) pages, about 48 w/o the references.

Pretty soon, both the print & online versions of journal articles will contain only the vowels of the text, while the consonants will appear in Supplementary Information files accessible only online and in a different PDF.

agnostic:

This is painfully true. It's now commonplace to have a crypto paper whose result is the proof of security of some mechanism in some model. The 12-page paper that appears in the proceedings states the results, discusses them, and maybe sketches the proofs. ("Informally, we develop an efficient protocol. To get the 30 pages of essentially unreadable proof, you have to go get the eprint version of the paper. (I'm not sure anyone ever looks at these proofs.)