What good references on programming style assert this dogma?
Because it looks like a caricature of the topic, somewhat like how some Agile people incorrectly argue about how first there was waterfall and then came the Agile manifesto.
Just because a lot of people learned crap habits fourth-hand, doesn't make it a dogma. Dogma - real dogma - must be come from authority.
McConnell's "Code Complete", from 1993, has a checklist of "Unusual control structures" on p366 of my copy. This is much closer to what the real dogma of goto looks like:
[] Are gotos used only as a last resource, and then only to make code more readable and maintainable?
[] If a goto is used for the sake of efficiency, has the gain in efficiency been measured and documented?
[] Are gotos limited to one label per routine?
[] Do all gotos go forward, not backward?
[] Are all goto labels used?
Most of the CPython examples fit these criteria. I don't have experience with the other projects. My own use of goto also fits this form.
