Gregory Maxwell said:
For example our articles on Free Software related
subject often
contain information sourced from mailing lists and form the
editors experience, sources which would not normally be acceptable
under the normal application of the original research standard.
"Personal experience" isn't verifiable. Such sources must be disregarded.
If someone says "Linus in 1993 contemplated starting again, abandoning
the GPL" and cites a personal conversation, then delete it. If he can
cite a mailing list that is archived on a public server, I'd cite that and
check with Torvalds himself if there was any question over authenticity.
Mailing list archives can be, but are not always, for good old fashioned
reasons of checking provenance, acceptable sources (at the risk of
stirring up that whole debate again.)