Re: Google Gmail mailing list bounces

[Bob Proulx <bob@proulx.com>]

Nate Bass wrote:
> I observed 3 different behaviors when reporting spam in Gmail. Don't
> worry, I did not report any GNU mailing lists :)

Awesome! :-)

> When reporting this current message (help-gnu-emacs) as spam Gmail
> will prompt first whether you want to "Unsubscribe and report spam" or
> "Report spam". Again, I did not click anything on this dialog so
> help-gnu-emacs will be unaffected. Unsubscribing redirects you to
> https://lists.gnu.org/mailman/options/help-gnu-emacs. There is also a link

That is the List-Unsubscribe header specified URL.  Good.

> to "learn more" at https://support.google.com/mail/answer/1366858?hl=en

Also reasonable but that documentation answer looked old and stale as
it did not mention the new behavior of unsubscribing.  Guessing that
the behavior was updated to unsubscribe but all of the nooks and
crannies where documentation such as the above was stuff has not been
updated.

> When reporting other lists (for example bug-make) it prompts the
> same dialog except it will not give a link to unsubscribe. Instead,
> it will send a message to bug-make-request with a subject of
> "unsubscribe" and a body of "This message was automatically
> generated by Gmail."

There is no difference between bug-make and help-gnu-emacs as both are
on the same server running the same software.  The only difference I
can think of here is if you were subscribed using different addresses
for one than the other causing Gmail to think you were subscribed for
one but not for the other.  Or something like that.  Because they are
otherwise identically run.

Please look at the message headers and see if they are different
coming to you.  I think the Gmail interface is the more actions or
something where it says "show original message" or something similarly
worded.

> Oddly, when I reported w3c-announce@w3.org it did not prompt anything
> and went directly to spam.

Please check the headers of that message too.  Perhaps that one did
not have the standard List-* headers and did not have a
List-Unsubscribe header?  If not then there would have been no way for
automated machinery to know.  But on the other hand if those do exist
for that message than that is strange that it would be different.

Thanks for the report!  Good stuff.

Bob