On 21/06/16 00:40, Pine W wrote:
Sounds good, thanks James.
If we could do scrubs with little risk of bad side effects then I would
be more supportive, but I have not-so-fond memories of what happened
last time.
In the long term, I would like to move away from Mailman toward
something like Discourse. It might make future suppressions straightforward.
Pine
Is an easier censoring its only benefit? :/
Note that it's only the archiving part (pipermail) of mailman what is
troubling for the redacting. We could use other archiving tool (eg.
HyperKitty), without needing to migrate from mailman.
Still, no matter the archiver used, that won't avoid the fact that any
content posted to (most of) our lists is almost instantly mirrored to
many mirrors (as well as hundreds of user's inbox).
Really, this is email communication 101.
Our mailing list archives are norobots precisely to avoid the complaints
of misguided people, and whoever is complaning to legal, probably don't
really care about *our* copy, but about the one on some other archive
out of our reach that appears highly positioned on Google results for
this person name.
Additionally, unless the scrubbing is really careful, history has shown
us that it is *very* easy to get the (already-fragile) numbering wrong,
leading to broken links (or non-sensical) in years-old messages (even
for posts seemingly unaffected, due to previous interventions).
The safest way is to edit the solution is to edit in-place the actual
pages, changing the private bits with "redacted" or similar.
That's about 3 edits per request (1 html page, 1 monthly archive and 1
edit in the mbox). Slightly more if the "really private content" is in
the "From" or "Subject".
Still, I -not knowing the actual “classified files” that were posted to
the mailing list- also lean towards declining such request. Also, when
you grant this request, what will you do with all the people we rejected
in the past?
Regards
--
https://wikitech.wikimedia.org/wiki/Remove_a_message_from_mailing_list_arch…