[
https://issues.apache.org/jira/browse/WHIMSY-233?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16736000#comment-16736000
]
Sebb commented on WHIMSY-233:
-----------------------------
Data handling
===========
There can be 700 mod requests per day, so handling them needs to be efficient.
Once processed, as much as possible of the data should be discarded.
However in order to support detection of messages that look like spam, some
data needs to be kept.
Also if a message is incorrectly flagged as spam, it would be useful to be able
to recover.
This means that there needs to be some way to track which other messages may
have been flagged as spam-like.
The Secretary Workbench uses a directory for each month of mails, with a
separate index file.
For moderation, a daily file would be needed, as the volume is much greater.
As far as I can tell, it should only be necessary to hold at most 10 days of
data; after that point ezmlm treats any response as too old (even if it has
still got the data).
> New workbench to support mail moderation
> ----------------------------------------
>
> Key: WHIMSY-233
> URL: https://issues.apache.org/jira/browse/WHIMSY-233
> Project: Whimsy
> Issue Type: New Feature
> Reporter: Sebb
> Priority: Major
>
> Mail moderation sends the same email to each of the moderators, so they all
> have to scan all the mails. It would be a lot more efficient if each mail
> could be dealt with once.
> The suggestion is to implement this in Whimsy.
> The code could be derived from the Secretary workbench.
> This issue is to keep track of requirements etc.
> Functionality needs to include:
> - store mails for posts, subscriptions and reponses
> - show only unprocessed mails (and dynamically update the list)
> - Flag a mail as accepted (and/or allowed)
> - Reject a mail with a suitable message (templates would be useful here)
> - Flag a mail as spam (don't reject)
> - Accept a subscription request (check that the email address is in LDAP so
> can validate it)
> - authorise the moderator so they only process mails for lists they are
> allowed to see.
> Optionally:
> - detect mails that look like spam, e.g. they are very similar to ones
> already flagged
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
