On Fri, Apr 19, 2019 at 4:22 PM Jay Hennigan <mailop-l...@keycodes.com> wrote:
> On 4/19/19 2:31 PM, Brandon Long via mailop wrote: > > I just don't think this is practical. > > > > For one, when you're only solution is to reject, the only way to get a > > signal that you're rejecting the mail wrong is manual review, which is > > impractical at best, and difficult to correlate with the opinion of the > > actual receiver. The spam/not spam signal from users is the best > > information you have on what your users want, even if the bad actors try > > to game the signal and a lot of user's use it as a hammer instead of the > > softer touch. > > I agree with the utility of spam folders in general. In the case of > webmail you can also deliver to the inbox with a visual indicator of > spam before the mail is opened. > Why is this limited to webmail? Anything you can do with webmail you can do with a thick client. I don't see many clients that work this way. > Even without these things, often we aren't sure that something's spam, > > so we rely on the folks always checking their email and clicking spam to > > inform us on messages we've already received but haven't been looked at > yet. > > This feedback is only really available for webmail, so you don't need a > separate spam folder. If unsure, deliver to inbox with a visual > "Suspected spam" flag on the individual message. Mail not flagged as > spam should have a clickable "This is spam" or better yet, "Report as > spam" button. This should be very distinct from the "Delete" action to > minimize false positives, perhaps even with a confirmation dialog box. > There's nothing to prevent client spam/not-spam markings from being synced to the server and used to train. Many of the popular clients have standardized on one of two imap keywords for that. Gmail listens to IMAP clients making spam decisions. When iOS makes access to the report junk easier, we see an increase in manual spam markings (took us a while to figure out what was going on). Mail that is flagged as spam can have a "This is not spam" button to > provide user feedback against false positives. > > In other words, you can get the feedback from the recipient as well as > flag suspected spam to the recipient without the need for a separate > spam folder. Based on that feedback as well as existing other metrics a > decision can be made to hard reject similar mail in the future either > globally or per recipient. > This is indeed a mechanism to not have a spam folder. It also seems to be either a terrible idea, or a its papering over a terrible anti-spam system. We've invested a lot of effort already in splitting people's inboxes out into separate categories to attempt to reduce the clutter, bring spam back into the inbox is the opposite of that approach. With priority inbox, we try to highlight what's important as well... and spam isn't it. And that's on top of the fact that user's interacting with spam/phishing is dangerous. When last I ran my own personal system, I was leaking a dozen or more spam messages a day through to my inbox, and finally decided to throw in the towel because I just didn't see the point of interacting with it even that much. Yes, as our spam filtering has improved, that does reduce the amount that user's spend in their spam folder, and we get less signal. No one said this was easy. Brandon
_______________________________________________ mailop mailing list mailop@mailop.org https://chilli.nosignal.org/cgi-bin/mailman/listinfo/mailop
