Code: Select all
But if I mark as Spam an email, what is responsible to send to the provider that have sent the email, this "spam notification".
If you look at the full list of headers of an email you will see lots of stuff created by the receiving server ( not the user's email client, or the webmail front-end ). Relays may stick their 2c in there too.
eg.
Code: Select all
X-VR-SPAMSTATE OK
X-VR-SPAMSCORE 0
X-VR-SPAMCLAUSE gggruggvucftvghtrhhoucdtuddrgedvfedrfeefiedgudeifecutefuodetggdotefrodftvfcurfhrohhfihhlvgemucfqggfjpdevjffgvefmvefgnecuuegrihhlohhuthemucehtddtnecunecujfgurhepuffvhfhrshggkffftgfgrfgioffqjfesthekjhdtredtjeenucfhrhhomhepoehfohhruhhmsehfrhgvvggtrggufhhorhhumhdrohhrgheqnecuggftrfgrthhtvghrnhepfefggeeuudekgeetkefftdetveejjeeileevvdffffehledvheelvefgtdefhfdtnecuffhomhgrihhnpehfrhgvvggtrggufigvsgdrohhrghenucfkphepvddujedrjedtrddukenvghtpedvudejrdejtddrudekfedrudelhedpmhgrihhlfhhrohhmpehfohhruhhmsehfrhgvvggtrggufhhorhhumhdrohhrghdpnhgspghrtghpthhtohepuddprhgtphhtthhopehfrhgvmhgrnhesphhimhgvnhhtshdrtghomhdpshhpfhepphgrshhspdgukhhimhepphgrshhspdhgvghokffrpefhtfdpoffvtefjohhsthepvhhrfeeh
It is the receiving server which will be configured to check if the account exists, forward bounce or ignore as a function of that and put spam tags into headers and maybe mess with the subject field by inserting something like [SPAM ***] in front of it.
Gmail is a particular PITA for just dumping anything it sees as spam-like according to it's not so intelligent AI.
If I find a useful vid, or newpaper article, I often clip the URL to someone with a couple of words like : "worth a look" or whatever because I can't be bother to write a book about it. Just look.
Gmail finds just a URL or URL plus one phrase to be spam and the recipent never gets to even know I tried to send him something. I don't get to know some jackass AI decided to breach its duty to deliver because it does even bounce it. I just find out next time I speak to the intended recipient and ask what they thought of it !
I try to avoid communicating with lambdas on gmail but if I have to I copy/paste a few random lines from a news article to give Google some spurious data to analyse and store in my profile as a subject of common interest with my lambda friend.
Bottom line, I don't think it is user's spam classification actions so much as AI heuristics.
Sending out batches of dozens of identical emails to widely different addresses and maybe doing so several times in the space of a few minutes maybe part of the trigger problem . Consider this thread:
https://forum.freecadweb.org/viewtopic. ... 74#p627974
Now running to 111 pages it must have a huge number of people getting a notification. One of them replies a couple of minutes later and you have another batch of ostensibly identical and minimalistic content emails with a couple of URL links. I could imagine that would fit a whole load of heuristic checkboxes.
Maybe @yorik would like to consider whether that is not at least part of the problem and possibly avoid it undoing his hard work fixing this in the last couple of days.
I hope the suggestions help avoid a recurrence.