so 10. 2. 2018 v 13:23 odesílatel Guilherme Gonçalves < guilherme.p.gonc@gmail.com> napsal:
Hi Martin,
I'm not authoritative on PII policies at all, but here's a couple of things that came to mind as I read your question.
2018-02-10 11:26 GMT+00:00 Martin Urbanec martin.urbanec@wikimedia.cz:
To prevent this tool from spamming I of course require its confirmation by accessing an URL with a random string (MD5 hash of user's email *and* random number from 1 to 100; I mean, those two things are in one hash).
Does this mean the URL for a given email address can be guessed in at most 100 attempts by someone who doesn't control the address? I think you'd typically want to draw your random numbers from a much larger range, or use as token something that was encrypted or signed with a secret only your server knows. It would probably also make sense to make your URLs valid for only a certain time.
*1000, but increased to 10 000 000, which should be big enough. I also can use more qualit hash than MD5 which will slow it down even more.
However...
Should I stop with collecting mails at all and use some WMF-maintained service for mass-emailing (mailman at lists.wikimedia.org maybe?) and make the tool to just send an email to the list itself?
If creating a single mailing list is an option (for instance, you don't plan on customizing the emails per user), this seems like a very good way to go.
It is, this just was the easiest way for me when I was writing the tool.
This question came to my mind before creating, so I do appologize for asking after programming.
Best regards, Martin Urbanec -- Můj kalendář najdete na https://martin.urbanec.cz/calendar.html
Wikimedia Cloud Services mailing list Cloud@lists.wikimedia.org (formerly labs-l@lists.wikimedia.org) https://lists.wikimedia.org/mailman/listinfo/cloud
-- Guilherme P. Gonçalves _______________________________________________ Wikimedia Cloud Services mailing list Cloud@lists.wikimedia.org (formerly labs-l@lists.wikimedia.org) https://lists.wikimedia.org/mailman/listinfo/cloud