That is only a problem if you have user info. On classical forums, everything was public if there was no PMs.
Also under GDPR IP address retention can fall under that in your logs depending on how you manage the info, same way analytics would need attention.
It should really be more simple, and I wish we'd have more people just building small sites here and there without having to care much about these details.
This scaremongering about GPDR always conveniently forgets that the stuff the GPDR makes scary to do is stuff you should not be doing anyway.
Collecting IP address will not reduce the need for manual moderation, nor even reduce bot spam...
(I do admin a topic-specialized technical web forum which has been running for over 20 years, which doesn't even need to ask for cookie consent.. but of course, has never been legally tested, so YMMV)
Second, the new user is subject to approval and all his messages are moderated until a couple of posts. If he starts posting decent material, why would you care he was banned 2 months ago -- are you really that punitative? And if he just posts decent, insightful material for his 5 messages, gets approved and then descends back immediately into trolling, well, at least it's been a couple of months and then the couple of weeks he was under moderation of total silence, plus you have gained whatever contributions he made under moderation. I am yet to see a troll that will repeat this process (which requires insightful contributions in-between each iteration of trolling) for more than a year. That's way less moderating effort that it takes to handle new users at all...
You'll see strangely repetitive user coming from different AWS's IP blocks. User showing constant sessions from 5 or 6 IP blocks that seem to belong to different countries. Or users having a pretty consitent IP but asks for help right after a new login happened on a different IP.
It can be replaced with other info, but it's still a simple and pretty powerful bit to know about your users.
It gets messy when you have a number of users and need to know why a specific user is sending you 150 requests per min for 3 hours: you'll insta ban that user, but still need to understand what happened. Did they get their login info leak and the whole internet is having a field day with it ? Is it an issue with your system and their browser is stuck in a weird loop ? Are you session management backend going bust and they're actually all different logged users ?
You'll only know if you have the IP, parts of the headers and some more debugging info, and activating the debug after the fact is often not good enough.
Do you have more info on this story about them?
https://www.theverge.com/23607095/vergecast-solo-acts-hive-s...
I found the story of Hive (the social network) fascinating in that it was basically that idea of someone jumping in, learning the ropes, building a platform. And getting crushed under how much knowledge and effort, ultimately money, is needed to run a decent platform.
Their team was truely incompetent, but looking under the surface they were really a group of a few amateurs trying to do something, and it's only after the backlash that they realized a whole team of well experienced engineers is needed for anything touching user info.
I'd imagine trying to run a raw phpBB on the web today would be in the same vein ?