[openstreetmap/openstreetmap-website] Limit number of edits per user and day (#2342)

[zstadler]

One example that got through is a new user [vecago7283](https://www.openstreetmap.org/user/vecago7283) who was able to submit 99 changesets within 3 hours (between  2023-10-21T14:56:34Z and 2023-10-21T17:52:27Z). According to a simplistic Overpass Turbo query, these changesets included 216722 unique nodes, 426 unique ways and 288 unique relations. Many elements were modified multiple times, making the revert effort more complex than usual.

IMHO,
* Malicious edits need to be considered as an evolving permanent threat to the OSM data integrity.
* Just like viruses, it can be safely assumed that agenda-driven vandals will steadily improve they methods and learn the safeguards OSM has and will have put in place
* > Maybe it would be feasible to reduce limits by factor of 10 and count each tagless node as 1/100 of object? That would reduce vandal bit impact without risk of blocking new people mapping landuse.
   
   Modifying un-tagged nodes belonging to ways can create havoc in the map. I would not advise them to be counted with 1/100 multiplier.
   ![image](https://github.com/openstreetmap/openstreetmap-website/assets/1304610/63ee0381-1c17-4348-bcdc-c47c7a3c76d7)
* Perhaps lower rate-limits should be made on deleting and modifying existing data and higher rate limits will apply to new data, which also happens to be easier to revert. Perhaps modifying an existing element would count as much as adding 10 new elements.
*  > Is it viable for new legitimate user to edit more than 5000 elements and try to submit as their first edit and become stuck

   Each element modification should be counted, not just the total number of elements modified.
   I also tend to disagree with this generosity towards new users.



-- 
Reply to this email directly or view it on GitHub:
https://github.com/openstreetmap/openstreetmap-website/issues/2342#issuecomment-1775719888
You are receiving this because you are subscribed to this thread.

Message ID: <openstreetmap/openstreetmap-website/issues/2342/1775719888 at github.com>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/rails-dev/attachments/20231023/87c725a2/attachment.htm>