>>/1005683/
Yes it uses IP but it requires someone to do the ban. It works like
> 1st IP before ban, 4 posts before posting media
< poster is banned, script updates it to 8 posts. 8 posts are required from a new IP before posting media
> 2nd IP before ban, 8 posts required before posting media
> poster is banned, script updates it to 16 posts. 16 posts are required from a new IP before posting media
< No recent bans found in 15 minutes or X minutes, decreasing posts required to 4 posts
You could also make it pick a random X number of posts each day X >= Y <= X so it cannot be assumed to be four posts required. Similarly, the script working in the background could detect the number of error post events for the board (i.e., cannot post, you are banned). As I can imagine he is cycling through IPs to find one not banned, the same approach described above could increase the number of posts required before posting a media. So, you are increasing the effort to spam media when banned and when trying to find an IP not banned.
It does require staff members to ban quickly or have the script work on those logs that he is cycling through proxies to find one that is not banned. You could add a logic that it ignores proxies from Australia and tune the scaling so if he's gone manic cycling through IPs but stops, the new posts required before posting media posts drops using some log2 order. Every 5 minutes, decrease the time in half, but I would make that number operate on some level of randomness so he cannot time the attack that he only has to stop for X minutes. Also, if he's trying to find new IPs when the timer is going down, it then also increase the number of posts required before posting media like autoscaling the effort depending on current spam demand.
None of it stops drive by spam posts. The aim is to increase the effort and to make it more difficult. The spammer is in control of their behaviour and if the spammer wants to spend all day trying to game the system, I think that's hilarious.