Originally Posted by
Josh
They have already announced on the forums that they intend to drop support for it entirely in the near future... Writing a bot for it now would be like writing a story with invisible ink.
Always... The fact is though, what I said is entirely accurate. I've re-read that post on about 5 separate occasions and can only conclude it to be true. There is no way that botting would ever be "impossible". Even if it comes to buying an Arduino and using that with some kernel debugging to gather information.
In reality, what it will come down to is overall detect-ability. Right now the bots can be detected server-side, so being able to detect them client side is not all that useful, it just means the ban rate goes up so chars last less time. If it gets to a point where they are detecting which bits of software are nasty and doing these "sweeping bans" as they used to, only catching everyone instead of some people, then it just means we have to further improve how the bots work to make them less detectable through the old method. The thing is, they are not going to scan and analyse every single person who is playing the game, as the resources required would be insane. What I think they will do is use their existing detection algorithm to find out who is botting. Once they know who is botting from their existing detection, they will look at what software they have running on their PC to try to find the bot itself. Once they find it, they will run some hashing function to get a few "samples" from the bot, then scan all users computers for a binary with that hash. If they find it, they can then delete all of those users. The key fact here is that their old detection system will have to catch the first person in order for the rest of the botters to also be caught. So, the trick would be to create a new, unique bot which cannot be detected by pattern generation - that way the system will never know you exist, so it won't know how to catch you.
The external machine in this mechanism really doesn't need to be all that expensive. You could do it with an Arduino for about $20, or with a Raspberry Pi for around $35, or you could use some custom hardware such as the Intel NUC for around $300, or go ham and do it with a desktop Windows PC. It makes absolutely no difference which you use, except that setting up the Arduino will be difficult. The Pi is relatively easy to work with, and someone could easily build a bit of software to make it a lot easier. Generally speaking, though, there is no reason you couldn't achieve exactly the same results using either a virtual machine, an emulator, or a compatibility layer such as Wine.
People will, as suggested, just stop botting high level accounts. As for WindBot 3, it's basically done but the bugs in it are almost entirely crashes and they are kinda frequent, so it didn't reach release-worthiness in time. It will probably still happen in some form. The logic from it could be applied to almost any of the above mechanisms.
They will lose a lot of money but 15k players paying €8 a month is still a metric fuckton of cash. They can also shutdown servers, remember, and they probably will. The company is relatively small and right now undoubtedly turns a large profit. I would also suggest that they have some money set aside ready to carry them through the financial depression, though.
It's actually only natural, also, that Tibia coins are purchased for in-game cash by botters. They will have to adapt NPC prices and the in-game economy to suit it, if bots don't exist. Try to find me one person willing to kill the mad mage enough times to loot the mage's cap or whatever it is for an addon... Nobody is going to do it because there are 25 waves each time you clean the goo, and the chance of looting it is less than 1.5%, so you have to kill over 1600 waves of shit just to get 1 item... Nobody is going to have enough cash to make it feel worthwhile, and certainly nobody is going to want to clean the place that many times for their own benefit.
