I'd like to use my lecture to test an idea I've had for years but never got around to, probably because it isn't economically interesting. It's about MikroTik Wi-Fi access points, which have been proverbially shitty for as long as we've known them. At first they were shitty in absolutely every respect, then they somehow sorted out the radio side, then they introduced CAPsMAN — and now they're shitty in only one remaining area: CAPsMAN has no concept of RRM, radio resource management. So if you place a few of these access points in a space where they can at least partially see each other and connect them to CAPsMAN, the system as a whole will pick frequencies in such a way that the APs then interfere with each other with remarkable efficiency. The interesting part is that MikroTik has all the tools that would let it solve this problem in a relatively simple way. The key one is each radio's ability to background-scan the frequency band we restrict it to, discovering both our own and foreign access points and measuring the signal strength at which it sees them. It can also uniquely identify the APs by their MAC addresses, which are part of the beacon signals they transmit. Unfortunately, MikroTik never actually uses this data anywhere — so during the lecture, we'll try to do it ourselves. The goal of the talk is therefore to build, in 45 minutes, an MVP project delivering functionality that goes well beyond what RouterOS gives us out of the box.