@redpillschool Linksys WRT3200ACM

You asked a long time ago about routers. This is what I would get now. The Wifi 5 on this is good enough. Then later say when Wifi 7 comes out then simply disable the signal on this and just use it as a router. Then you can put your Wifi 7 access points where ever you want.

The reason that this router is the best is that it was made by design to be loadable with custom firmware, hence the WRT in its name. The processor and memory on this are very good.

I check it out and there is an nginx package on this that also works with the admin portal (it has to switch to using another web server). So that means that you can put a reverse proxy vhost config right into the router so that via SNI (server name identification) headers that you can host multiple standard-port webservers on your network without needing a separate proxy box.

This router should be fast enough to last say 10 years or so (after you turn off the wifi antennas).

This is one of those rare situations in which it's better for you to be in faggot-ass America sitting there like a bitch with your Amazon marketplace. I'm going to have to struggle to get one of these. You suck. Fucking JQuery-level kinds of...

So I'll get one of these for the main DHCP router and then the others will be just access points providing 802.11r fast roaming with the 802.11v and 802.11k supplements to make it actually good. Then I'll just pop on Wifi 6 and 7 cheap nodes as they come up.

That brings me to my next point. The CPU on these things needs to be decent in order to have all of the thrills. So you need a somewhat modern ARM CPU to be able to use Wireguard VPN without it taking up to a 90% speed hit. There are MANY cheap MIPS CPU routers out there which are perfectly fine for being access points, but fall far short of being a general purpose router and application server (as the Linksys WRT3200ACM is).

So I'll get this and put it on 5Ghz only for the center of the house. Then the outside ones will be 2.4Ghz (and possibly 5Ghz).

The big benefit of Wifi 7 though will be that the wifi can simultaneously transmit on 2.4 and 5 so that the receiving can choose which data stream that it accepts. This will be really a huge improvement. It's not worth waiting for though, and I learned that 802.11r fast roaming allows Wifi versions to be mixed, while 802.11s possibly does not (I can't find the answer to this). However, for sure it's at completely necessary to run the same kernel versions for any mesh node. So that really isn't compatible with a chaotic mixing of various devices that might not work on specific kernel versions, etc (like WINE for example). It's more so for a larger scale coordinated projects where you roll out everything at once with like 500+ nodes.

My research has shown that I don't need a 802.11s mesh. What I need is a 802.11r fast roaming network with 802.11v and 802.11k to make it jump connections properly (so that it doesn't get stuck on a bad connection).

Furthermore I can combined wifi 5 and wifi 6 (ac and ax) devices without any issue. So that really frees everything up, big time. The problem is that before I understood this it was looking like that I'd be married to whatever I bought for a long time. Now that I'm free to mix and match then I can get whatever is the best value for the task.

The problem with 802.11s is that it uses batman protocol behind the scenes as an implementation detail, and this requires all devices to be using the same kernel. So that somewhat meant to me that all devices needed to be the same.

I'm thinking that I'd make the wifi in my house to be 5Ghz Wifi 6 only. Then the outside ones would use Wifi 5 or 6 (depending on the need) on 2.4Ghz (which is a lot slower). I'm never going to need crazy fast Internet outside. That is just a work thing.

I've made a list of Wifi 6 devices that can be purchased locally for OpenWRT router firmware. I'm now trying to figure out if it makes sense to get a router wit a very powerful CPU, memory and storage so that I could run my Golang webserver with Video4Linux2 webcams right off of that, saving stupid mini computers. But it's hard to get benchmarks on these processors to know if any of it is in the realm of usability. I have no idea if it's ridiculously overpowered or stupidly too weak.

I think that I'll need to get webcams that are delivering some degree of compressed video already, even if it's just a fast pass one. Basically becomes doing video encoding on a router probably isn't a great idea.

Damn though, I might be up on this a lot faster than I thought.

I tried playing Breath of the Wild on Yuzu emulation for about an hour, but it really did seem like a late beta game to me. There just didn't seem to be much going on. I'll give TOTK a try some day to see if it can hold me.

I'll add that the success and amazing graphical style of this game seems to prove that game design is still king, and that the latest consoles are not necessary to provide for a good-enough platform. Lately I find myself only playing indie games with "retro" graphics style. They're trying to be nostalgic for my generation, but also it is just much cheaper to make. It's more important that they concentrate on gameplay. For example Dead Cells, Slay the Spire.

I'm really not impressed by most AAA 3D games because of how much longer that they need to be completed. The market doesn't seem to be capable of fleshing games out relative to their graphical and scope potential. Like Cyberpunk 2077 only became half-way decent a few years after release. They'll still be improving that game, and releasing revamped versions of it for the next decade. At some point it will become a monument, but who wants to see a half-built statue? These games need however many years before release plus another 5-10 to really be amazing.

It seems that only Nintendo culture and economics are capable of cooking a game long enough to be flawless on release like how 80's and 90's games came on the console.

The Legend of Zelda: Tears of the Kingdom Review