Where to start? I’m a Wi-Fi geezer, one of those old guys who have been doing wireless networking literally since before some of our newest ranks have been born. That means two things here:
1. I have seen a lot of lunacy out of the IEEE 802.11 working groups, the Wi-Fi Alliance, and the WLAN industry as a whole
2. I can assure you, the future looks no better beyond higher-end promised data rates that most client devices will never achieve.
Let me share some real-world use cases to back up my pissing and moaning.
Sorry, Our Client Devices Don’t Do THAT
Some years back when 802.1X authentication became commonplace for Wi-Fi client devices, it got apparent fast just how fragmented the client device market was (and still is). I rant often about “consumer” versus “enterprise” when it comes to device capabilities and the general idiocy of the anemic Alliance to reform and address actual topics of importance, so let’s skip that and get to specifics.
- Printers. There was a time when 802.1X WLAN authentication using PEAP with MS-CHAP v2 was the de facto conventional method for connecting Wi-Fi clients to large business WLANs. And on those large networks, users occasionally need to print. Using printers. On the network. I remember a long and laborious phone call with a senior HP printer development person trying to explain that we needed their printers to do this very standard 802.1X authentication, and it went nowhere. We never found a common technical language to speak on the call, and it was obvious to me that his team lived not only on another planet, but that their planet is in another galaxy far, far, away.
- TVs. Recently, my team troubleshot a Samsung TV that couldn’t connect to a wide open SSID. That SSID happened to be doing OWE Transition mode, and the TV did not play well in the very presence of OWE even though Transition mode is “allowed”.
- This swimming pool thing. Again, troubleshooting a device that won’t connect to an open network.
This high-quality gadget was bought by a large Athletics organization to monitor pool stuff while it floats around. Except… it has very specific WLAN needs per the vendor.
Our gateway is only compatible with standard 2.4 GHz Wi-Fi networks that use WPA/WPA2 Personal security. Open networks (without a password) and more advanced configurations such as WPA3 or OWE are unfortunately not supported at the moment.
As a workaround, we recommend creating a dedicated 2.4 GHz network with WPA/WPA2 security enabled, or using a simple hotspot with a password for the initial setup.
Just a few of many…
Just Whip Up Another SSID!
So what”s the problem? Just crank up a new SSID or three to accommodate all of the oddballs. But wait- too many SSIDs will fill the air with management frames and soon the WLAN landscape will grind to a sucky slowness that sucks because of the slowness! Just ask Andrew von Nagy on that…
But it’s not just the SSID sprawl or expensive magic-in-the middle that MAYBE lets you get away with multiple PSKs on the same SSID that you need to do for endless oddball wirelessly-connected gadgets. I’ve also been told all of the following gems by out-of-touch Wi-Fi client device makers through the years when they wanted to sell a “solution” into our large WLAN environment:
- Our devices are 2.4 GHz only, and they NEED no 5 GHz in the area or they won’t work.
- We need each of our low-data Wi-Fi devices (hundreds of them) to have their own SSID, and preferably an AP dedicated to each with MAC filtering for things we can’t explain. Just do it.
- Our crap only works on a 192.168.1.x/24 subnet.
- Sorry, we MUST use our own wireless router and it MUST be on a 40 MHz channel in 2.4 GHz. Please turn yours off.
I couldn’t make this foolishness up if I tried…
They Say It, We Pay It
Now back over to the network side of Wi-Fi. Don’t you LOVE subscriptions? Don’t you love renting that which you also must buy? And don’t you love “smart” licensing that makes you feel stupid because even Grok can’t understand it?
Value.
This is what we have collectively morphed into. All hail recurring revenue! Now for an added bonus, let’s make each AP cost as much as a ten-pack of APs used to cost… No, wait- we can do better. Let’s make each AP cost as much as TWO ten-packs used to cost. Now we’re up to our ass-bones in value, I tellya. And don’t get too comfy, because in three years the vendors will start the cyclic brainwashing about how you have to replace everything you just bought with the next gen stuff, because value.
But Faster! Faster Forgives All Sins, No? No, it doesn’t.
802.11 “standards” are bizarre. Each one promises more feature shit that can’t be achieved beyond in lab settings under carefully controlled circumstances. But the marketing and hype would have you believe that each is a slam dunk that will deliver 50 Gbps wireless after you drop your coin.
– MU-MIMO? Doesn’t really work in the real world.
– TWT? Doesn’t really work in the real world.
– Bigass wide channels? Don’t really work in the real world.
– 4096 QAM? Good luck with that in the real world
– MLO? Remains to be seen, but absolutely will be limited in actual rel-world usefulness.
This list goes on. Like 8×8 clients (riiiiiiiiight). And it matters greatly because each non-achievable feature gets marketed out the arse and contributes to fictional performance numbers that the industry needs their faithful customer sheep to swallow to keep the magic revenue streams coming.
Put another way- WE ARE NOT GETTING WHAT WE ARE PAYING FOR. Run and tell THAT, bitches.
Do we even want to address the fact that after almost 30 years of Wi-Fi, the WLAN subsystems in Windows laptops are still festering little fragile crap volcanoes waiting to erupt with problems when the drivers get too old (or too new, sometimes)? Nah, my head hurts going there.
AI Stands For Ain’t Improving
Thankfully, Artificial Intelligence is here to make it all right.
No wait, let me rephrase that- thankfully artificial intelligence is here to make the vendors even more money as they convince us how much we need AI while none of the systemic underpinnings of dysfunction are actually being addressed. That’s what really matters. And like clockwork, here comes the next hype-cycle.
Essential that you buy into the next round of bullshit, that is. Because it costs money to get people like Sting to serenade your CEO while your staff weep with adulation.