the firmware releases were infrequent, the distro is not fully hardened, the linux kernel is old, and imagine having to get this notice now, at this point, and know that there wont be any further updates that fix future vulns. if u have this product, your firmware is now frozen and youre vulnerable to every future issue that affects it. looks like their support stops in 2029, but imagine these devices stay in place for another 10 years, this is a nightmare scenario for anyone with large deployments. in lots of cases we expect some devices to be in place for up to 45 yrs. still, this system wont make it past 2037, max. imagine running it for 12 more years without any fixes. untenable.
Buying large number of a hardware product automatically gives you some preferential treatment, as in both the client and the manufacturer must cooperate deeper just to absorb impacts and make logistics work.
You can't just click to buy a boatload of wireless earphones and have it delivered overnight, because regular boatloads has to go to regular stores. Regular 3-day shipping will become a 3-month delivery reservation or a 18-month supply contract. Potentially a special batch production, shipping, and trucks from port to your warehouse might have to be arranged. At that point the manufacturer might as well accept minor tweaks such as logo tweaks and configuration changes for that manufacturing run, or could even negotiate with you to take one just to prevent that batch disrupting regular consumer sales and supports.
But it won't increase combined total available budget of that operation above combined cumulative margins and cash flows, it won't print money, only distribute more effectively. So your contribution towards total "market cap" of the product might not make up for lost sales if product flopped even if you were an important customer.
yeah but even then, it's still just a contract. The legal math behind Cisco's decision might just be to extend support just long enough to meet obligations and tolerate the risk that the remaining customers might sue.
Yeah we now support 1.5M devices with our community network and enterprise network. And the most successful ones sent the least amount of data. Here is a list of all the supported devices: https://www.thethingsnetwork.org/device-repository/
Yes, this is an outlier and it took around 8 years for this to be emerge and be detected by the community run TTN mapper initiative. The main fun thing about this is more that it shows how cool it is to have a global community network where we enjoy sending messages for the sake of it just to learn technology. Like kids with 2 cans and a string between them.
So it takes more than 5 dollars. You also need a gateway but you can get these for 99 dollars. Still this technology has a very low barrier to entry for such a long range and low power capability.
For people reading this since I'm certain wienke knows this, you can also do point to point or mesh LoRA between cheap devices (e.g. Meshtastic). You don't necessarily need a multi-channel gateway, depends on what you want to do.
Good point. We'll add that to the next post if there is a new record. Also what can be confusing is that the business viable part is even much smaller than the technical viable part. The low power operations is really only at lower SF7 or SF8. And we push our partners and ecosystem to keep the payload as small as possible as every byte counts. It is a completely different way of thinking than WiFi or cellular. Yet I do agree it is presented with the same terms and paradigm and that is confusing.
On that point, have you looked at Myriota's research? (https://myriota.com/) In developing their system they went right back to the fundamentals and reworked everything around the idea of short messages. For example, an implicit assumption in the Shannon bound is that the message length is long, so they figured out that their short message system doesn't obey the Shannon bound and designed the coding and so on accordingly. It's fascinating stuff if you're into information theory.
