I know what they spend their time on - like all OSS devs they spend their time on what interests them, personally, or what someone pays them to spend time on.
Which is why I generally don't ask for support for a particular device, I just look at the list of what is already supported and pick something off the list. In the past when I was just doing this for my own amusement I was picking stuff up at the thrift stores and I'd always check the supported lists from my cell phone when looking at something interesting and cheap. During my recent deployment I looked at the supported list before even going shopping at all. The thing is that OSS has changed mightily in the last 30 years I've used it. When I first got started with it probably the biggest motivation among OSS devs was "sticking it to the man" Incredible amounts of effort were spent reverse engineering stuff like for example the X windows project did to the Nvidia cards for early X support. The feeling was if we broke into/reverse engineered enough stuff then "the man" would end up realizing how futile it was to keep locking us out and then would start cooperating with us. That worked many times - Nvidia quite obviously opened their specs because of this kind of pressure. But that was then, this is now. Now, the devs have "sold out" meaning very few are interested anymore in sticking it to the man and reverse engineering stuff. The dev who did the MR33 crack is one of the old school ones. He eventually had to reach for the soldering station to hardware break in. I bow to that level of skill and when I grow up I want to be like him. But almost all of the rest - just not interested anymore. Oh, they still say they want to stick it to the man and are entranced by the idea of sticking it to the man, but when it comes to the actual work of doing that? Cricket chirps. It's just easier to fork over an extra hundred bucks for a hardware device you can just follow the chipset makers released documentation for. The entire point of the GPL license is sticking it to the man. But today? Most project devs just want to get in bed with the man. 30 years ago a proposal like OpenWRT One would have been viewed as a complete sell-out. Everyone would be saying why are you spending time sucking up to MediaTek when we need to kick Broadcom in the ass and teach them some manners. But today? Completely the opposite. Buying a Raspberry Pi for your wifi router is probably the most practical way to run OpenWRT if your goal is to have maximum power maximum support on a power-sipping device. And you can even download case plans for your 3d printer for it. But, when you do that - you belong to "the man" Raspberry Pi Holdings made $250M last year. By running OpenWRT on it - you ARE the mainstream, you have sold out. Why are you even wasting your time? Might as well buy a complete device from Netgear and use their firmware like all the other sell-outs. You THINK you are doing something new and original and "cool" but your not - your just another lemming. When the devs turn their back on this ideal - which they almost all did years ago - it just sucks all the fun out of OSS. Your mainstream just like the people you claim to be fighting the good fight against. And the OpenWRT project isn't the only big OSS project this has happened to. Red Hat for example is so much "the man" these days that there's distros like Alma and Rocky that their entire point it sticking it to "the redhat man" Never thought I'd see the day 30 years ago when arms of OSS would regard other arms of OSS as sell outs and attack them, but here we are. Ted PS I don't have an answer for this, BTW. -----Original Message----- From: PLUG <[email protected]> On Behalf Of Russell Senior Sent: Monday, January 12, 2026 1:43 PM To: Portland Linux/Unix Group <[email protected]> Subject: Re: [PLUG] Router recommendations for pending Ziply Fiber upgrade "Whether that decommissioned enterprise gear is useful or not depends on the device and whether firmware replacement is practical. Sometimes it is, sometimes it isn't. As ever, this is a function of opportunistic and motivated reverse engineering." Obviously, there is a time when something isn't supported, which moves towards being supported, sometimes at zero miles per hour, possibly forever, most particularly if you waiting for someone else to do the work. I think you misunderstand what determines the things OpenWrt developers spend their time on. Almost all of the enthusiasm for supporting new devices comes from the outside, by the people who are likely to benefit. Take a look at: git log --pretty=format:"%h [%an %ae] %s%n" | grep 'add support' | less in the openwrt git history and the vast majority of the device support is coming from people outside the core developers. -- Russell Senior [email protected] On Mon, Jan 12, 2026 at 1:35 AM Ted Mittelstaedt <[email protected]> wrote: > > Been around for the MR33 for a while: > > https://sagacioussuricata.com/posts/meraki-mr33/ > > More discussion on this here: > > https://watchmysys.com/blog/2024/04/breaking-secure-boot-on-the-meraki > -z3-and-meraki-go-gx20/ > > And from here: > > https://openwrt.org/toh/meraki/z3 > > "... Thus, the only way to install OpenWrt is to physically remove the TSOP48 > NAND chip (U30 on the PCB) and flash it in an external programmer. You will > also need to reprogram the AT24 (24c64) EEPROM (U32 on the PCB)..." > > Now, granted, the Z3 isn't wifi6, it's 5. So even though the Z3's are > cheaply available on Ebay, etc. - they aren't worth the money to fuss with. > Unless your just wanting to get some cheap kit to learn on with your prom > programmer. > > There's also a long discussion thread on OpenWRT on the OpenWRT forum on the > Meraki MR46 which is an 802.11AX chip. Pretty much that one will end up like > the Z3 - unsolder chip, reprogram it, resolder it. > > Obviously, there's other devices out there that have secure boot bypasses > using that chipset - such as the Netgear WAX 220 - but it's going to be years > before these wifi6 devices are available cheaply in large quantity the way > that the Aps like the Luma, Meraki, etc. which fail to function if the > subscription is not paid, are. > > Buying 40 Meraki MR-52's for under $20 each to populate out a building with > wifi 5 - quite doable. > > Buying 40 Netgear WAX 220's for under $20 each? Not happening. > > Buying 40 Meraki MR-46's for under $20 each to populate out a building with > wifi 6? Well, not doable NOW since that device is currently sold - you can > get used ones as cheap as $75 - but the moment Cisco EOLS them the price on > them will crash on the used market. > > I simply don't see price crashing on the stuff sold to consumers like the > Netgear WAX stuff. And you won't see auctions like "lot of 20 WAX for $100" > the way you see for the Meraki stuff because it's generally 20 different > people buying those 20 WAXs while it's ONE org buying 20 or 200 of the > MR-46's. And when the 20 different people decide to get rid of WAX 220s, it > will be 20 different times over 2-3 years while the BigCo will just dump all > of them at once. > > This is why I spoke against the OpenWRT One on the forum although doing so > got me a lot of flak. > > With the used Enterprise gear, someone else has paid the depreciation. So, > it's always going to be dirt cheap. I felt instead of OpenWRT developers > sinking time into yet another $100 wifi AP, that they sink time into the > Enterprise gear so that we would have more $20 Aps. But the devs are techs > not businesspeople and to them, $100 is - apparently - nothing. They don't > consider the multiplicative factor of scale.... > > Ted > > -----Original Message----- > From: PLUG <[email protected]> On Behalf Of Russell > Senior > Sent: Saturday, January 10, 2026 6:12 PM > To: Portland Linux/Unix Group <[email protected]> > Subject: Re: [PLUG] Router recommendations for pending Ziply Fiber > upgrade > > On Sat, Jan 10, 2026 at 5:22 PM Ted Mittelstaedt <[email protected]> > wrote: > > > Obviously I want to see OpenWRT ported to the new hardware but the latest > > porting efforts now, too much gear seem to start out with: > > > > Instruction #1 - heat up your hot air soldering station, desolder the PROM, > > put it into your PROM programmer and flash in our fixed boot loader.... > > I've never seen that. Can you point at an example? > > -- > Russell Senior > [email protected] >
