5 ms·

I'd love to work with people who want to find ways to build low cost 802.11ax mesh implementations. Think high powered raspberry pis (or their equivalents) bein

I'd love to work with people who want to find ways to build low cost 802.11ax mesh implementations. Think high powered raspberry pis (or their equivalents) being customized to compete with the crazy high costs of Orbi implementations. And I absolutely have zero interest in the monetary value here. RPi + wifi shields + fun is what counts. If you're interested, I'd love to team up! And I bring all kinds of nerdy engineering skills to boot and I'd be doing this for fun with other engineers. The goal would probably be an implementation that would one day show up on hacker news as a post for other folks to build and enjoy!

310260 5y ago
I'd be interested in that! Would be a good learning experience and a way to understand WiFi more deeply. My experience so far is mostly in cellular.
- aliljet 5y ago
  I'm convinced every side project should lend you awesome learning experiences. Why else would you do it? These specs and the chipsets that are implementing them seem to be deeply maddening, but I bet that's a good thing, right? :)
  - 310260 5y ago
    >These specs and the chipsets that are implementing them seem to be deeply maddening Oh yes. Much to the frustration of even the phone OEMs. Hopefully WiFi is easier to handle than 5G lol
spmurrayzzz 5y ago
I'd love to follow along once you get this started. Sounds like a lot of fun. I'm personally limited on cycles currently (N.B. I am a cofounder / direct of eng. of an ISP that builds custom 802.11ax-over-mmWave last mile access networks and CPEs in-house), but since I have enough experience in that world I can probably jump in at points to help as I free up.
- aliljet 5y ago
  I'd bet you'd be a fantastic person to lend cycles to keep the guesswork within reason. And if you had cycles to put to the weeds, I doubt an engineer would turn you away. We should get in touch! pavgup at gmail dot com!
Tepix 5y ago
I'm intrigued by mesh networking, too. I'd like to see an open source implementation of Bluetooth Mesh networking for embedded SoCs like ESP32 that's not limited to a single chip (and ideally usable with MicroPython). This could make cool games possible running on all those hacker conference badges!
- aliljet 5y ago
  Years ago now, I was building baby monitoring tools, and putting BLE implemnentations into place on ESP8266s might have been worse than smoking cigarettes. I swear, my life was shortened, but I love the idea of device-to-device mesh systems. These 802.11ax systems are basically intelligent wireless backhaul systems, the devices that actually connect to the backbone could be boring, but it'd be so cool if devices helped to expand the mesh...
mch82 5y ago
This is a great project. Orbi systems are so expensive. Are you basing your work on any open source mesh projects? Do you have a project site or GitHub where people can follow along?
- aliljet 5y ago
  Great question. The *WRT (OpenWRT primarily) communities are clearly the winning universe around this and we should take (with generous attribution) what we can from that world, but I've yet to see a real DIY hardware implementation. These Orbis are honestly very powerful individual routers, so I suspect there's a real thing around their hardware, but I bet DIY can find a way... what do you think?
  - Nextgrid 5y ago
    OpenWRT is missing a big piece of the puzzle: configuration management and the ability to work with a "controller". OpenWRT is currently great at running stand-alone but has essentially zero support for being part of a "fleet" of devices managed centrally. This means something as simple as changing the network name or password requires changing it on every single access point manually, and even worse if your mesh system relies on sharing frequently-changing state between devices. OpenWISP tries to address this problem: https://openwisp.org https://openwisp.org - I suggest you check it out and solve the configuration management problem first. The actual "mesh" part is actually relatively easy. Most commercial systems use basic Linux networking tools, HostAPd (sometimes with custom improvements, but this all ends up upstreamed or reimplemented upstream given enough time) and custom glue code to tie them together. A "mesh" system is typically a user-facing network being broadcast by all APs (with shared settings such as name and password) and an invisible, "backhaul" network each AP hosts (either on a separate interface or on the same interface as the AP - I believe some wireless cards can act both as AP and station as long as the channel is the same) and the other in the path connects to, and the glue code handles configuring all of that. 802.11s is also an option that can be used, and I'm pretty sure all of this is already possible to configure manually in Linux - what's lacking is the "glue code" to set up & manage all of this automatically.
  - mch82 5y ago
m348e912 5y ago
Please reach out to Eric, the founder of wakoma.co He is working on this exact thing and I'm certain he would love to collaborate.
- aliljet 5y ago
  Very curious. At first glance, it's not clear he's targetting the Wifi 6 (or Wifi 6e) mesh implementations. If he's already running something, he's the place to send collaborators that are interested here. The goal here would be to compete with Netgear's Orbis with DIY low-cost hardware...
atdrummond 5y ago
I'm interested in the hardware side of this but also the software side - building a resilient, E2E messaging system not reliant on cell networks or centralized national inter/intranets to work. Is the latter too off-base for it to be worthwhile for me to reach out?
- krasi0 5y ago
  I am surprised that there have been no responses to your question as this would be like the holly grail for people interested in privacy oriented and robust peer to peer messaging. I wouldn't be able to add much know how to the technical aspects unfortunately.
Artistry121 5y ago
I have a finance / operations / marketing background and could probably help find a use for this at scale that can get it out. Even if not for the $$$ this could help get the tech out and build solid use cases. Would be fun to work on.
desertraven 5y ago
It looks like from the responses here, we could at the very least create a community?
- aliljet 5y ago
  Unquestionably. I started this conversation, I won't let it just hang. There's something seductive about building things that best at lower cost the available best consumption thing. We will find a way!

310260 5y ago

I'd be interested in that! Would be a good learning experience and a way to understand WiFi more deeply. My experience so far is mostly in cellular.

aliljet 5y ago
I'm convinced every side project should lend you awesome learning experiences. Why else would you do it? These specs and the chipsets that are implementing them seem to be deeply maddening, but I bet that's a good thing, right? :)
- 310260 5y ago
  >These specs and the chipsets that are implementing them seem to be deeply maddening Oh yes. Much to the frustration of even the phone OEMs. Hopefully WiFi is easier to handle than 5G lol

spmurrayzzz 5y ago

I'd love to follow along once you get this started. Sounds like a lot of fun. I'm personally limited on cycles currently (N.B. I am a cofounder / direct of eng. of an ISP that builds custom 802.11ax-over-mmWave last mile access networks and CPEs in-house), but since I have enough experience in that world I can probably jump in at points to help as I free up.

aliljet 5y ago
I'd bet you'd be a fantastic person to lend cycles to keep the guesswork within reason. And if you had cycles to put to the weeds, I doubt an engineer would turn you away. We should get in touch! pavgup at gmail dot com!

Tepix 5y ago

I'm intrigued by mesh networking, too. I'd like to see an open source implementation of Bluetooth Mesh networking for embedded SoCs like ESP32 that's not limited to a single chip (and ideally usable with MicroPython). This could make cool games possible running on all those hacker conference badges!

aliljet 5y ago
Years ago now, I was building baby monitoring tools, and putting BLE implemnentations into place on ESP8266s might have been worse than smoking cigarettes. I swear, my life was shortened, but I love the idea of device-to-device mesh systems. These 802.11ax systems are basically intelligent wireless backhaul systems, the devices that actually connect to the backbone could be boring, but it'd be so cool if devices helped to expand the mesh...

mch82 5y ago

This is a great project. Orbi systems are so expensive. Are you basing your work on any open source mesh projects? Do you have a project site or GitHub where people can follow along?

aliljet 5y ago
Great question. The *WRT (OpenWRT primarily) communities are clearly the winning universe around this and we should take (with generous attribution) what we can from that world, but I've yet to see a real DIY hardware implementation. These Orbis are honestly very powerful individual routers, so I suspect there's a real thing around their hardware, but I bet DIY can find a way... what do you think?
- Nextgrid 5y ago
  OpenWRT is missing a big piece of the puzzle: configuration management and the ability to work with a "controller". OpenWRT is currently great at running stand-alone but has essentially zero support for being part of a "fleet" of devices managed centrally. This means something as simple as changing the network name or password requires changing it on every single access point manually, and even worse if your mesh system relies on sharing frequently-changing state between devices. OpenWISP tries to address this problem: https://openwisp.org https://openwisp.org - I suggest you check it out and solve the configuration management problem first. The actual "mesh" part is actually relatively easy. Most commercial systems use basic Linux networking tools, HostAPd (sometimes with custom improvements, but this all ends up upstreamed or reimplemented upstream given enough time) and custom glue code to tie them together. A "mesh" system is typically a user-facing network being broadcast by all APs (with shared settings such as name and password) and an invisible, "backhaul" network each AP hosts (either on a separate interface or on the same interface as the AP - I believe some wireless cards can act both as AP and station as long as the channel is the same) and the other in the path connects to, and the glue code handles configuring all of that. 802.11s is also an option that can be used, and I'm pretty sure all of this is already possible to configure manually in Linux - what's lacking is the "glue code" to set up & manage all of this automatically.
- mch82 5y ago

m348e912 5y ago

Please reach out to Eric, the founder of wakoma.co He is working on this exact thing and I'm certain he would love to collaborate.

aliljet 5y ago
Very curious. At first glance, it's not clear he's targetting the Wifi 6 (or Wifi 6e) mesh implementations. If he's already running something, he's the place to send collaborators that are interested here. The goal here would be to compete with Netgear's Orbis with DIY low-cost hardware...

atdrummond 5y ago

I'm interested in the hardware side of this but also the software side - building a resilient, E2E messaging system not reliant on cell networks or centralized national inter/intranets to work. Is the latter too off-base for it to be worthwhile for me to reach out?

krasi0 5y ago
I am surprised that there have been no responses to your question as this would be like the holly grail for people interested in privacy oriented and robust peer to peer messaging. I wouldn't be able to add much know how to the technical aspects unfortunately.

Artistry121 5y ago

I have a finance / operations / marketing background and could probably help find a use for this at scale that can get it out. Even if not for the $$$ this could help get the tech out and build solid use cases. Would be fun to work on.

desertraven 5y ago

It looks like from the responses here, we could at the very least create a community?

aliljet 5y ago
Unquestionably. I started this conversation, I won't let it just hang. There's something seductive about building things that best at lower cost the available best consumption thing. We will find a way!