Community backed router

Is this something you'd be interested in?


  • Total voters
    21

TheLostSwede

Senior Member
I wanted to gauge the interest of the community here, to see if there are enough people that would want to get involved in a project. I’ve noticed a lot of people here are looking for routers that have features that you normally don’t find from the major consumer router makers, some of them hardware based, some of them software based.

I recently came across what should be a very high-end router, based on the QCA IPQ8072A SoC, with two 10Gbps ports, 1GB of RAM and even an M.2 slot for an optional 5G modem.
The router is apparently intended as a carrier device by the company that makes it and I have in all fairness not approached them as yet, since I wanted to see if there was any interest from the community here to try and source this product, get a “custom” OS up and running on it and have it supported by the community here or at least get it supported by OpenWRT/DD-WRT if possible.

br-6889ax.jpg


I obviously don’t know the pricing, but I would guess it would be cheaper than some of the branded routers in the market based on the same SoC. Normally the MoQ for something like this is one to 10k units, so that might be another issue. On top of that, there’s no word on what kind of state the software support is in, but as I have worked for a router manufacturer in the past, I do at least know some people that could get it to a fully working state, for a cost.

Obviously this is just a crazy idea of mine at this point, but with enough interest, it could be pulled off. I know not everyone is keen on OpenWRT as an OS, but it’s apparently not a huge effort to make a custom UI for it and the people I have worked with in the past, have already done it for different products. The main advantage of going down the OpenWRT route is that there will be regular updates that wouldn’t require a huge effort for a project like this. Other hardware can also be considered if price is a more important factor to those interested and I’m happy to look for alternatives, if that would prove to be a more viable route.

So please, thoughts, comments, feedback on my crazy idea.
 
Last edited:

AndreiV

Very Senior Member
Turris Omnia and Mox , the project is already there.

New Omnia model on its way soon :

Omnia 2022
 

TheLostSwede

Senior Member
Turris Omnia and Mox , the project is already there.

New Omnia model on its way soon :

Omnia 2022
Turris didn't exactly do a great job with their first couple of iterations and from what I have read, a lot of people regret investing in their hardware. Just reading the comment at the top of the forum page you linked to, isn't exactly a vote of confidence in their favour, as they clearly aren't interested in making their new version of "their" OS work on their original, expensive router, despite there being no technical limitations as such.

The Mox was, a pile of junk imho. It was an interesting concept that was built around the wrong hardware and not only had a lot of limitations that the average person were never going to be able to figure out, but due to different availability of their add-on bits in different markets, weren't a viable option even for those that could work it out.

This wouldn't be a custom hardware project, as it's expensive making hardware, as you have to pay for NRE's, moulds, certifications etc. and it doesn't make sense for something like this. This will be based around an already developed and selling piece of hardware that is normally sold to service providers.
 

AndreiV

Very Senior Member
ust reading the comment at the top of the forum page you linked to, isn't exactly a vote of confidence in their favour, as they clearly aren't interested in making their new version of "their" OS work on their original, expensive router, despite there being no technical limitations as such.

You should read more slowly, it says the exact opposite to your assertion. Just like ASUS/Merlin ect. they are moving on from the older version and migrating users to the latest.

Obviously you have zero real experience of the products, everyone I know that took up Omnia is extremely happy with their purchase and the router evolves with time. My kit is very reliable and stable , I definitely wouldn't ever buy another ASUS or similar router.

Mox is not junk and works very well , it suits the needs of those that want to buy only what they need instead of having endless bloated , pointless features installed.

You asked for opinions and then don't like what you get, but then you did say your idea was crazy.
 

TheLostSwede

Senior Member
You should read more slowly, it says the exact opposite to your assertion. Just like ASUS/Merlin ect. they are moving on from the older version and migrating users to the latest.

Obviously you have zero real experience of the products, everyone I know that took up Omnia is extremely happy with their purchase and the router evolves with time. My kit is very reliable and stable , I definitely wouldn't ever buy another ASUS or similar router.

Mox is not junk and works very well , it suits the needs of those that want to buy only what they need instead of having endless bloated , pointless features installed.

You asked for opinions and then don't like what you get, but then you did say your idea was crazy.
Or not, since only the units that run BTRFS will be updated according to that message, which seemingly doesn't apply to all of them.

I'll admit I don't own one of their products, but I can read the feedback and what I have read, has been anything but amazing. That said, I'm sure there equally are people that are happy with their purchases, but I guess it comes down to the expectations that people had vs. what was delivered and I followed both crowdfunding campaigns and based on those, both times they over promised or under delivered.

The Mox products is running on an obsolete, slow SoC, although the fact that it was slow was never a secret, but Marvell has discontinued it, alongside all of its products in that product family, so not exactly a great product when it comes to long term support.

Sorry, but your opinion was, buy this instead. That's not an opinion about what I'm asking about, it's you telling people that someone already made something you bought, that you think is great and that everyone should also buy. It doesn't answer the question I posted, it is your opinion about something entirely different and clearly not relevant.

So far, you haven't posted a single thing related to my original post here, so I would say your replies are the ones that are a bit crazy if anything. So thanks for nothing.
 

thiggins

Mr. Easy
Staff member
If this is going to be the tone of the discussion, then I'm going to lock the thread.
Listen with intent to understand, not to respond.
I’ve noticed a lot of people here are looking for routers that have features that you normally don’t find from the major consumer router makers, some of them hardware based, some of them software based.
This seems to be the primary reason for the project. @TheLostSwede Do you have your own list of "features that you normally don’t find from the major consumer router makers"? Putting your list of desired features out there may draw more specific feedback.

@AndreiV Can you say how Turris Omnia and Mox have made you swear off commercial routers like ASUS? Your reason sounded more aimed at reliability due to a smaller feature set.
 

Tech9

Part of the Furniture
I believe the best community router is actually software only - hardware independent or runs on readily available hardware. This is what projects like DD-WRT, Tomato, OpenWRT, Asuswrt-Merlin are, this is what I asked @dave14305 (OpenWRT on RPi) and @john9527 (OpenWRT on x86) to share with the community, if possible. Binding the software to specific hardware is not a good idea.
 

sfx2000

Part of the Furniture
I wanted to gauge the interest of the community here, to see if there are enough people that would want to get involved in a project. I’ve noticed a lot of people here are looking for routers that have features that you normally don’t find from the major consumer router makers, some of them hardware based, some of them software based.

Interesting concept - I've explored this path a couple of times - even taking it one step further by actually implementing a project with both HW and SW.

Hardware is really hard, and expensive to do if starting from scratch... even with "connections" in the industry, going from schematics to layouts to initial production is a slow and time consuming effort.

SW is pretty much the same - in my case, we probably went well beyond what was "current" state of the art at the time - 1GB DDR/8GB eMMC on a dual-core Marvell A53 chipset - and on the SW side, a solution similar to Alpine Linux, but for ARM, and containers for applications - the BSP was really looking to support anything from a Router to a NAS to a Set Top Box - initial product was a Router.

HW and SW were the easy part - the challenge was the business and legal side - working with partners, getting licensing sorted out, looking at the commercials into the channel (do ODM work for someone else, or OEM it ourselves, etc).

Ultimately at the end of the day, once we have a MVP (minimum viable product), we took a second look at the market, and the numbers didn't make sense for OEM, so we looking to sell/license the design out - to that end, I did manage to exit with an OD deal to another OEM/ODM and break-even for all of the NRE spent upfront, pay off the investors, and get a license/royalty deal for a period of time.

Not something to do lightly, at least from my lessons learned...

I recently came across what should be a very high-end router, based on the QCA IPQ8072A SoC, with two 10Gbps ports, 1GB of RAM and even an M.2 slot for an optional 5G modem.
The router is apparently intended as a carrier device by the company that makes it and I have in all fairness not approached them as yet, since I wanted to see if there was any interest from the community here to try and source this product, get a “custom” OS up and running on it and have it supported by the community here or at least get it supported by OpenWRT/DD-WRT if possible.

The main challenge from the SW perspective is to get the agreement in place with Qualcomm, and leverage into the QSDK - QSDK is a better approach than going with OpenWRT (nothing against OpenWRT, but to make the most of the chipsets, one needs the 'special sauce' that the closed source drivers for WiFi and the NSS subsystems can offer.

The QCA drops over on CodeAurora are partials - to get the full BSP, and supporting documentation, along with support - that means establishing a business relationship with Qualcomm and getting the NDA's and other legal docs done - and typically these days, Qualcomm needs a fairly significant volume commitment to even get their attention - we're talking 500K to 1M pieces usually.

And that means making a business and getting that B2B relationship going - Qualcomm generally doesn't support community efforts - to them, it's more work than it is worth as they have the OEM/ODM contracts in place with the established players in the industry that have a proven track record.

Not to say that it couldn't happen - just that it's much bigger than it initially looks like as a thought experiment.

Side note - The HW you found - it looks similar to an ASKEY design, at least from the mechanical ID...
 

TheLostSwede

Senior Member
This seems to be the primary reason for the project. @TheLostSwede Do you have your own list of "features that you normally don’t find from the major consumer router makers"? Putting your list of desired features out there may draw more specific feedback.
From what I have seen here, people want at least two, faster than Gigabit ports, up to date software, especially with regards to security related features, the ability to add software features, least not various ad blocking and VPN type options that aren't supported without various "hacks".
This wasn't so much about what I personally want, but rather if there was enough interest to make something that would be more user accessible.

Apologies for the way the first few comments went, that wasn't my intention.
 

TheLostSwede

Senior Member
I believe the best community router is actually software only - hardware independent or runs on readily available hardware. This is what projects like DD-WRT, Tomato, OpenWRT, Asuswrt-Merlin are, this is what I asked @dave14305 (OpenWRT on RPi) and @john9527 (OpenWRT on x86) to share with the community, if possible. Binding the software to specific hardware is not a good idea.
I have no intention to reinvent the wheel here, hence why I suggested using OpenWRT or DD-WRT, although for OpenWRT I would like to see a better UI/UX, as I find their standard one less than user friendly.
The reason I decided to ask, was because I found what appears to be some decent hardware, which imho the RPi is not and x86 isn't really suitable, mainly due to lack of drivers for various WiFi radios. Putting client radios in a router isn't a good idea imho.
 

TheLostSwede

Senior Member
Interesting concept - I've explored this path a couple of times - even taking it one step further by actually implementing a project with both HW and SW.

Hardware is really hard, and expensive to do if starting from scratch... even with "connections" in the industry, going from schematics to layouts to initial production is a slow and time consuming effort.

SW is pretty much the same - in my case, we probably went well beyond what was "current" state of the art at the time - 1GB DDR/8GB eMMC on a dual-core Marvell A53 chipset - and on the SW side, a solution similar to Alpine Linux, but for ARM, and containers for applications - the BSP was really looking to support anything from a Router to a NAS to a Set Top Box - initial product was a Router.

HW and SW were the easy part - the challenge was the business and legal side - working with partners, getting licensing sorted out, looking at the commercials into the channel (do ODM work for someone else, or OEM it ourselves, etc).

Ultimately at the end of the day, once we have a MVP (minimum viable product), we took a second look at the market, and the numbers didn't make sense for OEM, so we looking to sell/license the design out - to that end, I did manage to exit with an OD deal to another OEM/ODM and break-even for all of the NRE spent upfront, pay off the investors, and get a license/royalty deal for a period of time.

Not something to do lightly, at least from my lessons learned...


The main challenge from the SW perspective is to get the agreement in place with Qualcomm, and leverage into the QSDK - QSDK is a better approach than going with OpenWRT (nothing against OpenWRT, but to make the most of the chipsets, one needs the 'special sauce' that the closed source drivers for WiFi and the NSS subsystems can offer.

The QCA drops over on CodeAurora are partials - to get the full BSP, and supporting documentation, along with support - that means establishing a business relationship with Qualcomm and getting the NDA's and other legal docs done - and typically these days, Qualcomm needs a fairly significant volume commitment to even get their attention - we're talking 500K to 1M pieces usually.

And that means making a business and getting that B2B relationship going - Qualcomm generally doesn't support community efforts - to them, it's more work than it is worth as they have the OEM/ODM contracts in place with the established players in the industry that have a proven track record.

Not to say that it couldn't happen - just that it's much bigger than it initially looks like as a thought experiment.

Side note - The HW you found - it looks similar to an ASKEY design, at least from the mechanical ID...

Not my first rodeo as you know, it's a shame your project wasn't viable to make yourself, it sounds like it could've been a big hit.

I'm obviously aware of the limitations when it comes to the driver bits, all the router chip makers have the same limitations. That said, since this is an already made piece of hardware, it should be possible to get access to the software via the OEM, depending on their relationship and contractual agreements with QCA. I don't think any of the router chip makers are asking for those kind of volume commitments though, unless things have really changed a lot in the past five years.

Not an Askey product, but I have perused their hardware options too. They seem to be making nicer designs these days, but very complex designs at the same time. It's from Enermax's OEM/ODM business.
 

AndreiV

Very Senior Member
@AndreiV Can you say how Turris Omnia and Mox have made you swear off commercial routers like ASUS? Your reason sounded more aimed at reliability due to a smaller feature set.
Yes reliability and stability (and better security) but a smaller feature set, no.

It is about users CHOICE of whatever they run, not having baked in "features" that are not wanted by many and at best are not really any use at all.

My choice of router has open source software and hardware. I can open the case and add other hardware /sim cards and storage if I want.

I have everything I need but at the touch of a button can install features/apps from a list of @8000 packages.

Since changing router my LAN and WAN speeds are improved, Windows shares work perfectly unlike the constant issues when connected to an ASUS router and it runs perfectly without constant need for tweaks and changes.

Reality is I have more features available than an ASUS can provide but I am not forced to have them installed /running if I don't want them.
 

AndreiV

Very Senior Member
Sorry, but your opinion was, buy this instead

No, I was trying to point out that there is already a project that could be made use of.

As to people complaining about the Omnia/CZ products on their forum, take a look at ALL other providers forums, where will you always see the complaints/issues/faults listed? SNB has seen the comment " I was thinking of buying an ASUS router but having read the forums I see too many problems " many times a year.
 

Tech9

Part of the Furniture
The reason I decided to ask, was because I found what appears to be some decent hardware

As router/firewall it can't compete with already existing x86 options with OPNSense/pfSense. DIY guys will beat it in both price and performance.
 

RMerlin

Asuswrt-Merlin dev
These days more often than not, what makes a great product isn't the hardware, but the software stack. Hardware is just a platform to run the software.

Sadly, a lot of companies haven't realized that yet.
 

L&LD

Part of the Furniture
And a lot of people miss realizing that too.
 

TheLostSwede

Senior Member
As router/firewall it can't compete with already existing x86 options with OPNSense/pfSense. DIY guys will beat it in both price and performance.
And x86 hardware can't compete on the wireless side. I also doubt they can compete on price, knowing what these things really cost, without the brand name tax.
 
Last edited:

Tech9

Part of the Furniture
What’s the price for 1 unit and how long I need to wait for working software?
 

L&LD

Part of the Furniture
Once again, putting the cart before the horse. ZVP
 

Tech9

Part of the Furniture

Latest threads

Sign Up For SNBForums Daily Digest

Get an update of what's new every day delivered to your mailbox. Sign up here!
Top