Making the OpenWrt One
In a keynote on the final day of SCALE 22x, Denver
Gingerich said that he wanted to talk "a little bit about a router and
also the big picture around that router
". Gingerich is the director of
compliance at the Software Freedom
Conservancy (SFC), which is the organization behind the OpenWrt One router that
LWN looked at back in November. The
router is, of course, based on firmware from the
OpenWrt project, which got its
start because of GPL-enforcement activities and is a member project at the SFC.
He started by asking the question: "Why are we here?
" He did not
mean the much larger, existential question, but was focused on SCALE.
People came to the conference to learn and talk about "cool
technology
". The accompanying image (from his slides; YouTube video is also available) was of the
printing press, which was a technology that came about a few hundred years
ago and changed many things. One of those things was "power; a lot of
people who didn't previously have power
" gained a lot because they
could read and learn things they previously could not.
Power
Shifts in technology lead to shifts in power, but power is difficult to quantify, which means it is tricky to optimize. On an individual level, having power means being able to control one's life, which often translates to being able to control one's technology, he said. For devices, that largely comes down to whether an owner can repair it, modify it, and install new code on it.
Gingerich went through some of the history of devices, noting that early PCs allowed new software, including operating systems, to be installed on them. Some more recent systems, including iPhones and some Android devices, do not allow installing alternative systems. Others, such as the Nokia N900, were created using FOSS for the most part and were explicitly designed to allow alternative systems to be installed by users. That kind of control gives users agency, thus power, to customize their technology for their needs.
A more recent example of "computers" is televisions. His slide showed a Vizio TV that was installing an update, though not one chosen by the owner, but one forced on them by the manufacturer. One problem with the firmware that TV vendors provide is that it spies on the users' viewing habits in order to sell that data onward. That, like other examples of devices that remove power from users, is an example of unethical technology, he said.
Because most of our devices get their functionality from software, the path to ethical technology is via free software. The SFC has multiple initiatives on that front that are trying to ensure that users have the freedom they need. He said that he would be talking about three of those initiatives, starting with the most recent: building hardware.
Hardware
The OpenWrt One is a wireless router; it was ubiquitous at SCALE because
the SFC was the
network sponsor for the conference. It has a 2.5Gbps
Ethernet port with power over Ethernet and a USB C connector for power,
rather than needing some custom cable, along with "a lot of
other neat features
". Inside, there is an M.2 connector for adding
storage and a mikroBUS
connector for adding peripherals.
There are a number of reasons that the SFC decided to build hardware,
starting with a statement that wireless-router maker TP-LINK posted on its
web site. The company decided that an FCC ruling gave it the ability to
"claim that you could not install what you wanted on the routers that
they sold
". The SFC wanted to see if that interpretation of the ruling
was, in fact, true. The statement he showed was from the Wayback Machine,
but he said that router manufacturers still make that claim in private to
the SFC.
Another reason was that when the organization would receive firmware
releases from vendors, there would often be problems in building the
firmware following the vendors' instructions. "Instructions for how to
compile it would tell us to run make, we would do so and then we would get
an error.
" He showed two examples of those kinds of problems in his
slides, but he could have shown hundreds; "I chose to not bore you with
those today
". So the SFC wanted to see if it was really that hard to
provide a full source release that would actually build.
There are also problems with support life cycles in the router market. He showed a $400 router that was released in 2019, which the maker declared to be end of life in 2023. So after four years, a fairly expensive router was no longer being updated. In addition, there is often a need to get access to a serial port on routers in order to update and debug them, but many manufacturers make it quite difficult to access that port—often requiring modifying the hardware itself. Those anti-features were things that the SFC believed could be fixed in its own hardware device.
Finally, the SFC has noticed that court cases take a long time to
resolve. So, the organization wanted to see if it could "perhaps make a router a little bit faster than it took for a
lawsuit to happen
", he said to audience laughter.
The project started with vote among OpenWrt members on whether to build hardware using the project's name. That was in January 2024 and by "black Friday" (the day after US Thanksgiving) in November, it was released. The initial units sold out in a week, which showed everyone involved in the project that there is a market for these kinds of devices—ones that give owners agency in their hardware.
Results
Gingerich showed the certification that the OpenWrt One received from an FCC
testing lab, which meant that the FCC had no problem with the
device. "We had been telling router manufacturers for a while that this
was not a problem. And now we had the proof.
"
The OpenWrt One also showed that you can provide the software and instructions on how to build and install it on the device. Pointing router makers at that simple set of instructions can perhaps help steer them toward doing something similar for their devices.
In terms of support, there is no specific timeline for the OpenWrt One, but
community support for distributions like Debian and OpenWrt are generally
quite long. Debian still supports the Pentium 4, which came out 25 years
ago and OpenWrt currently supports routers that were released 14 years ago.
He expects the OpenWrt One to have at least ten years of support, if not 20
or 30 years. "We think we should be measuring support life cycles in
decades, not just a few years.
"
The device also solves the serial-port problem. On the front of the OpenWrt One, there is a USB C serial port connector, so no soldering or other modification is required. That was popular with the audience, which cheered and applauded. Meanwhile, it took about 11 months from conception to shipping for hardware; a recent lawsuit that looks to be winding down will end up taking around 47 months, perhaps more. Both approaches are important, however, as people should not have to buy new hardware to get agency in their devices.
There is, unfortunately, still a binary blob or two in the OpenWrt One,
however. These are bits of proprietary code that run on the wireless chip
and the Ethernet controller; there is another blob for RAM calibration as
well. Those are needed because of "power that we don't have yet
".
One way to solve that problem is to find hardware that does not require
blobs; such hardware does exist, at least for now, but only for networking protocols
that are 20 years old. If the features that are provided by that hardware
are sufficient, he suggested buying routers from ThinkPenguin, which has devices
without blobs that are also "very much compliant with the copyleft
licenses
".
The SFC decided to take the approach that it did to try to get new hardware to the same blob-free place. There are two main ways to fix the problem of blobs in today's hardware. The first is to reverse-engineer the existing blobs and replace them with free-software equivalents; the device was designed with the ability to replace the blobs, rather than putting them into an unchangeable ROM. The second mechanism is by gaining market power; selling more free devices like the OpenWrt One will show the part manufacturers that there is a problem that needs to be fixed.
There is still more work to be done to have more devices that will respect our freedoms, Gingerich said. A vote is coming to a close in the OpenWrt community on building an OpenWrt Two. It would be a different kind of device, with two 10Gbps ports, several 2.5Gbps ports, and Wi-Fi 7. That will likely be available in late 2025; the OpenWrt One will still be available once that starts shipping. The SFC is exploring other device classes where it might be sensible for it to make its own hardware.
Member projects
The SFC is also in the business of making "humanity-first software
"
by way of the member projects that it shepherds; "we do a lot of work
for them to make sure they are successful in what they do
". The idea
is that humanity-first software should be putting its users first by giving
them agency. He highlighted some of the "few dozen
" member
projects that call the SFC home.
The Reproducible Builds project is working toward the ability for multiple parties to create bit-for-bit identical binaries from source code. This will be useful for ensuring license compliance among many other uses. OpenWrt is another member project, which is part of why the SFC has worked with it to produce its hardware. OpenWrt came about when Linksys released code for one of its routers after a compliance lawsuit.
The Wine project "allows people to
get incrementally more agency in their computing
" because they can move
to a new operating system without leaving all of their Windows applications
behind. The SFC is also the home for Git, so that it can continue to be "free
and accessible to people and is not controlled by any particular for-profit
entity
". Two other distributed version-control systems, Mercurial and Darcs, are also member projects.
The Etherpad project is a "community
collaboration tool
" that allows collaborative document editing like
that found in Google Docs, "but with a lot more freedom
". The SFC
runs an Etherpad instance that
can be used, or, of course, you can run your own, he said.
Lastly, Gingerich called out Outreachy, which provides internships
in FOSS. There has been a push recently against the importance of diversity, equity, and
inclusion (DEI) initiatives, but the SFC believes that in order to
have computing freedom for everyone, those working on FOSS "need to be
representative of society as a whole
". That was met with a resounding
round of applause.
Lawsuits
The final SFC initiative that he talked about was lawsuits, which he
labeled with a :( frowning face on his slide. The organization is not at
all sad that the legal avenue for enforcing users' rights exists, but it is
unhappy that it has to use it at times. Those rights revolve around the
source code that is covered by licenses like the GPLv2 that is used for
Linux; users have the right to the code and "everything that you need to
compile and install changes onto that device
".
The first thing that companies need to do is to make users aware that they
have those rights, which is done by placing an offer for the source code in
the manuals or other materials accompanying the device. He showed a few
examples, including the offer that is printed on the box of the OpenWrt
One. The latter is different in that it actually encourages users to
obtain and modify the software, but it is also different because it will
truly provide the source code. For the others: "Unfortunately you do
not get the complete source code, you get something incomplete that does
not give you full freedom and agency
".
The GPLv2 is effectively a bargain with the manufacturers of devices: they
get a bunch of code that they can modify and ship on their device, but they
must give all of the rights they got from the software developers to those
who they sell or otherwise distribute the device to. "That is the
bargain; it seems fairly simple to me
", but companies sometimes do not
want to do their part. "They want to alter the deal
" (with a slide
of the classic Star Wars scene), but the license does not allow them
to do so.
There are steps that can be taken when that happens. The first step that the SFC takes is to nicely ask the manufacturer to fix the problems with their source release. So that anyone can check these source releases for themselves, the SFC has created the Use The Source web site, which has candidate source releases along with information about any problems found in them. He showed an entry for the OpenWrt One, which naturally showed that the source could be built and installed into the device. On the other hand, the entry for the Google Nest Hello has a detailed description of all of the different problems that were found in its source release.
After another example or two, he said that he would briefly talk about some
of the lawsuits, though SFC policy fellow Bradley Kuhn was giving a talk
just after that would be covering them in detail. One of the lawsuits was
against AVM as an entry
for its Fritz!Box 4020 router indicates. Sebastian Steck in Germany
bought one of the routers and did not receive the full source code, even after
asking for several years, so he sued with the help of the SFC and won.
The outcome is that users can repair and reinstall the software in those
routers. "It's
very exciting that purchasers of the device have the ability to use the
courts to get their rights.
" That, too, was met with applause.
The SFC bought a Vizio television in 2018, requested the source, and was unable to get it. The organization bought a new TV in 2021, thinking that the problems must have been solved by that point, but the new TV had the same problems, so a lawsuit was filed. The suit is being pursued as a purchaser of the device and is trying to obtain the source, in part to potentially start a project like OpenWrt but for televisions.
There have already been some wins in the Vizio suit, Gingerich said. A
federal court judge agreed that the "cause of action" was legitimate, thus
that purchasers do have the right to bring lawsuits to force companies into
compliance with the licenses. He suggested that those interested could
come to the trial in Santa Ana, California starting on September 15 to
"cheer us on—please do so very silently and respectfully
". The SFC
expects the trial to take a week or more.
If people are interested in bringing their own compliance lawsuit, the SFC
would be happy to coordinate and assist in that effort, so those who want
to take that path should get in touch. There are simpler paths available
as well, including reporting
GPL violations to the organization. Source candidates can be uploaded
to the Use The Source site, which will allow the SFC and others to help
determine if the candidate is complete. People can comment on the
candidates to describe problems that they found, but the hope in the long
term is that people will use the site "to discuss the neat things
they're able to do with all their rights
" by being able to modify their
devices. The site can also be used
to upload pictures of offers for source code, which will allow the SFC to
request the code; he suggested people look in the manuals and menus of
their devices to find such offers.
It makes a big difference what you buy, he said; for example, if you are in
the market for a router, the OpenWrt One is a great choice, with $10 of
each purchase going to the SFC and OpenWrt project. As more of those
device are sold, that increases the power that the SFC has to put
pressure on manufacturers to provide more freedom-respecting parts and
equipment. There are also lots of options for running OpenWrt on existing
routers or ones that can be bought used, which will help reduce the e-waste
problem; he pointed to a list of
supported routers that will run the latest OpenWrt releases for that.
Lastly, "what you don't buy also makes a big difference
"; the SFC is
a non-profit organization that relies on donations to fund its
activities and he, naturally, encouraged people to help out.
[I would like to thank the Linux Foundation, LWN's travel sponsor, for funding to travel to Pasadena for SCALE.]
| Index entries for this article | |
|---|---|
| Conference | Southern California Linux Expo/2025 |
Posted Mar 28, 2025 16:54 UTC (Fri)
by cen (subscriber, #170575)
[Link] (20 responses)
Overall this seems like a successful initiative so kudos to everyone involved.
Posted Mar 28, 2025 17:20 UTC (Fri)
by pizza (subscriber, #46)
[Link] (19 responses)
Is there a definitive source for this? Because the text in the proposal (and all other reporting) says it is expected to be 1x 10Gb SFP, 1x 5GbE, 4x 2.5GbE (switch), and 1-2x 1GbE. (It's not clear if the 2.5Gb and 1Gb ports are on the same switch or are different logical interfaces)
Personally I'd much prefer an additional 10Gb SFP slot instead of that 5GbE port, as it(along with one copper port) would allow me to finally replace the 12-year-old 1U server that has been my home router for the past few years.
(I don't even care about having wifi! But it needs to ahve at least two 10Gb (or faster) SFP slots and 1-2x additional copper ports. )
Posted Mar 28, 2025 19:18 UTC (Fri)
by zorg24 (subscriber, #138982)
[Link] (14 responses)
Posted Mar 29, 2025 1:02 UTC (Sat)
by pizza (subscriber, #46)
[Link] (12 responses)
Thanks, but I've long since learned to avoid trusting anything important to Arm-based SBCs that won't function properly with a mainline kernel and/or boot with off-the-shelf distros. (A special snowflake vendor provided "preinstalled SD card image" doesn't count. Oh, and this includes most Raspberry Pis.)
As I'm not going to be able to wait until the end of the year for OpenWRT to build this new board (assuming it has the dual SFPs that I need), this is what I'll probably end up going with:
https://www.amazon.com/Healuck-Firewall-Appliance-OPNsens...
Posted May 12, 2025 0:49 UTC (Mon)
by pizza (subscriber, #46)
[Link] (11 responses)
> https://www.amazon.com/Healuck-Firewall-Appliance-OPNsens...
I wanted to post a followup. The SoCs these CWWK designs are built on (N100/N150, or N305/N355 on the high end) only have a total of 9 PCIe 3.0 lanes. These lanes are split between a pair of i226 2.5GbE controllers, a slot for nvme storage, a slot for a wifi card, and one or two additional peripherals (second nvme slot, WAN card slot, and/or a third 2.5GbE controller) That means anywhere from 5 to 7 of the possible 9 PCIe lanes are already spoken for, leaving at most 4 (but more likely 2) for the 10GbE ports.
These designs all seem to use an i82559ES dual-port 10GbE controller, which is a PCIe 2.0 device whose documentation states an x8 link is necessary if you are seeking to run both ports at full speed. This means at _best_ (in a x4 setup and 0% overhead) this design provides only 80% of the raw bandwidth necessary for full utilization of both ports, and in an x2 configuration, it won't even be able to run a single interface at full duplex (or both at half duplex).
That's... quite disappointing.
Posted May 12, 2025 1:39 UTC (Mon)
by intelfx (subscriber, #130118)
[Link] (8 responses)
So... another instance of "you get what you pay for"? ;-)
Posted May 12, 2025 13:35 UTC (Mon)
by pizza (subscriber, #46)
[Link] (7 responses)
Generally I'd agree with you but when your headliner feature is so badly kneecapped... it rather defeats the purpose.
FWIW the rest of the system appears to be more than adequate. Variations of this thing exist with more 2.5GbE ports instead of the 10GbE SFPs, which is fine. Other variations exist with even more ports, but they're built on much more capable SoCs with at least *20* PCIe lanes to play with. But none of those seem to be optionable with with 10GbE.
...FWIW, the i82599ES is used instead of something more capable because it's really, really cheap these days -- first released 16 years ago, and as it turns out, was formally EOL'd (order books closed and support formally ended) just seven days ago.
Posted May 12, 2025 21:03 UTC (Mon)
by intelfx (subscriber, #130118)
[Link] (1 responses)
What I wanted to say is that cheap x86 boxes are not *that* better than Arm boxes. You just exchange the drivers issues for other kinds of issues. So if you have an aversion to Arm SBCs and your plan for dealing with this involves buying a cheap no-name x86 thing instead, there is a chance that you might be disappointed.
Posted May 13, 2025 1:54 UTC (Tue)
by pizza (subscriber, #46)
[Link]
I disagree; all of the problem those "cheap x86" systems have (eg underspec'd buses for the peripherals and nonexistant vendor support), "cheap ARM SBCs" also have in spades. The primary advantage for those Arm SBCs is their lower power consumption, but that's balanced by the huge disadvantage of being one-off special snowflakes that rarely move beyond "only works with the vendor's never-updated original pre-installed image".
Still, if those SBCs give "good enough" performance/features/etc that can be an overall win, though one has to consider how long it would take to come out ahead from the power savings.
Posted May 13, 2025 10:38 UTC (Tue)
by farnz (subscriber, #17727)
[Link] (4 responses)
Similarly, at a previous job, we had 2x10G links to our ISP, consisting of a primary and a failover link; if we'd wanted 20G service, we'd have had to have 3 links, two primary and one failover to cover "backhoe fade" between us and our ISP.
If that's why you want 2x10G, then this sort of box is useful; if you need high throughput, it's not so useful.
Posted May 13, 2025 10:45 UTC (Tue)
by paulj (subscriber, #341)
[Link] (1 responses)
You also need to obtain survey maps of where they have physically have placed their fibre, and /verify/ any claims they make about path independence of the fibres. Potentially down to hiring independent surveyors to verify such claims.
A certain large tech company lost connectivity for DC for a while once, discovering in the process their fibre suppliers had lied^Wwere mistaken in their claims about physical independence, when a JCB somewhere took out in 1 go a number of bundles of fibres that were not meant to be anywhere near each other. They significantly increased the level of verification of future supplier's claims after that.
Posted May 13, 2025 10:57 UTC (Tue)
by farnz (subscriber, #17727)
[Link]
If the claims about diverse pathing turned out to be false, that would have been our ISP's problem - they'd have been paying out on a 6 hour SLA while chasing their suppliers to fix it ASAP.
And we'd agreed an SLA payout that was large enough that the business was better off with the Internet link down than with it up; we weren't foolish enough to believe that a "business" service meant it'd be prioritised for repair, but did believe that if we were getting more in SLA payouts than it was costing us to get alternatives (like LTE sticks for everyone), we'd be OK.
Posted May 13, 2025 11:07 UTC (Tue)
by pizza (subscriber, #46)
[Link] (1 responses)
You make a valid point, but I do feel compelled to point out that having redundant 10Gb ISP uplinks but only 2.5Gb of internal network bandwidth seems backwards.
Posted May 13, 2025 11:21 UTC (Tue)
by farnz (subscriber, #17727)
[Link]
The reason it paid for 10G bearers is that change of bearer is a slow process, since it involves taking down a bearer (or running fresh fibre) then replacing kit on both ends, whereas getting faster service is just a software change - and they wanted to be able to upgrade us on-demand to a more expensive 1 Gbit/s or 2 Gbit/s without delay.
This is not an atypical configuration for SME dedicated internet access (as opposed to "business" service on consumer products); fast bearer, slow service on top. You don't need more than 2.5 Gbit/s of internal network when you've got under 2.5 Gbit/s of external network, supplied on 2x 10G ports.
Posted May 12, 2025 9:38 UTC (Mon)
by farnz (subscriber, #17727)
[Link] (1 responses)
You'd be looking for something like the Microchip Switchtec family devices, or the PLX (now Broadcom) PEX family of devices; a cheap design would put the 10G controller, WiFi card slot and WAN card slot behind the switch, so that you can feed 4 PCIe 3.0 lanes to the switch, and have 16 PCIe lanes out (4x PCIe 3.0 for the WAN card slot and WiFi slot, 8x PCIe 2.0 for the 10G controller), and have WiFi card, WAN card and 10G ports compete for the 4 PCIe 3.0 lanes worth of throughput. If you're going overkill, you'd use a switch with more lanes, and have 8 lanes from the host to the switch, with more ports on the other side of the switch.
Posted May 12, 2025 13:51 UTC (Mon)
by pizza (subscriber, #46)
[Link]
Not that I could tell -- And using one would be likely be more expensive (if even possible to fit in that tiny form factor) than just using a PCIe3.x-capable 10GbE controller to begin with.
(That is probably why their devices with more ports use more capable SoCs -- on the lower end, they sport the Pentium 8505 which sports 20 PCIe 4.0 lanes...)
Posted Mar 29, 2025 18:27 UTC (Sat)
by Mook (subscriber, #71173)
[Link]
Posted Mar 31, 2025 7:03 UTC (Mon)
by intelfx (subscriber, #130118)
[Link] (2 responses)
Out of curiosity, why do you need 2× 10Gbps links? Is it for some kind of failover, or do you actually require 20 Gbps of throughput on whatever side of the router (or is it two independent segments, each of which demands 10 Gbps)?
Posted Mar 31, 2025 12:22 UTC (Mon)
by pizza (subscriber, #46)
[Link]
The latter. I have my network partitioned into two major segments (four if you count the redundant/failover ISPs) and I want to be able to route between them at full wire speed. If I trunked the primary VLANs together over a single 10Gbps link into the router, it would effectively halve the maximum bandwidth available between them.
(Runs between buildings are trunked, and if 10Gbps becomes a bottleneck I can light up the redundant fiber pairs. I can't currently justify the expense of >=25Gbps capable switches.)
Posted Mar 31, 2025 16:51 UTC (Mon)
by Cyberax (✭ supporter ✭, #52523)
[Link]
I have a 2Gb uplink to my ISP (that needs an SFP+ module), and I also want to make my NAS available via a 10G connection to multiple clients.
Posted Mar 31, 2025 9:01 UTC (Mon)
by Wol (subscriber, #4433)
[Link]
Cheers,
Posted Mar 29, 2025 19:48 UTC (Sat)
by champtar (subscriber, #128673)
[Link]
Posted Mar 31, 2025 7:08 UTC (Mon)
by intelfx (subscriber, #130118)
[Link]
As indicated in the vote and alluded to in other comments, this thing is going to run on an MT7988, which amounts to 4×1.8GHz Arm Cortex-A73.
So, I guess, the money question is: will it run Cake at 1 Gbps? :-)
Posted Apr 3, 2025 11:29 UTC (Thu)
by farnz (subscriber, #17727)
[Link] (3 responses)
In the long run, I suspect that the SFC's approach will be far more effective than RYF; it's making it possible to fully Free this device in the long run, knowing that it'll not get Freed if in the short run, it's useless, or if Freeing it requires you to replace read-only memories for each test.
Posted Apr 3, 2025 12:23 UTC (Thu)
by gioele (subscriber, #61675)
[Link] (2 responses)
For this strategy to be viable, OpenWRT One will need to remain 1) in production and 2) relevant enough for the time it will take to reverse engineer the blobs to the point they reach a sufficient level of functionality.
I wish OpenWRT and SFC success in their endeavor, but a counterexample could be PC Engines's APU units that have been widely deployed for two decades and underwent practically no hardware changes, yet most of their NICs and WiFi cards (all except one?) do not have free firmwares.
Posted Apr 3, 2025 12:51 UTC (Thu)
by farnz (subscriber, #17727)
[Link]
At least in the case of the OpenWRT One, it's possible to free it; if I made the changes needed to comply with the FSF's RYF program, it is never possible to replace those blobs, ever - you are tied to non-free firmware forever.
Posted Apr 3, 2025 12:57 UTC (Thu)
by pizza (subscriber, #46)
[Link]
As someone who has written low-level radio firmware [1] and done a lot of hardware-level reverse engineering, the odds of this ever happening are vanishingly small.
(For the record, I *strongly* agree with the SFC's attitude towards firmware, and consider the FSF's "RYF" approach to be objectively and morally *wrong*)
[1] 802.11g-era Wifi, Bluetooth, and some custom stuff
Support
Port configuration is ambiguous..
Port configuration is ambiguous..
Port configuration is ambiguous..
Port configuration is ambiguous..
Port configuration is ambiguous..
> I wanted to post a followup <...> That's... quite disappointing.
Port configuration is ambiguous..
Port configuration is ambiguous..
Port configuration is ambiguous..
Note that, depending on use case, that box can be perfectly usable. For example, I have two interfaces between my home server and my switch, not because I need throughout increases (1G is plenty at the moment), but so that I have a redundant link, and when the wires break, I get a message from network monitoring telling me that I've lost redundancy, rather than losing service.
Port configuration is ambiguous..
Port configuration is ambiguous..
We wouldn't have needed to do any of that - the reason for redundancy was not because we wanted it, but because our ISP insisted on it as part of the service (since the service came with a 6 hour SLA, after which they'd be paying out).
Port configuration is ambiguous..
Port configuration is ambiguous..
Our ISP was fairly typical - it was paying for 2x10G bearers to us, to provide a symmetric 500 Mbit/s service on top of those bearers.
Port configuration is ambiguous..
Do any of the designs have a PCIe switch chip involved? The situation you're describing (PCIe 3.0 lanes on the host, PCIe 2.0 lanes on the device) is what switch chips excel at, since you can have 8 lanes of PCIe 3.0 to the host becoming 32 lanes of PCIe 3.0 facing the devices, with the switch chip operating on a per-TLP basis (so 8 lanes of PCIe 2.0 to the device consumes 4 lanes of PCIe 3.0 on the host side).
Port configuration is ambiguous..
Port configuration is ambiguous..
Port configuration is ambiguous..
Port configuration is ambiguous..
Port configuration is ambiguous..
Port configuration is ambiguous..
Port configuration is ambiguous..
Wol
long term support
I ran OpenWrt on a Ubnt rspro from 2011 to 2023, in the end I had to replace it because it was a bit underpowered, and it would have been impossible to source a replacement in case of failure.
OpenWrt Two
I do find a great irony in the fact that the OpenWRT One has chosen to maximize the chances of being fully "Free Software" in the long run by doing something that's directly in opposition to the FSF's RYF program. The SFC is clearly encouraging people to reverse-engineer the blobs, and produce something that's fully Free, rather than shoving the blobs in read-only memory and meeting the FSF's RYF criteria with something that will never be fully Free.
OpenWRT One is not FSF RYF compliant
OpenWRT One is not FSF RYF compliant
On the other hand, it's trivial to make the PC Engines APUs FSF RYF compliant - just add read-only media, and put the blobs on that media, then rearrange things so that the firmware can only be loaded from that read-only media (which could be as trivial as a rewritable flash chip with a track cut so that you can't assert write-enable).
OpenWRT One is not FSF RYF compliant
OpenWRT One is not FSF RYF compliant
