It's one of the most successful routers ever sold and yet network equipment manufacturers are still fighting tooth and nail to keep their devices closed source. It just doesn't make sense to me.
The manufacturers are mostly run by people who were trained in "standard" corporate governance. This includes the ways to protect corporate revenue streams by suppressing (legally, of course) competition, delivering a range of products by producing the top end model and crippling it to sell at a cheaper price point, and repeatedly reducing costs to increase profits in a "race to the bottom".
Until a new set of management philosophies is adopted for teaching, a large number of companies will keep doing the same thing, because in general corporate managers have a lead time associated with them, and we won't run out of the old school ones until 20+ years after philosophies change.
This is an opportunity for anyone who can do things differently, of course.
I see all this as a heartwarming story where a company was forced, with a "trap" set by GPL and its philosophy, to offer people for once a square deal: good hardware, fairly priced, you are free to do with it what you want. All this serves human needs better and the manufacturer could in fact turn a profit.
There is a faint, faint glimmer of hope that this is a peek of the far future of our techno-political-economic system. Of course with very different laws around intellectual property, company governance, customer protection, terms of participating in the market etc. We might be as far from it as the Enlightenment in 1750 (in a world built on overt serfdom and not even fully developed colonialism) was from the year 2000, but still. Makes me feel a teensy bit better about doing the right thing today, just because.
See how drones and 3d printers are evolving; who could fund the firmware development? who could fund the research? but now these cutting edge things are necessarily public and cannot both be a military advantage and cutting edge.
On the other hand this kind of "mistakes" are what is driving the new generation of POSIX clones for IoT untainted by GPL, like Azure RTOS, NuttX, Zephyr, RTOS,...
I'm firmly convinced that if a Chinese maker made a 100% open source keyboard or mouse, they could sell that for $30 instead of $3, and establish a global brand to boot.
Same thing for a lot of hardware, actually. Printers. Scanners. Etc.
http://www.miniware.com.cn/product/ts100-soldering-iron-pack...
An open source soldering iron, original by a Chinese company Miniware. I don't know how good their sales are, but the iron is so good it got that Louis Rossmann praised it (for the price, ofc). And it seems to be very popular (probably not nearly as much as Weller, but hey).
Isnāt that basically Keyboardio? Except itās a San Francisco company selling them for $150; expensive, but still within reason for boutique mechanical keyboards.
That's what gl.inet is (weird name though).
They can't though. Because if it was really open source, then another firm could just sell if for $3. Of course fabrication itself complicates things, but the gist remains.
Your standard bigco manager also believes a whole bunch of FUD about the lack of OSS secrity and what not, but it's 20 years unless upstarts eat their market.
Probably more likely for your average software company than hardware, but I suspect there's an inflection point in cheap hardware.
You can buy a new WRT today that supports FOSS firmware out of the box - https://www.linksys.com/us/wireless-routers/c/wrt-wireless-r...
And yet Linksys (and others) still sell their closed routers as well. One can only concluded that the Open Source support, while important for a niche group, is not enough for market dominance...
I bought one of these (WRT1900ACS) when I was working from home last year. It's good, but not great. Before anyone else buys one of these for their open source "support," you should know that Linksys/Marvell basically threw a buggy open source WiFi driver over the wall, failed to upstream it to the Linux kernel due to issues with the code, and abandoned it.
Although it works fine for my simple purposes, there's a discussion of some of its issues at the end of this PR: https://github.com/openwrt/openwrt/pull/2397
I have one and it works great for me. I'm not a heavy wifi user, mostly I want openwrt.
I've had several openwrt routers. Before this one I had a tp-link wdr4300, then an archer c7. The wrt1900acs has pretty fast, full-featured hardware.
I run firewall + adblock + privoxy + vlans. Because it has a USB port, I've added a USB GPS dongle so it does gps-based ntp time.
At first openwrt was a little daunting, but it has really grown on me.
One great thing about it is that the entire linux distribution is basically read-only, and all changes you make to your machine are in an infrequently-written overlay filesystem. If you back up /overlay/upper you will have all your config changes in a small tarball. All operations that do continuous writing like logfiles go to ramdisk, so it's easy on the flash and reliable during power failure.
Another thing is that if you follow the instructions, it's actually pretty straightforward to build openwrt for your specific configuration. I cut out the package manager and compiled everything I wanted into my image (or out of it, I turned off ipv6)
With a simple setup, you don't even have to bother with the gui. The config files are pretty simple and you can edit them directly.
I've also put openwrt on some network switches and once I got vlans going, my network got a lot more manageable.
I have a vlans:
- normal - machines can route to internet
- restricted - machines boot and have local dns - can get out (updates) only through the proxy
- test vlan - can't get to anything
the network switches are mikrotik and also running openwrt.
I have retired a rb750gl and rb2011ils, and now everything runs on a rb2011uias and a rb3011uias-rm (11 port)
I love the rb3011 - the rack mount tabs can be rotated 90 degrees and you can attach it under a shelf.
The two switches have SFP, and I can't help but think I should start messing with fiber.
Linksys is owned by Cisco, and I don't know what they do now, but at the time a Cisco low-end router had no specialty hardware to run a lot of their features. Those features were implemented in software.
So openwrt threatens their entry level and some of their mid-range devices, creating a conflict of interests.
Cisco hasnāt owned Linksys since 2013. Belkin bought it from Cisco and kept the brand.
Linksys is currently owned by Foxconn.
https://www.theverge.com/2018/3/26/17166272/foxconn-buys-bel...
I had a WRT1900AC for several years. It was a very nice product, with very good community support.
Official support, however, was not good in my experience. Several years later I finally bought a Ubiqui Dream Machine Pro, and absolutely love it. Kinda miffed that they suffered a breach a month after I bought it, though.
I recently sold my UDMP and bought some mikrotik gear, because the device hat very tight limits on what ubiquiti wants you to do with it. No wireguard was an annoyance I could live with, but disabling NAT was not possible and a switch backplane running at 1gbps were the final blow. Also I do not want to have to log into an online account to use my (maybe airgapped) router.
Iāve dithered on the UDM-P, the reviews are very mixed.
Iām in a strange place with UniFi as a whole, as my APs are limiting download speeds to about 275mbps while upload speed is line speed, as is wired speed. There is lots on forums and Reddit about strange issues like this with Ubiquiti and they could really do with some firming up of their software.
Ubiquiti hardware is an great but their software has some of the worst QA I've seen in my life, the forum is basically an continuously ignored issue tracker where I've found dozens of problems I'm having with no solutions (about 200 aps and 150 switches/routers). Ubiquiti software is absolutely abysmal.
still rocking the wrt1900 and openwrt/lede
When I bought my WRT few years ago it was like $50. The one you have shared is like $150+. Why are routers so expensive for doing one simple thing?
I would expect an AC3200 router to be more expensive than a WRT54GL (which is $40 today). More capable hardware with a more expensive BOM will come at a higher price.
Yes, it may very well be the most successful router ever sold, but have you thought about how many new models were NOT sold because the oldie WRT54G was chugging along all too well?
If its success has kept uncountable, "segmented" garbage devices from ever entering the market, I'd say WRT has been even better for the consumers than you think.
I think you both agree.
Unfortunately what is good for consumers in this case is bad for companies, because it reduces long term sales.
Certainly that's a good thing though. Conserving resources and discouraging needless waste of perfectly functional products is a good thing.
Good for the world, bad for the capitalists. Guess who wins in the end?
I would continue to buy their newer routers if they have open firmware a la WRT54G. New wifi standards came out, had to install routers for friends and family, and WRT54G itself kind of died after 3 or 4 years... (I bought a second one, but by then N standard was up and running, so 3rd was not Linksys)
But think of the economies of scale and the $ saved in terms of RnD and marketing
Cisco didn't want a threat to their lucrative enterprise market.
Imagine if they kept pumping out updated hardware supporting DD-WRT over the years, and eventually captured 80+% of the home networking market. Now consider that, during that time, a generation of future networking engineers cut their teeth on hi-po Linksys home routers, giving Linksys a segue into the lucrative enterprise market as this generation of people started gaining influence.
This ended up being one of magical events that could have been the turning point for a small, unknown company to take on a giant, and win. Instead, the opportunity was squished through a smart acquisition by Cisco.
That's why I'm so impressed with OPNsense and pfSense and a wide selection of build it yourself hardware selection with them. You can own and tinker with your own router top to bottom. Seems like a niche market and I'm wondering why they aren't catching on with this same community that embraced the WRT.
I think those that want to run an open source software stack, but not assemble the hardware themselves, are served pretty well by going to the OpenWrt website (the successor project around the original wrt54g open source release), and choosing a suitable router from the table of hardware they maintain, and then just install openwrt on top of the stock firmware.
That's what I've been doing ever since I jumped ship from ye olde WRT54G (currently I have a Zyxel Armor Z2, and I'm happy with it).
FWIW, "assemble the hardware themselves" means buy a > 5 year old desktop computer and add a multi-port PCI-express NIC. Or even a USB3 -> Ethernet adapter.
Moving to pfSense was the best decision I made for my home network.
I never dove into the WRT devices myself but it definitely has a niche.
Wonder if there's a chance some of the router projects and Pine64 could collaborate somehow to make a fully open router. Pine64 seems to be quickly developing some production chops and the various router projects also seem to be doing great work.
PC Engines makes a long-term series of pretty open router boards that works with vanilla Debian, current iteration is APU2: https://www.pcengines.ch/apu2.htm
It is pricier than low-end router, but they are high performance and are much easier to use.
If Pine64 threw a bunch of Ethernet ports into a Clusterboard that'd be a pretty killer platform for a router. Start with one SOPINE for the actual router stuff, then add more for things like NAS, print servers, home streaming, home automation, etc.
Turris Omnia is supposedly one of these routers. I have their old model from a few years back, and it's been serving quite well for all my needs. The OS is their custom version of OpenWRT, and you can do stuff like LXC, Wireguard and all that quite easily.
The only problem is the ARMv7 hardware, which doesn't really cut it with modern Internet speeds anymore, especially with Wireguard.
That said, I can't wait for pfSense and opnSense finally support Wireguard. And pihole should finally get a FreeBSD version. I'd much more prefer the sense systems over the wrt, but the time is not yet here.
ANY more work in this space would be great. The SG1100 seems similar already though. Most configs of the Pine64 I'm looking at are single Ethernet port though, I'm not a fan of the router on stick config, even the one in the SG1100 is confusing internally.
I would love to see some more prebuilt pfsense boxes with useful options (like built-in 4G) - there are some on Amazon without detailed specs and some small vendors that donāt feel like shipping in all of the EU (canāt blame them for the regulatory and tax challenges).
TekLager is a Swedish shop that has a fe quite nice options.
I believe the underlying BSD is the issue here, everyone that says they tried to do it says it is an awful experience. Similar story for the problems with realtek Ethernet chips.
I'd love to find a compact router/machine that has SFP/Gigabit switch and optionally PoE capability with pfSense support.
Sadly, my annoying Mikrotik is the only thing I've found until now :(
That's a big wish list for compact.
So, I worked for Broadcom for some years after this went down. This post is purely descriptive to give people some insight into the history from inside the company; I'm not commenting on who should have done what (although I was not directly involved, so if someone who was comments, take their word over mine).
Broadcom made an error of judgement here, but this incident fostered a deep distrust of open source, at senior levels, that persisted for more than a decade after; perhaps to this day.
Firstly at this point Cisco was, at the time, Broadcom's largest customer by a large margin. This caused huge tension in that relationship that was totally unforseen, and was very painful for a while.
Secondly, a at a certain point it dawned on Cisco and Broadcom that the GPL lawsuit was not like a normal business dispute , because businessmen after a certain point will settle for money even if they didn't get everything they want. Sure a few people will keep going to the detriment of their own business, but most aim to make profit, not expound a principle. Many companies in the position of the FSF would have settled for a cut of the revenue. But the FSF wanted the source code released, and they were prepared to kill the business to get it. So Cisco and Broadcom had to concede. The source code was released, and OpenWRT was born.
The fallout, though was that subsequently Broadcom router ICs were designed with hardware accelerators which were separate from the main CPU. They were driven by separate CPUs on the same SoC that did not run linux and whose drivers could not be demanded under the GPL. none of the open source firmwares can run these devices efficiently unless someone spends weeks reverse engineering them.
I'm not sure about the last point. I would think hardware dedicated accelerators were done because it was the cheapest way to achieve that performance not because it allowed to somehow bypass GPL. However, choosing to not run Linux but some proprietary OS could most certainly have something to do with that.
At the end of the day, was it a good thing? I would say it was. It opened many generations of home router hardware to being modded/replaced with user controlled software. It even created a market of its own where certain consumer router hardware is advertised as being designed to run custom/third-party software and where vendors themselves ship with some heavily modified software and release the sources for it from day 1 (which are the only wifi routers I shop for these days).
Indeed, hardware accelerators weren't introduced because of the GPL. What changed was that previously they were connected to the main cpu and driven by drivers that fell under the GPL; to avoid the GPL secondary CPUs were introduced not running linux at all.
I can't speak to Broadcom's motivations, but the end result has certainly been that they are the least open source-friendly WiFi vendor, behind Qualcomm-Atheros, Mediatek, Marvell and Intel (client only). When Linksys wanted to do a successor to the WRT54 series trading on its open source reputation, Broadcom wasn't an option because they've made themselves the NVIDIA of WiFi.
I donāt know why WiFi AP manufacturers donāt just give up and just use stock open source firmware on their devices. They are not even trying to get the sw right. The first thing I do when buying one anymore is ditch the built in tinker-toy firmware and install an open source one. Lots of companies that make hardware treat software as just another line item on the BOM like a bolt or a screw, and source the cheapest shit they can find, rather than treating the software as an integral part of the product that needs the same polish as the external box and marketing materials.
FRITZboxes are better in terms of their software. The names and descriptions for the various controls are written in proper language, and there are loads of graphs and stats for the nerds. My only gripes are that the interface relies too much on JavaScript (you get sent back to the login when you refresh the page...) and that, at least on my model, there is no way to perform a factory reset without plugging in a phone handset (who has one of those these days!).
Of course, OpenWRT still kills it in terms of support for standards. FRITZboxes have their own stupid mesh protocol that's only compatible with other FRITZboxes, not implementing e.g. 802.11s.
To be fair, the FRITZ suite also wants to (and does) support Cable internet (afaik the only non-ISP-supplied modem or router-modem you could even buy in europe), DECT, and a range of 433MHz home automation products. And of course, you mentioned their homebrew mesh stuff.
So there's a lot of non-standard tech available in those boxes and it is no huge surprise that this is kept proprietary.
IQRouter[0] uses OpenWRT, or at least is based on something like OpenWRT, and by default measures your bandwidth so it can manage traffic to reduce congestion. Recommended by Jim Gettys[1]. I've been pretty happy with it.
Probably because they can ensure their software works properly. I recently dug out an old Asus RT-N16 and the latest Tomato firmwares are all completely broken. WAN DHCP doesn't work. Took me a couple of hours to figure out. Turns out it was broken a year or two ago and nobody has noticed (it's a pretty old router; I doubt anyone still uses it). The official firmware worked fine.
The point is the manufacturers have a much higher incentive to ensure everything works than open source developers.
The ASUS firmware at least seems to support way more features than Tomato did, at least without resorting to the command line. E.g. my ISP requires the VLAN ID to be set. I doubt open source router GUIs have a nice option for that.
It's not that old, works tolerably for a small household if the link speed is below 100Mbps. Freshtomato worked fine last time I checked. Too bad these chips suffer performance loss with OpenWRT, though.
The sad thing is ten years later the market is still dominated by devices with half its RAM.
I had a Buffalo router that did that; IIRC it came with their proprietary firmware and a copy of DD-WRT on a CD. (Might have been the other way around; this was about a decade ago.)
I don't believe they would have been in much legal issues: they'd have to make sure the copy of DD-WRT they shipped was fine, but if you get updates / flash your own, there's no reason they'd be on the hook.
I still have one of these in a box. Maybe two as I used to encourage friends to buy them years ago.
I only stopped using it(with some custom firmware) about a year and a half ago because it was just too slow - and had gotten this weird issue where it would cut off the internet to some devices while keeping them on the network.
It was really by luck that I had one of these in my teenage years initially to play with. I sometimes wonder what hobbies I would have developed if I hadnāt lucked out and found working computer in the trash, or my parents had bought something that wasnāt such an easily moddable desktop (AMD K6-2 was the CPU in the first computer they purchased).
Anyway - the WRT54G really was a fun piece of hardware to play with.
> because it was just too slow
The WAN to LAN throughput on a wrt54g is only like 34mbits/s. Itās just too slow to handle a fast internet connection. I guess the fact that so many are still being used shows how ISP connection speeds have stagnated.
Or that there simply is no need for that high a bandwidth. Netflix, e.g., uses fancy compression algorithms and you can almost watch their HD offerings with ~3mbps. They do recommend 5mbps and 25mbps for their 4k content.
I so wished, I could get here a 6mbps connection for half the price of my current 65mps line.
So there's a need for it, it's just that you don't have a need for it.
I'm happy with my 1gbps connection where I can download a 50GB game in less than 10 minutes.
You might not have a need for it but others do. It really sucks to buy a new game after work and see that you won't be able to play it that night because it has a 5 hour download time.
Bro, when I want to play games with friends I frequently have to update to play because I play so rarely. Speed means lower latency to startup.
If you still want to live that WRT life with something like OpenWRT/LEDE (I think they re-merged now just under OpenWRT but I'm running LEDE currently) then I can highly recommend this [0] updated version. I have it and I can get gigabit speeds (wired) through it just fine and don't have any issues with the wireless other than at the far, far end of my house and only sometimes.
My next router will probably be a Ubiquity setup so I can setup 2-3 AP's for full coverage and coverage out to the (detached) garage but that setup is not cheap or simple and my current issues are so minor that it will be a while before I pull the trigger on that.
I'm still running a WRT54GL with Tomato firmware on at my parents place. I used it until I upgraded to a faster one, but the reason it's still running is that it provides the longest 2.4GHz range which is perfect for a large house. I've tried Ubiquity, newer ASUS routers and the range is shorter and their devices prefers to connect to the WRT54G. And my parents don't need super fast wifi, just a stable one.
One fascinating sidebar in the WRT54G history was the Fon [0] "Fonera" project, which was one of the reasons I bought WRT54G specifically. (Which I found in a box just recently, Fon stickers beside it.) Fon had the idea of trying to build a network of independent residential wifi that users could share roaming among each other. It was a paid wifi network, so people that had a Fonera AP at home could opt for either free access wherever they went as benefit of running an AP or a simple profit sharing option (but then they'd pay for their own roaming).
The original Fonera projects were all built on top of OpenWRT.
It was cute idea for trying to make guest-accessible wifi ubiquitous. It ran up against shifts in law in some countries making network AP owners more personally responsible for accesses to their wifi. Also, it never really hit network effects that the scale mattered. I ran a Fonera AP through a large chunk of college/grad school and can't say that I ever saw another AP in the wild to take advantage of the free roaming (and if I had it switched to the profit-sharing mode I never would have seen a dime).
Fon pivoted entirely out of the Fonera residential wifi project in 2016. It was a neat idea, but it didn't survive.
I don't recall which wifi router I used, but also I setup fonera for about a week. I also don't believe I ever saw one in the wild. I thought it was an interesting idea.
Man I used one of those forever, I think I finally threw it out once 100Mb switch and G wifi wasn't quite enough. Tomato was probably my favorite firmware for it. I remember bricking it with a bad update one time and having to jumper two pins with a paper clip to put it in tftp mode in order to load working firmware.
A buddy of mine got divorced and found himself in a tiny apartment with ethernet and not a router. I dug up my WRT54G but yeah, G wifi... In the end, we found an unused TP-Link Archer C7 for him, but that WRT54G brought back some memories.
The highly coveted WRT54G!
I picked up a number of these at thrift stores over the years. Occasionally I'd get lucky and get the "WRT54GL" version. I was sometimes persuaded to exceed my "$5 or less" budget for a "L" version.
They were great for having a little Linux-box to do oddball utility stuff-- ad-hoc OpenVPN endpoints, caching DNS server, captive Wi-Fi portal controller.
They were eerily solid for their built-to-a-price-point nature.
A few years back, I spotted two of these for $0.50 at the thrift store amongst all the outdated DSL modems and answering machines. My tech hoard was already large enough at that point so I made sure they worked, flashed the factory firmware, and turned around and sold them for $25 each on craigslist in under 24 hours. Easiest beer money I ever made.
Its spiritual successor was the Asus RTN-16. I still have one sitting on my bench, running TomatoUSB. I got it 9 years ago, and for the past 5 years it's been a 2.4ghz wifi bridge, connecting the hardwired devices in my office to the wifi router in my house. It just keeps working, so I keep using it.
Of course I can't forget the first time I got a WRT54G. My brother in law had one just sitting around unused (around 2006 I think) and while I didn't know a lot about them, I asked him about the router. I ended up trading him a well used laptop for it. The router was the locked down version. Then it died. Oh well.
Get a daily email with the the top stories from Hacker News. No spam, unsubscribe at any time.