Browse the linux kernel commits, it is littered with redhat, ibm, qualcomm, mediatek, amd, intel and realtek contributors. I'm pretty new to OpenWrt but have already been credited with contributing code and testing. My contributions were based on reviewing lines of existing code. Not once did I see a contribution from an employee of a router manufacturer.
Another example is OpenBSD, they list their financial contributors
http://undeadly.org/cgi?action=article&sid=20150708134520
There are manufacturers that provide GPL code for download or on request.
Do your research before purchasing and support appropriately.
True, but it's something that requires time to arrange things and quite a bit of money upfront, If someone volunteers to do that, please contact project members and see what they say about it.
if you think big brands aren't just buying hardware in bulk like that I have some news for you. For most sub-200 euro devices, they do that as there are basically no margins.
Most GL.Inet routers have been added to OpenWrt by the manufacturer's developers, even if they are not using company emails (especially lately, they started sending patches to the mailing list before the device is released). So it may not be obvious what was contributed by a manufacturer. Although in general the big brands don't really care.
Then again they are also known for their customer support people answering questions about doing electrical mods to their devices (i.e. things that 100% void any warranty) with partial schematics, so I concede that they are pretty unique.
Plus there is Gateworks and NXP developers sending for their own stuff.
I was thinking of using a reputable GoFundMe type service to handle the money/orders.
OpenWrt would pick the best supported chipsets in several categories:
All would be upgrade-able via LuCi.
The cost of each device would include a 20% donation to OpenWrt and provide hardware for the developers who sign up to contribute.
OpenWrt would develop and host the firmware.
ZBT would manufacture, to the projects specs, imprint with a logo and flash with a basic OpenWrt firmware. ZBT would be responsible for the hardware warranty, OpenWrt for firmware bugs/vulnerabilities.
Not sure about the distribution; direct from ZBT or regional distributor.
I repeat, someone must devote time to decide what hardware is good or not, deal with manufacturers, ship orders and keep stuff stocked, and afaik none has volunteered to do that. It's something you are underestimating I guess, but it's not something you can just do "whenever you feel like it".
"Making the firmware" is more or less a non-issue as on such a device you know already everything from the start and control the bootloader's configuration, anyone of the basic "contributors"that has added a device in the past can do it.
There is nothing to stock, If the project reaches its goal, ZBT sets and runs the production one time. Late orders would be not be filled unless another production run minimum is met.
Only one prominent manufacturer advertises the capabilities that are needed. Although to keep them honest several bids should be obtained.
There are only 2 SoC manufacturers, that open source, for the lower end device, Mediatek (mt7628 or mt7621). The mt7620 is an older mt7628 that is not as energy efficient. These include a 2R2T 2.4 radio on chip. The 5.0 radio is an addon chip and there are several to choose from. The mt7615 is having some teething problems but older chips with a track record can be used.. The person handling MT76 development could make a recommendation. Qualcomm is the other - I'm not familiar with Qualcomm's offerings to make suggestions.
This is not a project for volunteers.
This is a full on business. You need to acquire warehousing and shipping facilities/services
You need to provide warranty, no you can't offload warranty to a manufacturer when you put your label on the device. You need to employ someone to deal with customers.
It is a great idea, But the way your going about it is wrong.
When you open a case or irreversibly flash OpenWrt, you have voided your warranty anyway. The forum already deals, gratis, with users having problems with devices and actually is a more reliable source of flash dumps than any manufacturer. The available wikis are far more complete than any device manual I've seen. The GPL'd source is already available.
I agree, that It needs to be crystal clear for those that fund, the devices are specifically targeting OpenWrt users. The first thing most users will do is update and customize their packages.
Another hurdle comes to mind though and that is registering the device with the regulatory agencies, ie FCC. The FCC does not review the firmware, just the radiation and strength/specificity of the wireless bands. I have been able to find ZBT submissions to the FCC.
Oh my
What is the point of this excercise if your not offering warranty.
There is no shortage of products used and new that are supported by ooenwrt.
I don't see why this is required.
I don't understand your point.
https://openwrt.org/supported_devices/432_warning
While there is no warranty of ongoing support for any device under OpenWrt, those with insufficient resources are at great risk for “end of support”.
As quoted from the website, OpenWrt comes with no warranty. Here is why someone may want a device that is specifically built to run OpenWrt:
https://openwrt.org/reasons_to_use_openwrt#
Here is a second example, a Cudy marketed device that has attractive specs
https://forum.openwrt.org/t/cudy-wr1300-support/67593
and
advertises OpenWrt compatibility. They specifically disclaim warranty support if you load OpenWrt.
http://www.cudytech.com/openwrt_software_download
In my opinion
Openwrt is not set up to run a business.
Consumers expect warranty.
Openwrt cannot endorse a commercial product or have product branded as openwrt. However a product can claim it supports openwrt and use the logo with condtions on documentation and marketing etc.
A business is welcome to do all the things you mentioned above but openwrt shouldn't.
A business is welcome to make a clone for openwrt and provide the support and testing before releasing updates to customers. Allowing warranty to be maintained.
A business is welcome to support openwrt and to ask for help when needed, or they may choose to hire a professional/expert .
In summary it needs to be a business.
If this is where openwrt wants to go then it needs to spin off a business first then do the things you mention.
Why couldn't it be a non-profit with costs structured to cover overhead + OpenWrt Donation? The charter could be to support and advance Linux on embedded devices. Non-profits can sell a product and in this case, it would be to provide a supported development platform. It would not be any different than buying a Debian Install CD/DVD. The CD/DVD are linked on the Debian Website, approved to carry a Debian logo and Debian gets a donation from each sale.
https://www.spzlegal.com/nonprofit/can-nonprofits-sell-products/
OpenWrt users/developers are also consumers of OpenWrt and have never expected a warranty.
Here is an example of an mt7621a based router, with a custom logo.
Another mt7628DAN
https://www.alibaba.com/product-detail/360-Wifi-router-T2-security-5g_62489477175.html?spm=a2700.7724857.normalList.33.11f0302aVO5FeY
Exactly, an approach like this can only fail - badly. If only because the devices will have to be considerably more expensive than the underlying OEM devices ordered from China directly (shipping costs, customs, taxes, small production run, warranty/ support etc.), respectively the price of comparable devices. Likewise the life expectancy of these devices is limited (perhaps 2-3 years at most), during which time the whole stock must have been sold off completely, in order to avoid a loss. What happens if the OEM (silently) changes the hardware between hardware revisions - ideally you notice that before ordering the next batch (in which case you may not be able to continue offering your branded device), worse if you're now sitting on hundreds/ thousands of unusable devices. Next question, what market would you be targetting, cheap/ low-end (you really can't compete here), something in the middle (still hard), highend (that's easier to 'sell' to the user, but still excludes a lot of potentially interested users because of the required prices). The next issue would be the time-to-market delays, if you want to bless a device as officially supported, you need to provide 'perfect' results from day one (but you only get access to the devices after it's already selling by the OEM to the masses), this easily adds somewhere a year and more, before one could offer it as an "OpenWrt router" - combined with the general life expectancy of these devices, that doesn't leave much room (as in, it's already old before you can buy it).
Other projects have tried this in the past, remember Ubuntu pushing hard for a smartphone kickstarter - where did that go (and Canonical has deep pockets)? Or an even worse example, the KDE tablet, which failed twice, horribly.
When don't pay for a product or service you can't expect a warranty. But the law and the consumer demands a warranty when a product is traded.
The two Alibaba device links I provided both state the manufacturer would warranty the device EDUP for one year and Smart for 1 - 3 month. Both claim CE, FCC and RoHS certification.
Ubuntu and KDE may have bitten off more than they could chew. In particular the phone would need an unlocked connection. This is much simpler and sits behind the connecting modem.
From my standpoint, we need the following. Someone fluent in Chinese to negotiate the deal. A commitment from the project to develop/host firmware. Negotiate distribution to EU, US Australia/NZ and China. Legal support to set up the non-profit.
I'll craft an email to the project that lays out the idea. If OpenWrt gets behind it, it should have a good chance of success.
Will also look into setting up a poll to gauge intestest.
...something substantial to talk about, i.e. a detailed plan.
All I see right now is: "I have a dream...".
A detailed plan should include
If this plan shows that your idea can be realized with 100 manhours in 6 months, with <$500 investment and with no risk at all, it will certainly succeed.
If your plan however shows that it needs 5000 manhours in 2 years with >$20000 investment and with high risk due to warranty claims and unsold devices, success is not guaranteed.
Without this detailed plan, all discussions here are just hot air that will lead to nothing.
Hardware failure - return to manufacture directly under the manufacturer's terms. These terms will be verified with manufacturer in writing and put terms on the Go-Fund-Me site. If Manufacturer reneges, publish their agreement and their failure to honor it on the website(s).
Only 50% volume sold. - No production run unless pre-paid orders reach 100% threshold. If threshold not met, money returned under the terms of the Go-Fund-Me site. Investors may risk a handling fee.
Idea is not a daily enterprise but rather the setting up of a discrete production run. If run is successful, get and publish feedback from users on forum. If future demand sufficient, review advances in hardware support and setup another run.
I pm'd @tomomas about setting up a forum poll. If the preliminary interest is there, will hash out more details. If no interest, it stops there.