|
OrangePi has created many world-beating hardware products and is gaining recognition around the world from hobbyists and IoT developers. And is also behind much other board design as ODM. They are good at what they do. They gained respect and their relationship with developers community is good. You don’t need to worry about that. Most of their boards are also well supported by 3rd party community projects and you can get a top quality Linux from here: https://www.armbian.com/download/?tx_maker=xunlong. Unfortunately, there is no support for the one you seek for.
We, your existing and potential customers, want to use your products to create the next generation of devices that will carry many new ideas forward into the future, but we are unable to do so because there is not enough information available and software, documentation and support frequently cannot give us the information we need to fully use your amazing products in our projects Board schematics are made professional, accurate and are usually quickly available, software is as is. Xunlong doesn’t develop chip support. This is done by chip designer (Allwinner, MediaTek, Rockchip, …) and they gave you what they got from them. The one and only kernel are glued with »Debian«, »Ubuntu«, »Arch«, »Android« to "prove" it can run "Linux" … and as you already figured out, you can’t just plug it and run your project. They don’t fix kernel and not many other chip makers do that as well. Unfortunately.- chip support is above board support and Linux kernel is a community project per se. We (www.armbian.com) and a few other groups are interested to give those boards a proper Linux. But our resources are limited. Xunlong has a few of engineers and they can’t match a work of thausands of people. This is the common reach of board makers.
- they can support their work only and do that well. That is board design.
You have entered into direct competition with the most heavily supported products out there such as Raspberry Pi, Beagle, Arduino, Tinker Board and many others With exception of Arduino, all mentioned are toys. More or less. What they have in common is a big market share. They are popular and well supported. Not just because of popularity, but for simplicity. Well, Tinkerboard tends to be a bit more complex, but it remains a toy due to failed powering design. All those boards are general purpose boards while the board you are bringing up here (4G IOT) is a very specific.
I've been interested in your products for some time, and we purchased two 4G-IOT boards because Linux support was advertised. We also expected that we would be able to get full documentation for the hardware including the 40-pin header and were also interested in the optional 4G bands that were initially advertised. Remind your self that you bought a development board. You have a board (hardware) and your job is to make it work a little better. This can be our joint effort if you find people who are also interested to have a proper Linux on this board. This is how things work here around. Of course, Xunlong wants to have a superb Linux … and they are doing what they can. I am sure they could do it better but still, that would not change the situation much. I doubt they can throw millions to the kernel development. It would be also stupid since they help competition this way.
Any board can get a top Linux support. You only need to find, persuade and/or finance a group of competent people for about a year or two. Do the math. Rockchip is doing that, but they struggle. MediaTek as well. NXP too.
More transparency about the development process and where your company is struggling. Release what has already been completed of Linux for your devices, even if it isn't finished or fully working. Xunlong is a hardware design company with limited software development capabilities. There is nothing much to see.
Development progress can be observed here:We, and I'm sure others, want to help! You need to help on the right spot. For example - just a few of companies helps us with managing the cost of maintainace, while extensive development with business purpose on charity levels … is a joke.
OrangePi should request assistance from your community to help build things like Linux distributions if you are unable or unwilling to do so - a number of us are long-time developers and could help you build your community Kernel development is much bigger operation than board design. Linux is about the kernel and making many Linux distrubution without making a good kernel is pointless. This is not Intel world, where we only have one kernel. Here on ARM almost each chip design has its own kernel. Linux on ARM is very much about the kernel, not Linux distribution. The situation is way more different.
More details on the products you sell regarding pinouts and other information. Schematics are not a good way for new developers to find out how something works. As an example of information provided by Raspberry Pi, here is their getting started guide which details how things work on the board, and how to set everything up. Raspberry Pi is a simple and dump device, made for newcomers where technical manual could be written in short. Remember also that Rpi runs Linux inside propriatery OS and remember it has many hardware design flaws. That fact becomes important when you plan to do business.
My company alone has at least 3 projects in mind for the OrangePi range of boards right now, and if we can get the support we need to begin development we'll be ordering these boards in larger and larger quantities as our products begin selling into the market. We cannot move forward with those projects without proper support on your products. Linux works, but quality Linux and quality support costs. If you to skip learning, you need to pay or give something in return.
Remember. Xunlong doesn't make drivers or fix kernel. Others do that. We did it a lot. If you want the fully functional top shit mainline support of this 4G IOT board you might need to throw a million or more into R&D and wait a year or two. (it's a rough estimation). For just fixing it a 100K might do. Who knows.
We worked with 2G IOT for a while - the goal was just to make shipped old legacy kernel usable. We spent months of our time and money and receive no external help (except regular Xunlong donations, which are not on the level to cover such operations). Two months ago we ceased the operation and removed what has been done since we can’t do it alone.
we will need to seek out products from other manufacturers LTE modem and mPCI capable board and you are done. Or is that too expensive?
|
|