Qi Hardware is copyleft hardware. Our mission, provide free software developers with stable, mass market quality hardware that they can develop compelling end user applications on. I’ll contrast this with some efforts we’ve witnessed in the past where the hardware was not stable and not consumer ready. The openness of that hardware could not overcome the problems of instability, and software development floundering in the quicksand of changing hardware. At Qi the hardware designs are selected because they are stable and shipping in mass quantities and because they can be opened and safeguarded for free software development. In some cases the hardware designs already run some version of a Linux kernel, but the kernel is outdated and reprogramming the device to other purposes is complicated. We make that task easy. Our key commitment is delivering stable, working, high quality hardware to free software developers. With that firm foundation software developers can focus on making insanely great end user applications.
Our process starts by finding shipping hardware that can be opened. We work with manufacturers throughout Asia and either source hardware that can easily be opened or work with them to engineer open versions. Our hardware engineering efforts are focused on documenting these designs, publishing them under copyleft licenses, and then fostering a collaborative community effort to improve the designs. This engineering effort will utilize free software tools such as KiCAD and HeeksCAD, and copyleft the complete electrical and mechanical design.
Copyleft hardware is hardware with a future. In a proprietary system the end user is at the mercy of product life cycle. In the proprietary world there is a scheduled end of life — EOL. At that date software support ends; the hardware roadmap hits the wall. With copyleft hardware there is no EOL. The design is always open for enhancement and extension; and, with GPL software running on the hardware, end users always have the freedom to support themselves and each other. The roadmap for hardware is open. Because the designs are copyleft any engineer with an idea can create a new design, but only those willing to share back the design can take advantage of it. At Qi our roadmap philosophy will be simple — add capability in an incremental fashion. This ensures a focused software development effort.
Tags: copyleft, qi philosophy
I’m wondering about the hardware choices… for example did you choose the XBurst processor because it’s more open in some sense?
Excellent question.
We chose it because:
1. The semiconductor company (Ingenic) has their own team working on Linux and u-boot support.
2. They are small (ca. 75 people), so easy for another small company like us to work with.
3. This is somewhat ironic, but - because they have a weak engineering team, they focus on delivering raw digital computing performance. Other semiconductors like TI or Marvell have huge (and strong) engineering teams and their SoCs are growing and absorbing more and more free software features.
I see two ways forward: Ingenic (the XBurst makers) continue down a path of delivering manufacturing excellency, raw digital computing, etc. (say like Intel in the last 30 years). Then we can continue to use these chips. _OR_ Ingenic chooses to become an IP house (same as I would say TI nowadays, Qualcomm, etc), then we will need to look for another solution so that we don’t feed the people that will then grow their proprietary IP underneath us, limiting what free software can do.
The best would be if we could continue to use Ingenic for a few years, and in parallel develop fully open DSP chips so we can do processing of analog or RF data.
I could also imagine that we hook up with gta02-core and make Samsung-based products. Samsung of course is very hard to work with, it’s a miracle that we have any sort of direct contact into the company. Also they seem to not be able to decide whether their business model is manufacturing excellency (they have their own foundries and continue to invest huge money into that), or intellectual property licensing (right now they don’t invest much there).
Phew, long answer but I hope I got my point across.
What do you think about the XBurst CPUs? Any other chips you think are interesting?
is overclocking this processor possible?
Maybe from 336 MHz to 360 Mhz, but even then you may introduce instabilities under certain workloads. Not recommended :-)
good news, also more interesting thing - programm CPU slowdown (e.g. power saving)
Hi guys.
I find the idea of copyleft hardware very interesting, and I think you are innovative and daring to pursuit this as you do.
Just an idea, have you been looking into making open, ebook reader device with various wireless connectivity, user-customizable software and no lock-in?
As ebooks are growing and spreading these days, maybe this could be a nice segment for you. (Not that you should stop working on your Nanonotes, but perhaps to be considered as a future product).
As we know, eventually, free e-book/music/video stores that sells non-locked contents in a non-discriminatory way to all customers regardless of what type of device you are gonna read/play/watch it on will appear and break the lock-in-loving webshops of today.
Keep it up, I wish you the best of luck.
Hey,
> I find the idea of copyleft hardware very interesting, and I think you are innovative and daring to pursuit this as you do.
Thanks! We will do our best!
> Just an idea, have you been looking into making open, ebook reader device with various wireless connectivity, user-customizable software and no lock-in?
>As ebooks are growing and spreading these days, maybe this could be a nice segment for you. (Not that you should stop working on your Nanonotes, but perhaps to be considered as a future product).
>As we know, eventually, free e-book/music/video stores that sells non-locked contents in a non-discriminatory way to all customers regardless of what type of device you are gonna read/play/watch it on >will appear and break the lock-in-loving webshops of today.
We have thought about ebooks and still are actually. The problem is that the wireless connections you mentioned drive the hardware price up heavily and also development costs. That, of course, does not mean that we will not attempt it, just a bit later in our roadmap. Although the question is whether the world needs yet another ebook-reader, maybe it would be more interesting to build a device that can also be used as an ebook reader but is much more …
>Keep it up, I wish you the best of luck.
Thanks again!
/mirko
>We have thought about ebooks and still are actually. The problem is that the wireless connections you mentioned drive the hardware price up heavily and also development costs.
Actually, I’d be very interested in an e-book reader WITHOUT WLAN. It would be like an MP3-Player: You “consume media on the road” and “feed” the device at home on your PC.
I suppose most (all?) e-book readers have wireless access because the manufacturers want you to use their online bookstore - which is the only way to legally “feed” a Kindle, AFAIK? But I guess since these devices sell quite alright, other people don’t mind that and actually like buying books on the go. ;)
It’s like an iPod with wireless access, so you can buy music on iTunes. Everywhere!! j/k XD
I would love a copyleft ebook reader with eink-like LCD (http://www.pixelqi.com/ like the OLPC). That would rule.
The first one to bring such a thing to market is going to win bigtime.
+1
Do you have an OLPC? I applaud (same as everybody else I guess) the goals of the OLPC project, but on the delivery side they have totally failed for many years, from my perspective. Netbooks are delivering.
Pixel Qi? Same - everything they say is great. I also want that. Sure we should build products with those screens. Have you ever seen a Pixel Qi LCM in person? Are the things they are talking about real?
Mary Lou is doing a great job, it’s about attracting (big) investor money, then spending it to fulfill your own prophecy.
I am not very familiar with LCM technology or industry trends. So I am just a (small volume) buyer. I doubt Pixel Qi will have any real product for me anytime soon, but I will keep checking :-)
I think it’s easier to get a used Obama sneaker for charity auction than a Pixel Qi sample screen to hook up to our NanoNotes.
[...] software principles to consumer electronics. The three fundamental elements in our development are copyleft hardware, upstream kernels and community driven software. Each of these form a vital part of our Qi or [...]
I’m interested in the legal formulation of this concept “Copyleft hardware”. It appears to me that, often, patents are used to restrict access to hardware technology, and that some kind of “open patent” approach would be tedious and expensive, relative to copyright. I’d like to hear other people’s ideas about “copyleft” for hardware. Please put me on your list for that conversation.
Also, I’m interested in promoting a set of system-level components for electric vehicles - motor controller, battery management, precharge sense and sequencer, charger, volt-amp-hour meter, distributed power controls, and such. Currently, these components are either very expensive, or not available from manufacturers “holding their breath” waiting for that billion dollar contract from GM or Ford. I will be interested to see how your business model might be applied to that sort of business.
Also, I’m interested generally in the “decentralization” of hardware manufacturing, where, in “the world of the future”, “rapid prototyping” tools may achieve manufacturing economies relative to traditional manufacturing techniques - which have otherwise expensive up-front costs - and would also move control of that manufacturing technology into the hands of “the average person”. Promoting the use of “personal” design tools, such as KiCAD, or “whiteboard” collaboration tools, such as Coccinella, is a great start. Perhaps Qi Hardware can help people to understand how to make the most efficient use of the existing low-volume and high-volume manufacturing industries?