snif.
I agree a little with everybody. In fact I dislike Allwinner, I don't know much wrong about Texas Instruments (but dislike PowerVR by Imagination Technologies) but at the end of it, ARM itself is hostile to free software,
and Intel and AMD are already selling backdoor-only hardware, and so on. Aparently you're left with some MIPS and Power 8 (if you can trust IBM) and maybe Arduinos, Myquists, etc. .
Buying from vendors that alienate users is collaborationism (no Godwin intended, afaik the word began with Napoleon, not the nazis). In fact reverse engineering and adding free software support for those
vendor products is adding value to those vendors offerings, so it's collaborationism too, if you want to be strict. Learnign to use proprietary software also adds value to the software, so it is collaborationism too.
In fact the more I think about it, the only conclusion I can reach is that the best option is not using computers. But that will hardly help getting "just computing" to the world. And "just computing"
is useful. And humanity is not so well off that it can ignore useful resources. History proved humans can live without electronic computers, what is unproven is that computing needs to be evil. Can "just computing" be achieved ? If it can, can it exist at the same time than evil computing or only after evil computing is erradicated? The answers are hard and mostly a leap of faith in humans.
In fact there is a deeper problem that you could name with the oximoron "immaterial sustainability". Not only proprietary software is leading to obsolescence and therefore extra waste and resource depletion. The fact that hardware and computers on sale change so fast and there's so much diversity makes it very hard to build collective knowledge that is sufficient to manage that software and hardware corpus. It's not only that we don't have time to reverse engineer and free devices. It's that we don't have time to audit all free software, don't have time to keep the internet safe, don't have time to (learn and) spread the good practices and the understanding throughout society so that the use of computers, networks and services is concious.
So innovation is not clearly positive. It can bring new positive features, but it can also bring antifeatures, and it sure brings a requirement to judge whether the new features are positive or not. The more open the innovation is the less hard it is to judge and the more likely it is that positive features outweight the work of judging it and maintaining it up to date and the more likely it is to detect antifeatures or unintended problems.
So in a way the only ethical thing to do is pull the brakes (I mean to slow down, in fact car brake pedals are pushed, not pulled... and bicyle brake handles are... mmmm... what vehicle brakes must I have been thinking of? an old train? a horse car?). Not too unlike "ungrowth" (is that English?, "decroissance" like by Latouche). We should buy the fewest products as possible, not to feed evil vendors, and if we deem it necessary to buy one, we should try to buy it second hand so that the money does not go to the vendor (it stil increases the vendor product's value, but in a way that it is competing with its older self, so it could really slow down the abuse a little) and so that we reduce waste. And we should try to build as much independent and shared knowledge about the existing products as possible (knowledge includes software, drivers, repair plans, whatever). To minimize adding value to the evil vendors we should only publish knowledge about products that the vendor is no longer selling and priorize knowledge that prevents obsolescence of the old products. I reached this logical conclusion from trying to understand how Luc Verhaegen
got frustrated. He seems to have reached the conclusion that constructively building a better world may be counterproductive.
The problem with that approach is society as usual. If you try to keep simple, save the world and personally survive to convince others, you'll have to face competition from those who only concentrate in surviving and can choose to do whatever you do or whatever else your ethics don't allow you to do but theirs do. It may mean not finding a job, not finding support in your community, or simply being frustrated because your device can't do something the neighbour's crap does.
So building new products that are more just, sustainable and hopefully able to displace the evil products is a survival alternative you may be tempted to pursue. And I don't think the market allows to do it all at once. So you'll find some products that fix one thing, other fix others, and none will score on all fronts. Hopefully we'll get there little by little. As a buyer this means you're still screwed, you can just choose it to be just a little softer here or there. Must be a coincidence but today I've seen two mails: one form a friend asking what smatphone to buy and
one from
tinkerphnes.org that said:
Therefore we discussed in a small circle what the core of Openmoko
and Openphoenux is.
It was easy to find what it is not:
* it is not a 100% fair phone (we don't have the resources to track
components - it is enough challenge to have it working and being produced)
* it is not a 100% open phone (we have not found a feasible solution for
WLAN and GPU)
* it is not a 100% secure phone (we can't do security audits of every
component)
* it is not a cutting edge phone (we do not get the latest and greatest
chips as mainstream manufacturers do)
* it is not a geeks (only) phone (we want everybody to be able to use
it)
But yet they try to do what they can instead of saying since I can't do it all I'll do nothing. It is very easy to find what we still don't like. I bet they want to make that list shorter when and as possible.
I don't think it's their fault it's not as easy as it should.
Other projects will pick other priorities, not because they wouldn't like to achieve everything, but because they just have so many resources. Rhombus Tech decided to use a cheap A20. They tend to document very openly their decisions, but I don't have a link to this one handy. Maybe they thought making sure a cheap and ubicous SOC had enough free software for RYF was more important than judging a company dismeneavours. I don't know.
So yes, Luke Leighton is no fool, I hope he can deliver a fully free system, and even if he can get enough source/licensing from Allwinner I doubt he'll manage to reform and redeem a megacorp like Allwinner. If he does, praise to him, but it's hard. If he doesn't reform Allwinner enough so that they don't rip free software authors off ever again (ever? how can we tell?), then buying Allwinner will still be funding people that not only do not help improve the world, but who actively undermine those who do. I mean I don't sympathize with proprietary software authors, but I understand people do not have to share my views, so I must understand people may do that and be consistent with their beliefs. Taking the work of those that build common knowledge and making it exactly the opposite of what they work for is much harder for me to empatize with. It's just bully. After all proprietary software is not even illegal yet, but not complying with GPL is. And that someone does this and no lawcase is brought, not imports are barred, etc. is dismal.
So not buying anything is reasonable, buying the A20 EOMA-68 card is not ideal, but it is possibly not worse than buying many other things, and better than buying most of the products out there (I doubt Allwinner is the only GPL violator). And I guess Luke starts with A20 because he thinks most people don't think like me. Buying a laptop or desktop without a CPU card is not such a bad idea. A little waste until they (or whoever) get a better CPU card out, but it helps a project with a nice concept. Why not?
About RYF certification, I understand it's hard enough as is. It could include providers not doing anything illegal (breaking GPL, breaking labour laws, buying stolen minerals, whatever) or providers not doing anything unethical, or not contributing funds to political parties that endorse software patens or whatever, but one has to stop somewhere. It would be a bit odd to include in a certification that involved companies must obey the law. One would expect that certification or not, companies that don't obey it are brought to court. Even when that expectation fails how does including it in the certification help? And concentrating on what you can do with the supplied hardware and software, or how likely an average consumer is to end up running proprietary software, or some such makes sense. This certification or any other is not the only thing to consider when decding your purchase.
I think the frustration with the appreciated concept and the unappreciated SOC provider, or the frustration with RYF certification and GPL violations is just a reflect of how bad the computer market is. It is more a market or society failure than a Rhombus Tech or FSF failure.