an Open Source Hardware Certification

10 respuestas [Último envío]
tonlee
Desconectado/a
se unió: 09/08/2014
lembas
Desconectado/a
se unió: 05/13/2010

I think this is a very interesting development although it's unfortunate they use the "open source" lingo [1] and even point to OSI etc...

All this wonderful free software we have is useless unless we have machines to run it on.

1 https://gnu.org/philosophy/open-source-misses-the-point.html

muhammed
Desconectado/a
se unió: 04/13/2013

RYF already does this right? It may be good to have another certification program, but it depends on the criteria. One of OSHWA's primary goals is to "Move towards common expectations of what qualifies as open source hardware, including how non-open elements of putatively open source hardware is handled."

So, what will "open" mean and how will the OSHWA handle non-free components? It's possible that OSHWA will marginalize liberty issues (and RYF) with the media's assistance, like the OSI did to liberty issues (and the FSF) in the 1990s. On the other hand, it's possible that the OSHWA will extend the public's understanding of "open" into liberty topics.

Edit: good chance I'm wrong; I can't say that I have a good grasp on the hardware subject yet, and I only skimmed the OSHWA website

jxself
Desconectado/a
se unió: 09/13/2010
Mzee
Desconectado/a
se unió: 07/10/2013

I don't think it's a duplicate as there is a difference between "Open Hardware" (i.e. hardware whose interiors and circuits are known and well described) and hardware which "just" respects your freedom.

jxself
Desconectado/a
se unió: 09/13/2010

"hardware whose interiors and circuits are known and well described"

Best not to get confused here. "Open Hardware" doesn't actually mean that. Maybe that's what you think of with the term, but as has been pointed out in this thread, the actual definition of their so-called "Open Source Hardware Certification" is still up for definition. So it'll be necessary to see and wait if that make it in or not.

"which 'just' respects your freedom"

It seems you're trying to present the so-called "Open Source Hardware Certification" as something more fundamental that the RYF program. I disagree. If you've kept up on discussions, the FSF has indicated their willingness to go in that direction themselves, pending discussions of how to do it well (there are definitional problems going in both directions.) So the so-called "Open Source Hardware Certification" would remain a duplicate in any event.

Mzee
Desconectado/a
se unió: 07/10/2013

"So it'll be necessary to see and wait if that make it in or not."
True that but that would be MY definition of truly open hardware. ;-D

"It seems you're trying to present the so-called "Open Source Hardware Certification" as something more fundamental that the RYF program."
Right now this is pure speculation but once a definition has been developed which allows free hardware (not free of charge but free in the sense of free software) I indeed think it would be much more fundamental than the RYF program is right now. Provided this hardware would be made available we would, for the first time ever, have to chance to have a truly free computer system. Including all hardware and software parts. Unfortunately, we are still very far away from such a system right now and all we have is free software and a free BIOS on some mainboards.

muhammed
Desconectado/a
se unió: 04/13/2013

Mzee, it sounds like you are describing an ideal that (people here say) doesn't exist today. To realize the ideal, I think that we must have some decentralized way of manufacturing free designs. It's analogous to us having the opportunity today to compile software ourselves, from source code (but maybe not necessarily as decentralized as this).

It would not matter if most people will not know how to evaluate free hardware designs. So long as some people evaluate the free hardware design. Like how only some people today know how to read source code and evaluate it. Those people can warn others if they notice something amiss in the hardware design, just like community members sometimes warn us about bugs or privacy issues in libre software.

You can read about free hardware desin here:

https://www.gnu.org/philosophy/free-hardware-designs.html

Michał Masłowski said that no free hardware exists these days. He described free hardware as "anyone may audit/study all aspects of the hardware, share/modify designs, and manufacture."

https://trisquel.info/en/wiki/hardware-type-and-software-freedom

I assume that Michal's standard for free hardware is the same as the FSF's standard for free hardware design (please correct me if I'm wrong).

Does Michal's definition of free hardware match yours, Mzee?

According to Michal's standard, there are (apparently) no free hardware on the market today. How is OSHWA going to "move towards common expectations of what qualifies as open source hardware"? They will only find hardware to certify if they choose a lower standard than Michal's.

Can the OSHWA meet your personal wants for hardware if their definition turns out lesser than Michal's? You said that you want:
1. "hardware whose interiors and circuits are known and well described"
2. "have to chance to have a truly free computer system. Including all hardware and software parts. "

I hope that I didn't misrepresent anything here (especially Michal's view). Please correct me if I got something wrong.

Mzee
Desconectado/a
se unió: 07/10/2013

Hello muhammed,

"Does Michal's definition of free hardware match yours, Mzee?"
It does but I don't see why OSHWA would have to come short of this definition. They could, for example, create several "levels" of freedom and the highest level of freedom would be achieved by just following Michal's definition. Thank you for that link, btw!

P.S: Unfortunately, Michal is very right when he states that there is no such thing as truly free hardware these days.

Chris

I am a member!

Desconectado/a
se unió: 04/23/2011

I doubt anything is going to change for the better as a result of this effort. They're already showing a willingness to compromise on users freedom from the sounds of it. It's based on a foundation that misses the point ("open source"). We're already seeing companies like Intel, Atheros, and others taking steps backwards.

It's time to start designing hardware from "scratch" to whatever extent that's possible and figuring out how we can cut corners to bring down the manufacturing costs (not necessarily worse computers- just lower costs- as you can't manufacture 200,000 of something for a user base of 5,000).

Once you've achieved that design then work needs to be done to promote the effort and raise funds for actual manufacturing. Even $250,000 is a lot of money particularly when you don't have numbers to build off. That is if you know you can sell 250 of a given computer then it's easy to formulate a plan to manufacture 250 computers. But when you've been selling x86 systems of three different classes and the most popular class only has 120 computers and contains a 1920x1080 screen, etc. then all of a sudden it's not so clear that the cheaper lower end system you're able to design would have enough demand to meet the 250 you'll need to manufacture. A wrong move here and there goes the business. It's one thing for a business to seed the development, but it's another to take that risk, if the effort fails.

So the first thing that would need to happen is that risk needs to be spread out. That could be done through a well-marketed campaign, but still, you need to hit that 250 number. Otherwise it will fail no matter what.

Unfortunately I have no idea for a solution on some issues. Like wifi. All next-gen wifi chipsets are dependent on proprietary software and a good chunk of the systems on the market now lock users into these chips so you can't swap them for ones that are free software friendly.

You can for the moment skip the wifi as its not designed into the SOC, but what happens when it is? Then it becomes even more difficult. In theory you could just not utilise it maybe. An internal USB port and an older wifi chip might work. However you can't get older wifi chips manufactured due to quantities required and significant increases in manufacturing costs. Do you want to pay $250 for a wifi chip? That's assuming you have demand at $250 a chip for at least 100,000 chips a year, maybe more.

tonlee
Desconectado/a
se unió: 09/08/2014

I think the goal for now is a free software notebook mainboard. If fx I could get a newly manufactured libreboot x200 mainboard, I would get one.
I agree, you will not get non knowledgeable buyers to buy a free software computer, especially not because it will likely be more expensive. It has to be organized within the circles of free software users. Even at a somewhat higher mainboard price, I think it will have sufficient buyers if a campaign is executed. About the remaining librem 15 non free software, can it be reverse engineered? What if crowdfund money is at hand? Or is the encryption unbreakable?