How close does the Pyra's design come to respecting your freedom?

[diligentcircle]

It seems to me that banning convient access to not-completely-100%-freer-than-free-software fails to respect my freedom to run non-free software.

I wouldn't call a requirement for an endorsement a "ban", but I digress. It's not convenient access to proprietary software that prevents a vendor from getting RYF certification, but a recommendation of that proprietary software in either the advertising for the device or the software meant to run on the device.


It's even perfectly possible to comply with the RYF requirements for some people, but not for others. I think someone mentioned in that other thread the possibility of having multiple "editions", but you probably wouldn't even need to go that far; you could just as easily have a special "libre" version of the site, like Think Penguin does.

 

[Grench]

It seems to me that banning convient access to not-completely-100%-freer-than-free-software fails to respect my freedom to run non-free software.

Suppose the Pyra ships with DBPManager, configured to use repo.openpandora.org

If I want free software, I can start DBPM, search and hit the big download button.

If I want non-RMS-approved software, I first need to muddle around the forums, then poke config files, eventually adding "evilrepo.openpandora.org" to DBPManager's repo list.

As much as I dislike having non-free-blobs forced on me by big corporations, I also resent  having the FSF's set of ideals shoved in my face.

 

^ hear hear!

Exactly.

FSF: Your device will be censored to only provide you with options that are agreeable to the FSF's policies and dogma - then can be called 'Free'. If your device cannot conform, then it will be forbidden to wear the FSF seal of approval. FSF seal of approval also requires a payment of ....

They loose me right where they claim censorship is required for freedom. By the time they get to requiring payment for licensing logos to claim 'free', that's where I get the giggles.

 

[Neelix]

I wouldn't call a requirement for an endorsement a "ban", but I digress. It's not convenient access to proprietary software that prevents a vendor from getting RYF certification, but a recommendation of that proprietary software in either the advertising for the device or the software meant to run on the device.

Except that as you've already outlined, allowing such convenient access anywhere where it might be construed as official counts as a "recommendation" in the eyes of the FSF, which means you have to make such things inconvenient in order to get the certification.

It's even perfectly possible to comply with the RYF requirements for some people, but not for others. I think someone mentioned in that other thread the possibility of having multiple "editions", but you probably wouldn't even need to go that far; you could just as easily have a special "libre" version of the site, like Think Penguin does.

Can't have multiple editions either... In order to do that you would need to advertise the non-RYF-certified "edition" under a completely different product name with no indication that it's really the same thing just without the certification.

Honestly I personally think the FSF are shooting themselves in the foot. By making their certifications so unattractive for use in real world commercial settings they are actually doing their cause more harm than good.

-Neelix

 

[WizardStan]

ScummVM doesn't need the binaries, just the data files

The script files are proprietary binaries, they're compiled byte code. They are part of "the data files" and the game won't run without them. If SCI games count as "free" because they can be played with ScummVM then all Java applications would by similar definition be free because they can be run in a free virtual machine. Obviously that's not true.Furthermore, even just talking about "the data" as the graphics and audio, those are most definitely restricted by hard copyright, you have absolutely no freedom to modify and especially no freedom to redistribute them.

it depends on what kinds of things are in those scripts

It really doesn't though. They are code that you are not authorized to modify or even read, it doesn't matter how simple or complex they are. I brought up the "reverse engineering" part because it is very relevant: there was no sanctioned document, no official support for the file formats involved, everything was discovered by disassembly and experimentation. The formats are closed and proprietary which is the very reason it needed to be implemented via reverse engineering. If precompiled proprietary bytecode via a free interpreter is fine then what about precompiled fully open machine instructions? ie, you should be fine running anything through an x86 emulator, as you *just* need the "data files" (which includes compiled machine instructions).

For example, I've looked at Ren'Py's scripts, and those aren't a problem.

Even if I put "copyright 2015, not authorized for modification or redistribution" in them? The source may be open but as the copyright holder I'm telling you not to modify nor redistribute my code. That would definitely make it non-free.

Nouveau is a product of reverse-engineering, but it would be silly to reject that.

Nouveau is self contained; if it still needed a prioprietary firmware blob to run would it still pass? I think not.

It seems to me that banning convient access to not-completely-100%-freer-than-free-software fails to respect my freedom to run non-free software

I need to defend Onpon on this, unfortunately. I may disagree with the FSF and think that RYF is most definitely the wrong thing for us, but I do recognize its purpose and see that there can be no other way to accomplish what it is they want to accomplish. One more time: it restricts the company so that it can guarantee the freedom of the user. RYF certification means that if a company sells a product claiming it can perform some tasks (even if it is a vague "look at all this software it can run") then the end user can be guaranteed that these tasks can be accomplished with totally free software.Like suppose EvilDragon used NDS emulation as a selling point, and someone bought the Pyra because of that only to find out that the only viable DS emulator is Exophase's closed source one, and supporting closed source software is something this person does not want to be part of. Sucks to be them? They should have investigated further? Yes, yes indeed. !OR! if the company itself was also of this mind they could apply for the RYF certification and do everything they are requried to in order to ensure the end user can purchase without needing to research every little nuance.

Very very VERY important distinction here: RYF is 100% voluntary! The company must choose to follow the restrictions, doing all the research on behalf of their users to ensure all software is free. If EvilDragon believed in the ideals of the FSF as strongly as Onpon then he could follow the recommendations, achieve RYF certification, and everyone who purchased it would be confident that their freedoms are entirely respected. Now, this would mean certain restrictions, it would mean that EvilDragon couldn't make any sort of official recommendation, including hosting, of non-free software, but again, this being an entirely voluntary thing, EvilDragon would only go through with it if he already felt strongly against such closed source applications.

So the RYF is not useless, it is not unecessarily restrictive, it is a guarantee to the user that the company that has gotten this certification believes in "freedom" as strongly as the user they are appealing to, so strongly that they are forgoing any proprietary software.

This does not apply to the Pyra project, this is not something that EvilDragon wants to do, this hardline view is not inline with what he wants to do, therefore the RYF is not something he should even be considering. And that is perfectly acceptable. Anyone who says otherwise is a zealot who doesn't live in the real world. In the ideal world everything would be "free" but that will never happen.

Consider it like food that is certified kosher, or something like that. Imagine a Muslim person who wants to eat a lasagna. They can buy a pre-made one and research the company, look at the ingredients, make sure everything is ok for them to eat; or they can buy individual ingredients, again making sure they're getting only what they can eat, and make it themself; or they can just buy the one with labeled "halal" and not have to worry about it, safe in the knowledge that the company has accepted the restrictions on their diet and done all the necessary research on their behalf.

That's what RYF is, it's the "kosher" or "halal" of the computing hardware world. A company takes the restrictions on behalf of users they know care about such things. It is no more the FSF shoving anything in our faces than the vegetarian aisle at the grocery store is calling you a murder: if you don't agree with it, don't do it.

edit: just to drive the analogy home, saying "I want RYF certification but I don't want to adhere to the restrictions" is like saying "I want to be a vegetarian but I still want to eat bacon".

 

[Binky]

Surely what you're talking about is clarity of labelling?

As long as all advertising is clear that features x, y, and z depend on closed-source lumps while a, b and c use open source software, then users know exactly what they're getting.

If the Pyra is a cafe, then your food logic says that putting steak on the menu is infringing my right to be a vegitarian.

Freedom is about giving the individual choice, so an ideally free device would let the user decide what software to run. It would offer, but not depend on, closed-source software.

RYF is not about giving the user freedom any more than a vegan cafe is about my right to eat whatever I like. Both favour a non-mainstream set of ideals.

Advertising wonderful pyra software, then slipping in a draconian EULA at the last minute is bad, but deliberately outlawing clearly labelled non-free software is just as bad as Apple-style walled gardens.

 

[levi]

I can see the benefit of RYF certification for desktop PCs where I decide on the exact components inside it, and to a certain extent on laptops.  I generally go for Intel graphics for example on my PCs because I'm in charge of the OS on those machines and have found binary blobs less reliable than open source code, and more of a pain to configure.  On a machine like the Pandora/Pyra though, where others are in charge of configuring the OS (assuming I continue to use the stock OS), and where the choice is largely between binary blobs or no device at all, I'll take binary blobs.

 

[WizardStan]

If the Pyra is a cafe, then your food logic says that putting steak on the menu is infringing my right to be a vegitarian.

Not at all.If the Pyra is a cafe and it puts steak on the menu then it can't call itself a vegan cafe. It's as simple as that. If the Pyra wants to voluntarily call itself a vegan cafe then it needs to ensure there are no animal products in its menu. Plain and simple. The vegan customers that the Pyra cafe would be appealing to would then be able to eat there knowing that the cafe has done everything in its power to ensure they don't accidentally eat a piece of meat which would violate their beliefs.

This is an entirely voluntary thing. Someone said "hey, wouldn't it be great to get vegan certification? That's an entirely untapped market!" and instead of saying "no, we can't conform to the strict requirements of veganism" people are complaining "Vegan certification is so stupid, you can't even serve bacon! What if someone does come in wanting bacon?" and the answer is, if you aren't a hardcore vegan yourself and have no problem serving meat and animal products to customers then don't become a vegan certified cafe. Period. So simple.

The RYF is a guarantee to the end user that everything that has been promised in any way by the company can be reasonably accomplished with just "free" software. It is restrictive in exactly the same way that a vegan cafe can only serve fruits and veggies, it is a restriction that the cafe takes on voluntarily to cater to a very specific group.

RYF certification caters to a very specific group. If EvilDragon felt as strongly about "freedom" as Onpon then it would be a no-brainer and he'd go for it, restrictions and all, because that's what he believed in. He doesn't, we don't, it is not a good fit for the community, so where's the point in considering it?

 

[diligentcircle]

The script files are proprietary binaries, they're compiled byte code.

Hm, I'll have to investigate this at some point. Thanks.

Furthermore, even just talking about "the data" as the graphics and audio, those are most definitely restricted by hard copyright, you have absolutely no freedom to modify and especially no freedom to redistribute them.

As I say on my website, I oppose copyright and non-libre culture, but think it would be senseless (and even destructive) to refuse to view, read, or otherwise consume non-libre works, with the exception of software (software is an exception because it controls how your computer runs).

That's why I said that it matters what kinds of things are in those scripts, and that I determined that Ren'Py's language isn't something to worry about; I looked at a Ren'Py script and it seemed that the instructions were all trivial. I could have gotten the wrong impression, though.

If precompiled proprietary bytecode via a free interpreter is fine then what about precompiled fully open machine instructions? ie, you should be fine running anything through an x86 emulator, as you *just* need the "data files" (which includes compiled machine instructions).

You raise an interesting point. I'll have to re-think this. Thanks.

Surely what you're talking about is clarity of labelling?


As long as all advertising is clear that features x, y, and z depend on closed-source lumps while a, b and c use open source software, then users know exactly what they're getting.

The RYF certification is supposed to be for devices that don't need any proprietary software. If you advertise that the device you're selling can support a feature if you only add a proprietary program, then that means one needs proprietary software to get the full experience.

If RYF allowed devices that work in a capacity less than advertised with libre software, someone would be able to, say, sell a computer with no network support, no sound support, no USB support, and a GPU that can only output text, with libre software, while it supports everything you'd expect on Windows. But, the computer happens to support Libreboot, so it can technically run entirely libre software. They'd be able to advertise office suites, Web browsing, whatever, even though just about nothing useful can be done without Windows. "But you can goof around with Nano in Dragora, so this system respects your freedom!"

 

[Grench]

Comparing RYF's mandates to those of a religion is a pretty accurate analogy.

Having fundamentalist RYFers popping in periodically to demand that the rest of the community be censored from any non-RYF materials and that conformation is required for businesses to be allowed to sell to their 'flock'... that part sounds familiar too.

 

[Binky]

This is implying that there is a "full experience". That the device is an appliance with a finite number of functions.

At some levels, that makes sense, but at many others, it doesn't.

At a driver level, claiming "The Pyra supports wlan" and then it turning out you need binary secrets to use it could be justifiably seen as bad.

At a userland software level, claiming "The Pyra runs thousands of games and applications" and then it turns out that some of them are closed-source, is really not a problem. I doubt anyone has used every pnd out there anyway.

The certification seems to think that there are so few functions that every possible program can have an exact open analogue.

The definition of "reccommend" used is far too broad too.

Making the device usable without closed blobs is "respecting yout freedom"

Hiding all closed-source tools and pretending they don't exist is nannying.

As a separate point, why does the whole device need an umbrella certification for openness? Surely a user who is that picky about thier purchases can read up on what works with open software and what doesn't? Just as a diner in our cafe can look through the menu for the dishes with a little green 'v' next to them.

I really don't see a problem with advertising "Foo, Bar, Baz (baz requires binary blob)"

That way, users can judge whether they still want the device if they are unconfortable with the closed-source baz.

The fact that the device can do baz doesn't make it any worse at  foo or bar.

 

[comradekingu]

If you host a non-free repository, like Debian does, and tell the users "if you want non-free software, you can use this repository", you're effectively recommending that proprietary software, or at least offering to do so. You can disagree with the FSF on this, but the FSF's position is that this is not acceptable, and not doing this is a requirement of RYF.

This would have saved a lot of time with that argument back then.

The actual text is «Our endorsement is strictly conditioned on the seller's discharge of its commitment to ensure that all product software will always be free software. This applies to all software that the seller includes in the product, or provides with the product, or recommends for use in conjunction with the product, or steers users towards installation in the product, except for certain microcode and firmware.»

And it is explained «By way of explanation, a general-purpose facility for installing other programs, with which the choice of programs to install comes directly from the user, is not considered to steer users toward anything in particular. However, the product suggests using the facility to install certain programs, that is steering users towards those programs.»

 

[Neelix]

Consider it like food that is certified kosher, or something like that. Imagine a Muslim person who wants to eat a lasagna. They can buy a pre-made one and research the company, look at the ingredients, make sure everything is ok for them to eat; or they can buy individual ingredients, again making sure they're getting only what they can eat, and make it themself; or they can just buy the one with labeled "halal" and not have to worry about it, safe in the knowledge that the company has accepted the restrictions on their diet and done all the necessary research on their behalf. That's what RYF is, it's the "kosher" or "halal" of the computing hardware world. A company takes the restrictions on behalf of users they know care about such things. It is no more the FSF shoving anything in our faces than the vegetarian aisle at the grocery store is calling you a murder: if you don't agree with it, don't do it.

This is exactly correct... The problem is one of terminology though. Kosher, Halal, and the FSF's ideology are all restrictive ideologies. If the FSF came up with a unique name for their ideology that didn't have other connotations I'm sure there wouldn't be such a problem with it, anyone with no interest in it would just dismiss it as irrelevant. However, the FSF calls it's restrictive ideology "freedom" which to anyone who doesn't subscribe to the FSF's ideology has a completely different meaning.

-Neelix

 

[diligentcircle]

At a userland software level, claiming "The Pyra runs thousands of games and applications" and then it turns out that some of them are closed-source, is really not a problem. I doubt anyone has used every pnd out there anyway.

I very much doubt that statement would disqualify a piece of hardware from the RYF certification. Think Penguin's page on the wireless devices with RYF certification it sells clearly state that the device works with several non-libre GNU/Linux systems, by name, and also mentions that it will work with many others. There are currently only 9 distros that are approved by the FSF as completely libre, so it's obvious that most of the systems Think Penguin is claiming compatibility with are proprietary.

As a separate point, why does the whole device need an umbrella certification for openness? Surely a user who is that picky about thier purchases can read up on what works with open software and what doesn't? Just as a diner in our cafe can look through the menu for the dishes with a little green 'v' next to them.

RYF is meant for completed products, not components. Figuring out what components work with libre software is h-node's territory.


Finding completed products that work with 100% libre software is not easy. Actually, I think the number of computer models that will work with Libreboot or blob-free Coreboot can be counted on two hands.


Even getting computers at the level of Think Penguin's isn't that easy, and it's largely based on luck. You have to contend with useless wireless cards, digital restrictions that prevent you from changing the wireless card, unaccelerated 2-D graphics, and malicious features Intel and AMD like to put in their CPUs every once in a while. And even when you've gotten past that, you still have a proprietary boot loader.

 

[WizardStan]

I looked at a Ren'Py script and it seemed that the instructions were all trivial

So in your mind there's some arbitrary level of complexity before code becomes code and the restrictions someone might place on it start counting?Because RenPy is not at all trivial; it's an API. It is Python with some additional useful functions. You can write incredibly complex applications in Python wrapped up with the Ren'Py API in order to produce some game. It seems very dangerous to try and define code vs code (and therefore worthy of "freedom") based on how complex it "feels" when looking at it.

with the exception of software (software is an exception because it controls how your computer runs)

Define software?What about DVD or BD menus? For all common understanding of software that I'm aware of, they are software. There have been some very complex applications written in DVD script, I'm pretty sure it's turing complete. BD-J is full blown Java, capable of so very much more. In neither case are you able to look at, modify, nor distribute the software. If the opposition is to non-free software then DVDs and BDs are out as well.

Do you want to keep going down this rabit hole until I've got you questioning your very existence? 'Cause it's a long way down from here.

Surely a user who is that picky about thier purchases can read up on what works with open software and what doesn't? Just as a diner in our cafe can look through the menu for the dishes with a little green 'v' next to them.

Yes, you are absolutely correct, this is a thing that CAN be done. Just as a diner is perfectly capable of looking at a menu and choosing what they do or don't want to eat a user is perfectly capable of looking things up themselves and deciding whether to proceed or not.That's not the point. If that were the point then pizza parlours wouldn't advertise themselves as halal, and vegan restaurants wouldn't exist. Everyone would happily go into any restaurant, research every ingredient on the menu, and pick the one that didn't conflict with their beliefs.

The point is that by labeling themselves, either halal or kosher or vegan or whatever, they are telling their customers that they don't have to do all that work, they don't have to do massive amounts of research just to know their burger was prepared according to whatever edict they need to follow. The company is promising they've done all this work for them.

And that's what RYF is. It. Is. Voluntary. I don't know how many times I need to say this. It is something that the company willingly chooses to adhere to in order to appeal to a very select set of customers, the same way that a cafe will label itself vegan. If you do not believe 100% in the things that the FSF stands for and what the RYF pushes then you simply do not get RYF certification. It really is that simple. No one is placing restrictions, these are simple edicts that the FSF have laid out to be RYF compliant, the same as there's an understanding of what vegetarian, or halal, or even lactose intolerant.

Like seriously, imagine you go to a restaurant and they have "lactose free ice cream" and you're like "hell yeah! I'm lactose intollerant, I haven't had ice cream in a while, I'mma gonna try it!" and then you find out that "lactose free" was apparently just a marketing gimick and you're sitting on the toilet for the next million years? Pretty crappy, eh? "Oh, you should have done your research and then you'd've known that it wasn't really lactose free" no that's bullshit. If a company says "lactose free" you expect it to not contain lactose. If a company says "RYF certified" then you expect all software they promote to be libre free, that is it's definition. If you cannot meet the definition then you do not call yourself RYF certified, you do not call yourself vegetarian, you do not call yourself whatever. It's actually really simple.

The certification seems to think that there are so few functions that every possible program can have an exact open analogue.

No it doesn't, it's just saying that what you advertise the machine to be able to do it needs to do so with "free" software. Just as there's no true vegetarian analog to bacon, so to there are going to be no "free" alternatives to some applications. This is not a failing of the certification, the certification serves a very specific purpose and it does so quite well. If your device needs non-free software, or you believe it is suitably enhanced by this non-free software (in much the same way that everything is improved with bacon) then you don't call yourself RYF certified. No problem.

 

[bzar]

I may disagree with the FSF and think that RYF is most definitely the wrong thing for us, but I do recognize its purpose and see that there can be no other way to accomplish what it is they want to accomplish. One more time: it restricts the company so that it can guarantee the freedom of the user. RYF certification means that if a company sells a product claiming it can perform some tasks (even if it is a vague "look at all this software it can run") then the end user can be guaranteed that these tasks can be accomplished with totally free software.

I was with this up to the point when community-made software that did not adhere to the restrictions meant for the company became a problem. The moment the restrictions start leaking from the company to the users is where I draw the line.

 

[_wb_]

At this point, given the limited choice in SOCs/GPUs and wifi chips, a fully RYF-certified handheld seems hard to do, just in terms of hardware drivers.

The "vegetarian" analogy is quite nice. The FSF are hard-core vegans who only eat fruits that fell of the tree naturally without harming the tree. Most people eat meat and don't really care in which manner the pigs and cows are bred and slaughtered, as long as they can buy conveniently packaged steaks and sausages cheaply in their supermarket.

There is a spectrum here, and there are a lot of intermediate positions. For example, both in dietary terms as in the FSF analogy, my position is to be a vegetarian. I like cheese and eggs too much to not eat them, and I don't care about "hurting" plants. But I still don't want to eat dead animals. In particular, I really don't like it if I go to a restaurant, order a vegetarian lasagna, and find little pieces of meat in my mouth, because the restaurant chef seems to have the idea that "vegetarian" means "includes a larger than usual amount of vegetables".

So I don't think we should try to go for RYF certification (it would be impossible anyway), but I do think we should try to be better at labeling things. Now I have the impression that the words "Free Software" and "Open Source" are thrown around a bit easily, while it is rather hard to tell which parts of the Pandora (its OS, and its default software repo) are actually FOSS and which are not. There is no easy way to take the "vegetarian" option. In Debian, there is a clearly marked "non-free" part of the repository, and you can easily disable it. I would suggest that in the DBP repo (or whatever it will be called), we do the same: have a mandatory field in the package metadata that says whether the software is Free (FOSS), Gratis or Commercial.

 

[comradekingu]

Consider it like food that is certified kosher, or something like that. Imagine a Muslim person who wants to eat a lasagna. They can buy a pre-made one and research the company, look at the ingredients, make sure everything is ok for them to eat; or they can buy individual ingredients, again making sure they're getting only what they can eat, and make it themself; or they can just buy the one with labeled "halal" and not have to worry about it, safe in the knowledge that the company has accepted the restrictions on their diet and done all the necessary research on their behalf. That's what RYF is, it's the "kosher" or "halal" of the computing hardware world. A company takes the restrictions on behalf of users they know care about such things. It is no more the FSF shoving anything in our faces than the vegetarian aisle at the grocery store is calling you a murder: if you don't agree with it, don't do it.

This is exactly correct... The problem is one of terminology though. Kosher, Halal, and the FSF's ideology are all restrictive ideologies. If the FSF came up with a unique name for their ideology that didn't have other connotations I'm sure there wouldn't be such a problem with it, anyone with no interest in it would just dismiss it as irrelevant. However, the FSF calls it's restrictive ideology "freedom" which to anyone who doesn't subscribe to the FSF's ideology has a completely different meaning.


-Neelix

The problem with your argument is that you don't consider more alternatives to be more free.

Proprietary software with all the restrictions that brings on the user is already available. That level of vendor freedom is already there beforehand, no change differentiates nothing. GPL is voluntary, and its protected by the same copyright laws, if you consider provisions to be irrelevant to the work put in a license, you only give something to the user in a meaningful way.

Copyleft is the word you are looking for. The FSF calls BSD just as free as its GPL, so that isn't the argument. Both of which are voluntary on the user-end. Thinking one has a more exclusive set of restrictions, is a valid reason to decide against, or between them, depending on how you see things.

It is not restrictive, but natural for free to be a baseline spec, fair is the word to use between parties.

Free (as in freedom) is a subjective bias, you have to want some level of freedom which you think is not available to you. On a non metaphysical level that means freedom from restrictions that someone else imposes on you.

To label all of these attempts as restrictive, is a level of conservatism reserved for only those who think history to be stagnant, or having arrived at best / for the time being / finality.

The only footing your argument has, is for free to also mean provisions greater than those already granted by current laws, and in particular copyright, for the vendor.

To that I can only say that for a balance where vendor doesn't lose its meaning, its free for less people. But its by no means exclusive, you are free to do as you like.

It is misguided to assume software freedom gives you user freedom in every aspect of the word freedom.

The FSF is very clear about this, in that FSF says political freedom can not be achieved with the GPL alone. Check out Richard Stallman, who originated it, he is a political activist in many ways which aren't the GPL.

Even if FSF coined the term Free software, you can start using it differently at will. To be fair they say free, the (as in) freedom, is only there to qualify the free, which is needed in English since it shares meaning with gratis. All the freedoms state explicitly "the user" since that is whom they are for.

And you don't really get to pick if you want to have an ideology, unless you want your brand to be meaningless to this space, "its creative commons" is a prime example of that.

 

[sebt3]

As far as I can tell there is no accessible ARM SoC that would match RYF criteria.

maybe an intel SoC could match such criteria.

But imho, if you're in the market for a RYF device, the pyra isnt for you. What are you going to do with the nubs and gaming buttons ? After all (nearly) all games worth using these are copyrighted...

 

[comradekingu]

Curiously for RYF, that is actually opposite.

«The exception applies to auxiliary processors or low-level processors, none of whose software is meant to be installed or changed by the user or by the seller. This can include, for instance, microcode inside a processor, firmware built into an I/O device, or code compiled into an FPGA. The software in such auxiliary and low-level processors does not count as product software.

We want users to be able to upgrade and control the software at as many levels as possible. If and when free software becomes available for use on a certain auxiliary processor, we will expect endorsed products to adopt it within a reasonable period of time. This can be done in the next model of the product, if there is a new model within a reasonable period of time. If this is not done, we will eventually withdraw the endorsement.»

AFAIK only modern intel processors ship user installable microcode updates. What something is meant to do is up for debate. In my opinion hiding something in firmware is not good, and the distinction to user freedom is just as diverse/clear cut as software, as soon as you can start to license something, it matters.

 

[WizardStan]

I was with this up to the point when community-made software that did not adhere to the restrictions meant for the company became a problem. The moment the restrictions start leaking from the company to the users is where I draw the line.

Then don't get RYF certification. It's that simple.And it isn't that community-made software is a problem for the company making the device, there is absolutely nothing in the RYF that says the community cannot make its own software. THAT would be counter to the FSF's ideals. The problem for the company only comes when the company does anything that looks like it endorses non-free software.

Imagine you're a hardcore vegetarian and you start going to our Pyra-cafe because it calls itself vegetarian, and then you find out they do keep bacon in the freezer, just in case anyone asks: they don't really advertise it, maybe the servers occasionally slip that they have somthing "extra". Regardless, if you're a hardcore vegetarian and you find out your favourite "vegetarian" restaurant stocks bacon you will feel betrayed, especially if you manage to hear about this bacon and assume, because this is a certified vegetarian restaurant and therefore everything they serve must be vegetarian, that it is some kind of veggie bacon, so you order it just to try it. And when it turns out it isn't, you can no longer trust that "vegetarian" identification.

RYF protects the user from this. If a company really believes 100% in the FSF and wants to get RYF certification then they will have no problem following the requirements. We cannot, nor should we try, to follow the requirements, therefore the Pyra cannot be RYF certified. Move on.

edit: to again drive the analogy home, the vegetarian restaurant can't serve bacon and call itself vegetarian, but if you order a veggie sandwhich, walk out the door, and put bacon on it from a nearby bacon vendor then there's no problem, the cafe is still vegetarian and you, as the end user, still have the right to bacon. The cafe just can't serve bacon itself and still call itself vegetarian.