"Respects Your Freedom" certification?

[Neelix]

What people don't really take into consideration here is that the "freedom" to run proprietary software and advertise it as such isn't straight-forward.

We all saw what happened to that steam x86 "console" that claimed certification with steam before the whole thing was finalized. Lo and behold it wasn't supported by valve in the end. And if one ships with Android, one gets sued by Microsoft. I dare you to name me any single supplier who doesn't pay that extortion tax.

Well I wouldn't recommend entering an exclusive advertising arrangement like that with a proprietary software company on an open device either, but that doesn't mean that there isn't worthwhile non-free software for which a recommendation wouldn't come with such strings, or that there won't be such software in the future in the software ecosystem of the new device.

Claiming any certification before it is finalised would seem to be a stupid move, but I don't see how that is relevant here.

- Neelix

 

[comradekingu]

Because supposedly the doors are open at valve, and that gives the company no power to do what one fears FSF might do? When in fact its much the same, plus there is an EULA to consider, NDAs, license etc. On top of whatever goodwill you might source from the company.

The corporate world of today doesnt play with open cards, whereas im pretty sure if you sent FSF, EFF, or OSI an email you would get a straight answer and could be able to share it with the community to see how the wind blows.

TL;DR To sum it up it seems everyone was in favour of non broken by design hardware, and that a certification is up for discussion. Then whatever happens to the software is a continuous discussion, insofar as its decided on the fly and that it has to do with the hardware. We could start now by making the steps that are better for everyone, like stating the license for every part of software there is on the repo and having it available as per what one would expect instead of email this, bundled here, link to git there. It isnt that that needs to be superstrict, but some guidelines are good to have. The wiki is a nice place to document it.

 

[Binky]

We could start now by making the steps that are better for everyone, like stating the license for every part of software there is on the repo and having it available as per what one would expect instead of email this, bundled here, link to git there. It isnt that that needs to be superstrict, but some guidelines are good to have. The wiki is a nice place to document it.

Why not make license (and copyright holder) mandatory fields in the pxml?

 

[Neelix]

Because supposedly the doors are open at valve, and that gives the company no power to do what one fears FSF might do? When in fact its much the same, plus there is an EULA to consider, NDAs, license etc. On top of whatever goodwill you might source from the company.

The corporate world of today doesnt play with open cards, whereas im pretty sure if you sent FSF, EFF, or OSI an email you would get a straight answer and could be able to share it with the community to see how the wind blows.

I'm afraid you've completely lost me here. I can't make heads nor tails of what point you're trying to make, nor why you would think Valve's EULA is in any way relevant to what I've been saying.

Put simply I don't advocate allowing any third party, be it Valve, the FSF or anyone else for that matter to dictate terms that restrict how ED can advertise either his product, or what other software can be used with it.

TL;DR To sum it up it seems everyone was in favour of non broken by design hardware, and that a certification is up for discussion. Then whatever happens to the software is a continuous discussion, insofar as its decided on the fly and that it has to do with the hardware. We could start now by making the steps that are better for everyone, like stating the license for every part of software there is on the repo and having it available as per what one would expect instead of email this, bundled here, link to git there. It isnt that that needs to be superstrict, but some guidelines are good to have. The wiki is a nice place to document it.

That seems a reasonable course of action in any case.

- Neelix

 

[comradekingu]

I'm saying the freedom to advertise proprietary software is upholding the non-free means of valve allowing you to publicize their content, which is worse than the FSF. It seems the arguments against FSF supposed control can be applied to the "freedom" that somehow incurs when you ship non-free software as per default, following the tight nit control of the proprietary world.

If its an argument about control, i think its largely based on dogma and not pragmatism. In the real world, the only "restriction" is not being able to restrict users and encumber them with models of business that we cant afford the lawyers for, and that others do already, no? 

I'm sure if there is no under the table horse-trading involved there will be _countless_ companies doing that. _Fully_ free seems like the banner to separate under

Since it was pretty much down to discussion and 'we will see', what is the big deal-breaker now? Shipping emulators by itself might not be the strongest move, since the laws concerning legality seems to change a lot and is different in a lot of not only European, but also world countries.

As for the other things that are on the repo now, they cost money. Im pretty sure those who don't want them don't want to pay the bundle price, even if its a special deal. Why not make it a deal to GPL it and throw some money at the devs in that same token. I guess thats feasible for the newer games, but not the arcade games.

The software development fund was mentioned, thats what i would expect from it, i don't know about others who contributed, but software development for the greater good funded by the community is not far fetched at all.

 

[_wb_]

So to summarize this thread, I think this is the consensus: (of course feel free to correct me)

• When deciding which hardware parts to use, freedom (i.e. availability of FOSS drivers) is an important consideration.
• The software that will ship with the hardware, should be as free as possible; if the hardware allows it, 100% free.
• We want to keep a community software repo like what we have now (maybe with a better indication/separation of licenses), that is, both free and non-free software can be uploaded and downloaded, gratis or commercially
• We want to be able to advertise that the hardware can (optionally) run non-free software (e.g. Android, RISC OS, GNU/Linux distros that include non-free software, free or non-free emulators that emulate free (rarely) or non-free ROMs, proprietary games, maybe Steam if that ever happens, etc), e.g. with YouTube videos showing footage of non-free games running on the device.
• At the same time, we don't want to give the impression that non-free software is required in order to enjoy the device. In fact, we would also advertise the wonders of Free Software running on the device, and emphasize the potential to modify anything and develop on the device itself using GNU and other Free development tools.
• If "Respects Your Freedom" certification can be obtained with the above in mind, then we should go for it.
• Otherwise (e.g. if the certification would restrict advertisement options too much), then we try to meet the criteria to the largest possible extent, but we don't apply for the certification to avoid it backfiring.

 

[Grench]

So to summarize this thread, I think this is the consensus: (of course feel free to correct me)

• When deciding which hardware parts to use, freedom (i.e. availability of FOSS drivers) is an important consideration.
• The software that will ship with the hardware, should be as free as possible; if the hardware allows it, 100% free.
• We want to keep a community software repo like what we have now (maybe with a better indication/separation of licenses), that is, both free and non-free software can be uploaded and downloaded, gratis or commercially
• We want to be able to advertise that the hardware can (optionally) run non-free software (e.g. Android, RISC OS, GNU/Linux distros that include non-free software, free or non-free emulators that emulate free (rarely) or non-free ROMs, proprietary games, maybe Steam if that ever happens, etc), e.g. with YouTube videos showing footage of non-free games running on the device.
• At the same time, we don't want to give the impression that non-free software is required in order to enjoy the device. In fact, we would also advertise the wonders of Free Software running on the device, and emphasize the potential to modify anything and develop on the device itself using GNU and other Free development tools.
• If "Respects Your Freedom" certification can be obtained with the above in mind, then we should go for it.
• Otherwise (e.g. if the certification would restrict advertisement options too much), then we try to meet the criteria to the largest possible extent, but we don't apply for the certification to avoid it backfiring.

Pretty good synopsis.  The bold line above is a bit of a concern IF it means that we have to advertise the 'free' version just as much as the 'non-free' version of the same program.  Frankly, they shouldn't care about free/nonfree in determining which products to use for showcasing and advertising the unit.  Just use the 'best', whichever it may be for the particular feature(s) being showcased.

The whole, 'equal placement and promotion of a free alternative' bit is actually very restrictive to 'free speech'.

The project shouldn't to be 'required' to promote free alternatives any more than it should be 'required' to use non-free software.

Any time someone tries to promote free or freedom by listing a series of restrictions, demands, regulations and criteria that others must follow, I feel they have already compromised the ideal of 'free' as in 'freedom'.

 

[Rockthesmurf]

I would agree with Grench. Make good device. Make good software for device. Advertise good software on good device to get good community.

At no point should 'good' have any relation to free/open/closed/commercial, just show the best we have, if it is free that's great, if it is paid that is also great (some users in the community have actively voted in previous polls to say they want more commercial software).

The only proviso I would say is that if 95% of software is free, then it could be good to state that, I personally feel most users won't really care what the license is, or what badge it has, but they'll be interested to know that 95% + of software is completely free (no adverts, no IAP, no up front cost, completely free). It isn't so much 'taking sides' one way or another, it is telling the user a fact about the distribution of free software out there.

 

[comradekingu]

Its not because "they" want us to, its because we, those that have the time, will do it. Free software is how we differ from other things, it would be strange not to promote that, because not only does it make for a difference, its also (imo) the better way to go.

Free software (as in freiheit) has advantages over freeware that far transgress the monetary cost.

Open discussions like this, free software, friendly hardware.

Edit: please say gratis when you specifically mean free as in beer. I have no qualms with commercial software, its the closed software that brings problems.

 

[Mr_Loon]

@ _wb_ : Good summary, agree 100%

....

And if one ships with Android, one gets sued by Microsoft. I dare you to name me any single supplier who doesn't pay that extortion tax.

I have no proof of course but I'd be utterly astounded if Chinese companies like JXD pay anything to Microsoft for the use of Android. Based on the fact that they often include pirated roms & emulators on their devices it seems rather unlikely to me that they are overly concerned with paying M$ their 'due'.

 

[comradekingu]

http://boards.openpandora.org/index.php/topic/14049-respects-your-freedom-certification/page-5#entry267892  indeed. And that means there is no winning either way.

 

[God Ginrai]

Honestly, this is my opinion on the issue:

• We should try our very best to have FOSS drivers for all of the hardware. I don't think anyone wants to deal with issues our talented community of devs can't fix on their own, and they shouldn't have to reverse engineer something just to be able to fix it
• We should strive towards free software without limiting ourselves. It's not worth it to try for some certification if the process of tip-toeing around it is a detriment to the ease of use, the ability to develop for, or the ability to monetize the product.

-God Ginrai

 

[WizardStan]

We should strive towards free software without limiting ourselves.

Depends on the limit. If there's a wifi chip with open drivers that is rated at 1.5MB/s and a wifi chip rated at 3.0MB/s but only has closed drivers I am going to vote that 1.5MB is fast enough and the speed boost is not worth the potential hassle of dealing with binary blob drivers. If one SoC has a GPU which pushes 60M polygons through a closed driver and another which pushes 50M polygons through an open driver, the extra 10M does not seem worth the trouble I've seen with the Pandora's existing drivers.Where exactly the magic cut off balance between "better" and "free" exists I can't say, but it's such that I think it'd be worth at least keeping some lower spec'd components in mind and not just disregarding them because there's something faster.

 

[_wb_]

• We should strive towards free software without limiting ourselves. It's not worth it to try for some certification if the process of tip-toeing around it is a detriment to the ease of use, the ability to develop for, or the ability to monetize the product.

It is a rather strange thing to say: "strive towards free software without limiting ourselves". Having to rely on non-free software is a serious limit in itself (probably the single most important limit you can have as a user/developer). In my view, "free software" and "no limits" are more or less synonyms.

So I'll interpret the remark in the same way as WizardStan did: you're not talking about "limiting ourselves", but about "limiting our design options". Insisting on having FOSS drivers available, assuming hardware vendors cannot be persuaded to open up their blobs and reverse engineering is too hard or will take too long, would make the list of possible GPUs and wifi components shorter, and it may "limit" us in the sense that while designing the P2, we may have to pick something with lower specs than a non-free alternative. But picking a component with no hope of a FOSS driver, even if it has better technical specs, also limits us in the sense that we have to rely on the hardware manufacturer to get bugs fixed in the driver and to keep it up to date - that could for example force us to use an outdated X version or to not be able to use different kernels than the ones supported by the component vendor.

Hardware specs will become "outdated" anyway, there is no way around that. Software can stay alive much longer - but not if all we have is a binary blob. If we want the P2 to be a kick-ass little machine, with a long-term community, then non-freedom is a way bigger limitation than technical specifications that are somewhat below cutting edge at release date.

This discussion is moot anyway unless someone has evidence that the hardware choices with only blobs (e.g. PowerVR) have better specs than the hardware choices for which FOSS drivers exist. As far as I can tell, this is not at all the case.

 

[God Ginrai]

We should strive towards free software without limiting ourselves.

Depends on the limit. If there's a wifi chip with open drivers that is rated at 1.5MB/s and a wifi chip rated at 3.0MB/s but only has closed drivers I am going to vote that 1.5MB is fast enough and the speed boost is not worth the potential hassle of dealing with binary blob drivers. If one SoC has a GPU which pushes 60M polygons through a closed driver and another which pushes 50M polygons through an open driver, the extra 10M does not seem worth the trouble I've seen with the Pandora's existing drivers.Where exactly the magic cut off balance between "better" and "free" exists I can't say, but it's such that I think it'd be worth at least keeping some lower spec'd components in mind and not just disregarding them because there's something faster.

 

• We should strive towards free software without limiting ourselves. It's not worth it to try for some certification if the process of tip-toeing around it is a detriment to the ease of use, the ability to develop for, or the ability to monetize the product.

 It is a rather strange thing to say: "strive towards free software without limiting ourselves". Having to rely on non-free software is a serious limit in itself (probably the single most important limit you can have as a user/developer). In my view, "free software" and "no limits" are more or less synonyms.

So I'll interpret the remark in the same way as WizardStan did: you're not talking about "limiting ourselves", but about "limiting our design options". Insisting on having FOSS drivers available, assuming hardware vendors cannot be persuaded to open up their blobs and reverse engineering is too hard or will take too long, would make the list of possible GPUs and wifi components shorter, and it may "limit" us in the sense that while designing the P2, we may have to pick something with lower specs than a non-free alternative. But picking a component with no hope of a FOSS driver, even if it has better technical specs, also limits us in the sense that we have to rely on the hardware manufacturer to get bugs fixed in the driver and to keep it up to date - that could for example force us to use an outdated X version or to not be able to use different kernels than the ones supported by the component vendor.

Hardware specs will become "outdated" anyway, there is no way around that. Software can stay alive much longer - but not if all we have is a binary blob. If we want the P2 to be a kick-ass little machine, with a long-term community, then non-freedom is a way bigger limitation than technical specifications that are somewhat below cutting edge at release date.

This discussion is moot anyway unless someone has evidence that the hardware choices with only blobs (e.g. PowerVR) have better specs than the hardware choices for which FOSS drivers exist. As far as I can tell, this is not at all the case.

It's funny how you both completely disregarded my first point to address my second. Had you taken the first point into account, you would realize that I do want FOSS drivers, and consider it limiting to have closed drivers.

When I asked for us to not limit ourselves, it was with the intention that we don't sacrifice functionality, usability, or credibility in order to get some certification. Making EvilDragon walk on eggshells just so he doesn't lose said certification is certainly a limiting factor, and if that's what he would have to do, then we don't need this certification. As for usability, if we make it harder for the users to find their information or applications, then we are limiting ourselves. Honestly, the debian issue pointed out earlier in this topic illustrates for me pretty well that this certification is something we do not need, because the people who enforce it are too strict for a product like the Pandora.

Also, while I would love to have FOSS drivers, we need to look realistically at what EvilDragon can get. It is hard enough to get SoCs in the small quantities that the Pandora needs. Because of this, we're more than likely going to have to deal with whatever he can get. However, if he can get a choice of FOSS or not on some SoCs, then I am all for the FOSS driver SoC as long as there are not major performance gains to be had by the closed driver SoC.

-God Ginrai

 

[WizardStan]

It's funny how you both completely disregarded my first point to address my second. Had you taken the first point into account, you would realize that I do want FOSS drivers, and consider it limiting to have closed drivers.

Sorry, I didn't mean to make it seem as if I were disregarding your first point. It was perfectly clear and I didn't think it needed addressing, I agreed with it entirely. Your second point seemed a little ambiguous, especially the way you emphasized limited. I just wanted to make sure we were talking about the same types of limits, since your two points taken together do not necessarily define what you consider to be limits. When in doubt it is always better to clarify.

 

[_wb_]

When I asked for us to not limit ourselves, it was with the intention that we don't sacrifice functionality, usability, or credibility in order to get some certification. Making EvilDragon walk on eggshells just so he doesn't lose said certification is certainly a limiting factor, and if that's what he would have to do, then we don't need this certification. As for usability, if we make it harder for the users to find their information or applications, then we are limiting ourselves. Honestly, the debian issue pointed out earlier in this topic illustrates for me pretty well that this certification is something we do not need, because the people who enforce it are too strict for a product like the Pandora.

Sacrificing functionality and usability to get certification: it depends on what you mean with that. Getting certification means it cannot ship with non-free stuff pre-installed, so if you consider that a sacrifice of usability (after all, "already pre-installed" is more user-friendly than "have to install manually"), then it is a problem. Credibility: in my opinion, getting the certification increases the credibility of the project, but some people may disagree so much with the FSF that for them, getting FSF endorsement decreases credibility.

 

[God Ginrai]

When I asked for us to not limit ourselves, it was with the intention that we don't sacrifice functionality, usability, or credibility in order to get some certification. Making EvilDragon walk on eggshells just so he doesn't lose said certification is certainly a limiting factor, and if that's what he would have to do, then we don't need this certification. As for usability, if we make it harder for the users to find their information or applications, then we are limiting ourselves. Honestly, the debian issue pointed out earlier in this topic illustrates for me pretty well that this certification is something we do not need, because the people who enforce it are too strict for a product like the Pandora.

 Sacrificing functionality and usability to get certification: it depends on what you mean with that. Getting certification means it cannot ship with non-free stuff pre-installed, so if you consider that a sacrifice of usability (after all, "already pre-installed" is more user-friendly than "have to install manually"), then it is a problem. Credibility: in my opinion, getting the certification increases the credibility of the project, but some people may disagree so much with the FSF that for them, getting FSF endorsement decreases credibility.

My concern is with how hard it may be to obtain this license, and how easy it may be to lose it. EvilDragon will want to tote Emulation and Homebrew when trying to sell his system. If a closed-source emulator or homebrew software is more popular, it is his best interest to shove it to the front so that people who are interested in said software see that they can use the Pandora 2 to use this software and hopefully generate more buzz and sales. If this may get him in trouble with FSF, then is it really worth it to try to get their certification in the first place? Isn't the OpenPandora repo hosted on the official server? This would make you have to drop things like DraStic from the repo. Honestly, I don't see this as a benefit for users.

An like you said, there are enemies of FSF out there. I feel that trying to get that certification for the Pandora 2 will be just like trying to pick a side in a Vim/Emacs war.

-God Ginrai

 

[_wb_]

If we want to get the certification, then of course we need to be clear (with the FSF) about what will be allowed and what not in order to keep the certification. In particular, we need to know where the boundary is when it comes down to "recommending for use in conjunction with the product". I think we agree that ED should be able to make videos etc advertising the unit running non-free software (or free software which happens to be emulating non-free roms). So if certification does not allow that, then we shouldn't go for it.

As for the repo: I don't think it is a problem that the repo contains non-free software, even if it is hosted on ED's server.

I don't think there are enemies of the FSF here - only people who are somewhat annoyed by some of their advocates. If getting FSF certification feels like picking a side, then of course there's nothing to stop us from getting some kind of certification from other sides, if those exist. We can pick all sides that are in line with our goals.

 

[ZXDunny]

I'd agree with _wb_ as usual. If we can get FSF certification without a significant restriction on our freedom, then it can't hurt.

There will be some restrictions as to what we can advertise, what we can ship etc - but so long as we can point people to where they can get the non-free stuff, it's all good.

D.