Let’s Limit the Effect of Software Patents, Since We Can’t Eliminate Them (Wired)

Posted Nov 6, 2012 18:06 UTC (Tue) by dlang (guest, #313)
In reply to: Let’s Limit the Effect of Software Patents, Since We Can’t Eliminate Them (Wired) by bojan
Parent article: Let’s Limit the Effect of Software Patents, Since We Can’t Eliminate Them (Wired)

As others have noted, trying to limit it to 'open source programs' is tricky, we have arguments today if some software is open or not.

I think it's better to leave it at all software, but this still ends up with a tricky definition. What is "generally used computing hardware"?

Here's where I think the attention needs to be paid.

I would define it as hardware where the command set and interfaces to the hardware are documented and available to the general public.

This would mean that the OS running on a general PC may be covered, but a driver accessing proprietary hardware (for example, the NVIDIA driver) would not be covered, but the ATI driver would be (since ATI publishes the interface for their cards)

Let’s Limit the Effect of Software Patents, Since We Can’t Eliminate Them (Wired)

Posted Nov 6, 2012 22:32 UTC (Tue) by Zack (guest, #37335) [Link]

>>I think it's better to leave it at all software, but this still ends up with a tricky definition. What is "generally used computing hardware"?

It would be a tricky definition to get right, and there probably would always remain a gray area. But it shifts the software patent problem far away from where software patents are doing most harm.

If this limit on the effect of software patents ever gets accepted, and (at least) a single "generally used computing hardware"-platform is established, a programmer can just sit down and write software on this platform and distribute it for users of this platform.

It pushes the bickering to the edges of "generally used" and specialised hardware, reducing the problem to spats between dedicated hardware manufacturers and leaving the average programmer free to implement improved mailclients, one-click-shop webpages, and other everyday software to their hearts content.

Let’s Limit the Effect of Software Patents, Since We Can’t Eliminate Them (Wired)

Posted Nov 7, 2012 5:36 UTC (Wed) by Seegras (guest, #20463) [Link] (3 responses)

You're missing the point. Software already is not patentable. Because its math. The Problem is that the patent office and the courts have a different definition of math than mathematicians and computer scientists.

Kind of legislating Pi=3.

(and no, conversely, just because someting can be expressed mathematically does not mean its math. But that doesn't concern software).

PolR tries to make the case.
http://web.archive.org/web/20200730120644/http://www.groklaw.net/article.php?story=20121013192858600

"software is math!" and how we win

Posted Nov 7, 2012 12:10 UTC (Wed) by coriordan (guest, #7544) [Link] (2 responses)

That leaves three options:

1. Replace the judges and examiners with computer scientists
2. Convince the judges and examiners to switch to the computer scientists' definition
3. Change the law so that the (broken) opinion of the judges and examiners doesn't matter

#1: Isn't realistic. (That is to say, pursuing this goal isn't an efficient use of our limited resources.)

#2: This is where "software is math" fits in, but we've been making this argument for years when submitting amicus briefs and responding to patent office consultations. Still might work, and we'll keep trying, but the Supreme Court is our best hope and they've been pretty reluctant to take a stand. And if we win in the Supreme Court but haven't built any support in Congress, the megacorps will rush through some legislation.

#3: This is where RMS's suggestion fits in. We've done a lot of work on this in other countries but very little in the US. People say it's impossible in the US, but the hard truth is that it's the only way to really win.

"software is math!" and how we win

Posted Nov 7, 2012 20:02 UTC (Wed) by dlang (guest, #313) [Link]

#1 has many problems.

the limited number of computer scientists is the first one (would you rather have them doing useful stuff or reading patent apps)

But the biggest problem is that setting any industry to police itself is a major problem.

"software is math!" and how we win

Posted Nov 11, 2012 22:27 UTC (Sun) by SecretEuroPatentAgentMan (guest, #66656) [Link]

> 1. Replace the judges and examiners with computer scientists

Are you aware that examiners need a technical degree and a background in their arts? Or that some judges are former patent attorneys with technical insight?

> 2. Convince the judges and examiners to switch to the computer scientists' definition

Exactly what is this definition?

> 3. Change the law so that the (broken) opinion of the judges and examiners doesn't matter

So who gets to overrule the judges?

Definition of "generally used computing hardware"

Posted Nov 7, 2012 11:56 UTC (Wed) by coriordan (guest, #7544) [Link] (9 responses)

> What is "generally used computing hardware"?
>
> Here's where I think the attention needs to be paid.

I'm thinking the same.

But I'd go for a broad definition. The PC might be extinct in 10 years. 95% of the population might be doing all their computing on watches, phones, glasses, and other wearables.

Maybe courts would interpret "generally used" as referring to whatever hardware people generally use to run software, so it would include the things in my previous paragraph. But it's better to not leave this down to chance. We should find a wording that makes this meaning clearer, IMO.

Definition of "generally used computing hardware"

Posted Nov 7, 2012 21:02 UTC (Wed) by dlang (guest, #313) [Link] (8 responses)

If you just define 'generally used computing hardware" to be thigns that people run software on, that will include toasters and other similarly bizarre items.

If you define it in terms of ABI documentation and ability to load their own software on them, you then put manufacturers into a corner

they can either

1. document the hardware and make it so that others can run software on it, making their software running on the device immune to patent lawsuits

2. keep their hardware secret and locked down and be liable for any patent lawsuits against their software.

In other words, defining things this way not only eliminates patent claims against all opensource software, it also gives manufacturers a strong incentive to document their hardware sufficiently that opensource software can be made to run on it.

For things like video cards, you would have the following situation

The NVIDIA drivers would be vulnerable to patents.

The Raspberry Pi GPU firmware could be vulnerable to patents (I don't know if they have the info needed to write stuff to run on the GPU available or not, the source of the firmware blob is definitely not available), but the Pi drivers would not be as the interface to the firmware is published.

The ATI drivers would not be vulnerable to patents as the ABI for the cards is published

Apple devices would be vulnerable to patents as they are locked down with undocumented hardware and the owner is prevented from installing new software on them.

Google Nexus devices would not be vulnerable as they are documented.

Other Android devices would only be vulnerable in their drivers, and even then, only to the extent that they have 'secret' ABIs

book readers would be vulnerable as they are locked down

Game consoles would be vulnerable as they are all locked down

All of the above sounds very reasonable to me :-)

what about running VLC on an iThing?

Posted Nov 8, 2012 11:21 UTC (Thu) by coriordan (guest, #7544) [Link] (7 responses)

You could be right. I'm just still thinking it through...

What about the stuff in application space? Patent problems in driver space are a small minority (although surely growing due to video and other acceleration in hardware).

If someone gets a completely closed iThing and runs VLC, should that person be safe or not?

I'd say yes, and that the iThing is "generally used computing hardware". But if the hardware has to be documented, then the person running VLC on an iThing wouldn't be safe.

I think people should be able to run VLC everywhere without patent risk, and that whatever falls outside the "generally used computing hardware" definition should only be liable insofar as the act is specific to that non-general hardware.

what about running VLC on an iThing?

Posted Nov 8, 2012 19:17 UTC (Thu) by dlang (guest, #313) [Link] (6 responses)

good point

I guess we would need to make it something along the lines of if the software only uses published interfaces (be they software or hardware), it's safe, but if it uses 'secret' interfaces, it's not.

so something written to run on a iThing would be safe, but the OS for the iThing would not be.

And Rockbox on the iPod nano 2?

Posted Nov 9, 2012 15:40 UTC (Fri) by coriordan (guest, #7544) [Link] (5 responses)

What about the iPod nano 2 interfaces? They were secret (encrypted even, IIRC), but after a few years the RockBox guys figured them out and there are now free software systems for those devices. Does Rockbox's work make the iPod's interfaces public?

(If not then Rockbox would have patent risk.)

I also wonder if a library cover secret interfaces with open ones, thus creating a loophole in the law. (This is only possible for certain definitions of "secret".)

And Rockbox on the iPod nano 2?

Posted Nov 9, 2012 19:59 UTC (Fri) by dlang (guest, #313) [Link] (4 responses)

> Does Rockbox's work make the iPod's interfaces public?

No, If it did it would make everything on the iPod safe.

I would say that Rockbox documenting what they _think_ the interfaces are should protect Rockbox, but it would not protect Apple software (after all, there is no way of knowing is Rockbox got things right, or just 'right enough to work')

As for a library to hide secret interfaces. As long as the open ones are documented and can be used by others, I don't have a big problem with this 'loophole'. Either the library makes enough available for people to use it to write other software (in which case it is functionally equivalent to the 'secret' interface), or it doesn't, (in which case it can't be used for things that want to take advantage of the 'secret' commands)

I'll note that I am one of the people who doesn't consider a system non-free just because it has loadable firmware blobs.

Probably a bad gamble

Posted Nov 9, 2012 21:30 UTC (Fri) by coriordan (guest, #7544) [Link] (3 responses)

...but then the Rockbox developers and users (me) have patent risk.

The API approach is a gamble, but I think it's probably a bad gamble.

With the transition from PC/Laptop to carryables/wearables, I think we're heading into a decade where a lot of free software users will be putting free software on their devices against the manufacturer's wishes.

Offering patent protection in return for documentation probably isn't enough. If software patents were that big an issue for the hardware manufacturers, they would be working on their own campaign to get that protection without having to document their interfaces. The current relative silence on the issue suggests it's not that motivating, so the gamble is risky.

(That said, the general v. special distinction also has problems. What's "special" about an mp3 player? It just does a subset of what my PC does. Is "special" supposed to refer to medical devices? How do we make that clear in a legal text?)

Probably a bad gamble

Posted Nov 9, 2012 22:48 UTC (Fri) by dlang (guest, #313) [Link] (2 responses)

given that the librarian of Congress just denied the right to jailbreak tablets because it would have included the ability to jailbreak e-readers as well, I don't think that you are going to have any luck trying to get patent protection for running software on such devices.

That being said, in the case of Rockbox, all the software that actually had a real potential for being hit by patents is going to be the same on all devices. It's only the bootloader and drivers that will be special for the iPod.

Probably a bad gamble

Posted Nov 10, 2012 14:53 UTC (Sat) by coriordan (guest, #7544) [Link] (1 responses)

If Rockbox is to be ok (except bootloader and drivers), the proposal should say "a program *that can be effectively run* on generally used computing hardware".

But I'm not sure that that's broad enough. It could leave a problem for video acceleration in drivers. I'd stick with looking for ways to ensure that mp3 players are wholly included in "generally used computing hardware".

The terrible decision about tablets does prove that legislators can make anti-citizen decisions, but the goal (enforce companies' DRM against computer users) shouldn't bind them to making bad decisions on patent questions.

Probably a bad gamble

Posted Nov 10, 2012 23:05 UTC (Sat) by dlang (guest, #313) [Link]

> the proposal should say "a program *that can be effectively run* on generally used computing hardware".

why do you need to include the word 'effectively' here? that is subjective.

> It could leave a problem for video acceleration in drivers.

I'm actually Ok with that being a problem.

A closed source driver with closed ABIs like NVIDIA should be a problem

A closed source driver with open ABIs like ATI is Ok with me. I will try to avoid using it and instead use the open source driver that will be written instead.

A open 'shim' driver with open ABIs like the Raspberry Pi is also Ok with me. Everyone uses the same ABI and any 'secret stuff' is in the firmware.

being able to replace the firmware and have the ABI from the firmware to the hardware be open is even better.

> I'd stick with looking for ways to ensure that mp3 players are wholly included in "generally used computing hardware".

I wish you good luck with that, however in the current political climate, I think you will run into the same problem that they had with tablets including e-book readers

Let’s Limit the Effect of Software Patents, Since We Can’t Eliminate Them (Wired)

Posted Nov 15, 2012 19:51 UTC (Thu) by JanC_ (guest, #34940) [Link] (1 responses)

I would define it as hardware where the command set and interfaces to the hardware are documented and available to the general public.

This would mean that the OS running on a general PC may be covered, but a driver accessing proprietary hardware (for example, the NVIDIA driver) would not be covered, but the ATI driver would be (since ATI publishes the interface for their cards)

That would also mean that it would be in the interest of hardware companies to reverse engineer & publish the specs of their competitor's hardware (or pay a 3rd party to do so), as that would grant them free access to the same technology, and in turn that would give open source developers proper specs to almost all hardware...? :)

Let’s Limit the Effect of Software Patents, Since We Can’t Eliminate Them (Wired)

Posted Nov 15, 2012 20:58 UTC (Thu) by dlang (guest, #313) [Link]

reverse engineering cannot produce the full specs for a system.

It can produce specs that are 'good enough' in practice, but even theoretically, it can't tell the difference between 'undefined result that happens to be X' and 'deliberalte calculation that results in X'

So the documentation I am referring to would need to be produced by the company producing the hardware (or by a 3rd party on their behalf)