HDMI Forum Rejects Open-Source HDMI 2.1 Driver Support Sought By AMD

[F.Ultra]

The wikipedia is not always right. There is a key part of patent law Wikipedia nicely skips over.

https://www.brettoniplaw.com/requirements%20for%20a%20patent.html

35 U.S.C. §103 - Non-obviousness


Yes doing a clean room implementation is doing a POSITA test with Patents to Invalidate the patent. This is hard to setup at times. to make sure the POSITA people in the clean room at not tainted.



Asetek gets away with it because of this bit in 35 U.S.C. §103 - Non-obviousness

Anyone who has an Asetek design AIO fail knows this one way too well. Even so Asetek had to buy out 2 companies because their patents would not work because they had correct clean room implementations that could have nuked their patent. One had the pump on the CPU block 90% off to the Asetek. Asetek worked these 2 companies they were going to lose in court so completely acquired the company. Asetek has never solved the fact that over time the liquid inside a sealed AIO disappears. There are in fact older non sealed AIO designs that have the pump on the CPU as well. This is the problem the non sealed is the obvious design for quality AIO that lasts so that the user can top it up as the liquid disappears .

Yes patent law for you. You patent something that is broken garbage by design patent law will allow you to keep on enforcing your patent to punish those who use the garbage design.

Patent law its about obviousness. Not just simple obviousness either. It has to be obviousness that it a good design that a person would have came up with.

Clean room with copyright is not going to cause the other party to consider buying your company outright to protect their copyright. Clean room with Patent if they are a larger company they see you doing clean room on their patents that could prove a key patent they are making money from as obvious buying your company out to stop you comes really possible.

Clean room usage with patent is pure offensive. This is not to make a competing implementation this is purely to destroy the patent by proving it obvious good design. Yes once you destroy the patent you can just clone their product. This is how some china clone product makers get away with it legally because they destroy the patent that could be used to block them shipping around the world.

The whole thing with POSITA is that it is an enormous gray area in patent law. This step is supposed to be done when the patent is applied for and yet we see software patents being issued left and right (my position is quite simple, any and all sw patents to date does not fulfil the POSITA requirement but yet here we are). And here comes what I mentioned in the previous comment about genius, because if you are a genius then you are not a POSITA. So in a trial when you have one expert saying that the invention is innovative and one says that it isn't then it's very easy for the court to simply define the other expert as being a genius and not a POSITA.

So a second implementation does not solve this problem, they can always claim that it was created by a genius and not by a POSITA. Remember, to the average Jury, we developers are all autistic geniuses.

And the Asetek example does still not demonstrate clean room, those 2 other companies simply had solved the issue in a different way that didn't infringe on the claims of the Asetek patent. This is similar to how Arctic have solved it in their Liquid Freezer series where they have the impeller mounted on the side specifically to avoid the wording of the patent. This is again not clean room, this is designing with the patent in mind.

[Panix]

Sounds like an opportunity for Michael to do a Digital Foundry style investigation of the matter

The 124 firmware for Chinese adapters doesn't enable VRR on Windows for some reason, so I doubt Cable Matters will update their description. VRR is also still quite fickle even without the adapter, the newest 6.7.6-201 kernel on Nobara seems to have a regression that causes VRR to malfunction in the same way it does when HDR is enabled even in SDR (refresh rate jumping).

If you use HDMI - & Nvidia gpu, do you have the same problem?

I understand the open source zealots' position - I support it, too, myself - but, it has to be conceded that there's just a lot of hassle here - using active adapters - spending extra money and then requirement of a firmware upgrade - and then, look - there's still not a consensus that it works properly or fully (as in full features) - so, that is a valid concern, imho.

These closed source companies just have so much power and influence - to me, that's very discouraging and although I don't like to sound defeatist - it seems like a lot of hassle to support this FOSS principle - I hope that these 'fixes' can implement all these features soon/eventually - the HDMI Forum and the other collaborators should lose....somehow.... just don't know how.

[hartree]

If you use HDMI - & Nvidia gpu, do you have the same problem?

I understand the open source zealots' position - I support it, too, myself - but, it has to be conceded that there's just a lot of hassle here - using active adapters - spending extra money and then requirement of a firmware upgrade - and then, look - there's still not a consensus that it works properly or fully (as in full features) - so, that is a valid concern, imho.

These closed source companies just have so much power and influence - to me, that's very discouraging and although I don't like to sound defeatist - it seems like a lot of hassle to support this FOSS principle - I hope that these 'fixes' can implement all these features soon/eventually - the HDMI Forum and the other collaborators should lose....somehow.... just don't know how.

Not sure, I don't have an Ampere or Lovelace card to test it, maybe I'll pick a cheap one up to test it because I'm interested in NVK.
Nvidia has disadvantages too, afaik I'd need to route the display output through an AMD or Intel gpu anyway to get gamescope-session support which is the killer feature for linux gaming in my opinion, a true console-like interface with all the bells and whistles.

Overall I'm happy with AMD's efforts and never thought this was too grave of an issue, for gaming even the PS5 only supports chroma subsampling in 4k120 due to its 32GBit/s HDMI port, and I have never seen as many complaints for it as I've seen from concern trolls in the Linux space, and it's not fixable with a $15 adapter there as it is on linux now.

[oiaohm]

The whole thing with POSITA is that it is an enormous gray area in patent law. This step is supposed to be done when the patent is applied for and yet we see software patents being issued left and right (my position is quite simple, any and all sw patents to date does not fulfil the POSITA requirement but yet here we are). And here comes what I mentioned in the previous comment about genius, because if you are a genius then you are not a POSITA. So in a trial when you have one expert saying that the invention is innovative and one says that it isn't then it's very easy for the court to simply define the other expert as being a genius and not a POSITA.

So a second implementation does not solve this problem, they can always claim that it was created by a genius and not by a POSITA. Remember, to the average Jury, we developers are all autistic geniuses.

There is problem legally you stated as counter argument does not hold up when a clean room has been done correctly. There is such things as intelligence tests that stand up under law. Its really bad to make this argument that the other side has to be geniuses to design this thing and the other party like sorry no the person who did this work is a bellow average idiot. This explains the clean room usage. You can stock the clean up with not the best developers/people yet still put expert on stand who is a genius to handle the cross exam well.

Clean room with patents is about proven POSTA. This include making sure you staff the clean room with average or slightly bellow average personal.

Lot of software patents are because they are in fact bad ideas not good ones. POSTA cannot be used against a patent taken out that is truly provable bad idea.

And the Asetek example does still not demonstrate clean room, those 2 other companies simply had solved the issue in a different way that didn't infringe on the claims of the Asetek patent.

This is wrong you are looking for two companies Austek by a shell company did hostile take overs on. You will find what you wrote there is wrong.

Something people never think of is that is simple to patent a bad idea because bad idea is not classed as obvious under patent law. So item claiming patent protection might have patent protected design because is a bad design.

[smitty3268]

Except that I don't believe that applies at all in this scenario. This is clearly a case of somebody made something

I have zero clue what patents are involved here (or even if they are). I just think it's a bad idea to assume things about this situation just because you'd like them to be true, if you don't actually know for sure.

What you are saying definitely does not apply in many situations. Does it here? Maybe. It's certainly possible. I just don't know.

[nfp0]

I'm sorry if this was discussed before, but this is becoming a big topic.

I understand an optional closed-source binary blob is not the ideal solution, but what's stopping AMD from providing one to at least get them out of the false-advertisement territory, which is illegal in most places?
An AMD dev even said on the Gitlab issue that the basic functionality was already implemented. They just need to release it as an optional closed-source component.

We have the basic functionality up and running, now we have to go through each of the features with legal and determine if/how we can expose them while still meeting our obligations.

[shtripok]

Interestingly, in the world of single board computers (SBCs) and cheap set-top boxes (usually Android), DisplayPort ports do not exist at all.
HDMI only.

[F.Ultra]

There is problem legally you stated as counter argument does not hold up when a clean room has been done correctly. There is such things as intelligence tests that stand up under law. Its really bad to make this argument that the other side has to be geniuses to design this thing and the other party like sorry no the person who did this work is a bellow average idiot. This explains the clean room usage. You can stock the clean up with not the best developers/people yet still put expert on stand who is a genius to handle the cross exam well.

Clean room with patents is about proven POSTA. This include making sure you staff the clean room with average or slightly bellow average personal.

Lot of software patents are because they are in fact bad ideas not good ones. POSTA cannot be used against a patent taken out that is truly provable bad idea.



This is wrong you are looking for two companies Austek by a shell company did hostile take overs on. You will find what you wrote there is wrong.

Something people never think of is that is simple to patent a bad idea because bad idea is not classed as obvious under patent law. So item claiming patent protection might have patent protected design because is a bad design.

you have to be more specific, Google returns 0 results for "asetek hostile takeover".

Also show me a single software patent being invalidated due to a clean room, not that I doubt that such a unicorn exists, but because I would like to look at the details.

Interestingly, in the world of single board computers (SBCs) and cheap set-top boxes (usually Android), DisplayPort ports do not exist at all.
HDMI only.

Well set-top boxes makes sense since they are designed to be connected to TV:s and not monitors.

[oiaohm]

you have to be more specific, Google returns 0 results for "asetek hostile takeover"..

I said it was shell companies because this one is not easy to find.

Also show me a single software patent being invalidated due to a clean room, not that I doubt that such a unicorn exists, but because I would like to look at the details.

The most recent example I can think of would be the Oracle vs Google 2012 ruling over the patent side of the Java/android dispute. Oracle tried to argue that Clean Room was not enough to invalid patents The court ruled in Google favor so invalidating 2 patent Oracle still had in the claim with because of clean room. This is why it came a pure copyright case after that.

So F.Ultra such a unicorn do exist to say clean room should not be taken lightly even in a patent case. Clean room effect in patent case if it staffed the right way is patent termination.

Of course it lots money spent to get court to get ruling that your clean room work has nuked the patent. Yes the 2012 ruling for google over the patents google did not get costs or damages when winning. This is why you don't see many of these cases.

Clean room is a legal offensive measure if you find yourself under attack for patent infringement with the goal of destroying the patent. Remember you destroy a patent you are most likely going to be stuck with costs with no rights to get damages because the destruction of the patent is your payment.

F.Ultra think about it you prove the patent does not apply to your product you have chance of getting costs and damages. The reality if you have clean room work you will try every other legal method to avoid the patent before you use it as a company. The cost of no damages and no costs because you used clean room to invalidate the patent. You have the same problem if you prove the patent was invalid due to prior invention. This is where the patent system really bad. The patent system does not encourage patent invalidation in fact it punishes any party attempting this.

F.Ultra all methods of patent invalidation in court are rare unicorns to happen due to the this costs and damages thing. This also leads to mistakes of thinking items like clean room don't apply in patent cases when they do. Clean room and prior invention usage in patent case is a party has tried everything else and is now going for invalidate patent. Also you required very deep pockets to pay legal bills to get this far and be willing to take the loss to win.

[F.Ultra]

I said it was shell companies because this one is not easy to find.



The most recent example I can think of would be the Oracle vs Google 2012 ruling over the patent side of the Java/android dispute. Oracle tried to argue that Clean Room was not enough to invalid patents The court ruled in Google favor so invalidating 2 patent Oracle still had in the claim with because of clean room. This is why it came a pure copyright case after that.

So F.Ultra such a unicorn do exist to say clean room should not be taken lightly even in a patent case. Clean room effect in patent case if it staffed the right way is patent termination.

Of course it lots money spent to get court to get ruling that your clean room work has nuked the patent. Yes the 2012 ruling for google over the patents google did not get costs or damages when winning. This is why you don't see many of these cases.

Clean room is a legal offensive measure if you find yourself under attack for patent infringement with the goal of destroying the patent. Remember you destroy a patent you are most likely going to be stuck with costs with no rights to get damages because the destruction of the patent is your payment.

F.Ultra think about it you prove the patent does not apply to your product you have chance of getting costs and damages. The reality if you have clean room work you will try every other legal method to avoid the patent before you use it as a company. The cost of no damages and no costs because you used clean room to invalidate the patent. You have the same problem if you prove the patent was invalid due to prior invention. This is where the patent system really bad. The patent system does not encourage patent invalidation in fact it punishes any party attempting this.

F.Ultra all methods of patent invalidation in court are rare unicorns to happen due to the this costs and damages thing. This also leads to mistakes of thinking items like clean room don't apply in patent cases when they do. Clean room and prior invention usage in patent case is a party has tried everything else and is now going for invalidate patent. Also you required very deep pockets to pay legal bills to get this far and be willing to take the loss to win.

The Oracle vs Google 2012 case was not about clean room for the patents. For both patents, Google successfully argued that they didn't use the methods from Oracle:s patent claims, the '104 patent required a symbol reference but Googles implementation did not contain one and the '520 patent used "simulated execution" while Googles implementation used parsing instead. So again a case of deliberate patent workarounds and not clean room.

So this one doesn't count, so so far no unicorn.