Posted by speckx 6 hours ago
Some questions that come to mind are:
Do you actually own the hardware?
Do you stop owning the hardware, meaning you're not allowed or able to use it once the company that's sold it to you stops providing or updating hardware?
Does every piece of hardware come with a guarantee that it will work on every piece of hardware that it's attached to forever?
These are just a few of the really troubling questions when people forget that software has to run on hardware. And the closer the software is to the core operation of a piece of hardware and equipment, the more topics like right to repair or right to maintain or right to optimize come up.
I never thought that a video card replacement driver would make me think of companies that are limiting Farmers from repairing and updating their own tractors.
The closest you can get is something like the Raptor Talos II, although parts of the hardware are still under patent.
Unbelievable mafia tactics from those clowns.
https://static.lwn.net/images/pdf/Cease_and_Desist_Notice_Mi...
Fuck you, Synaptics.
https://www.eff.org/deeplinks/2019/06/felony-contempt-busine...
Which is made funnier by the lawyer's name being Bateman. Sorry, could not resist that one.
What I find more ridiculous is the timing:
> You are to provide Synaptics, no later than 11:59 p.m. Pacific Time on October 11, 2026
The letter is dated "October 5, 2026". WTF are these lawyers smoking that they think 5 days is any reasonable time frame.
They just try to keep the pressure up so that the other entity gives up before they have time to estimate whether any of the claims are valid. One person against big corporation, which is threatening with big penalties. A common tactic.
OCuLink doesn't even use drivers. It's literally just a PCIE passthrough. All usual tools apply, whereas DisplayLink encrypts all data, so a big fuckyou for getting logs.
Quit giving trash companies pounds of money for abuse at scale.
EDIT: Fixed now.
In any case, I am not a subscriber and I see it.
> The following subscription-only content has been made available to you by an LWN subscriber. Thousands of subscribers depend on LWN for the best news from the Linux and free software communities. If you enjoy this article, please consider subscribing to LWN. Thank you for visiting LWN.net!
> Are paywalls ok?
> It's ok to post stories from sites with paywalls that have workarounds.
> In comments, it's ok to ask how to read an article and to help other users do so. But please don't post complaints about paywalls. Those are off topic. More here.
https://lkml.rescloud.iu.edu/hypermail/linux/kernel/2610.1/1...
According to patent prosecution records, the independent claims for GB2484736 originally lacked any network transport limitations and were deemed entirely unpatentable against the prior art listed above.
To get the patent granted, DisplayLink’s attorneys amended the claims to specifically add delivery of the display data over a "general-purpose data network." The validity challenge relies on proving that adding a general-purpose transport layer (like USB or Ethernet) to existing image-compression methodologies was an obvious step and completely unoriginal in 2010, given the widespread existence of USB monitors and network-attached graphics at the time.
As far as I can tell, these patents only go up to claim 10 in the first place:
U not cleared to see it
So only the combination may be new. But this reminds me of a USPTO thing from my past.
The landmark ruling you are referring to is the U.S. Supreme Court case KSR International Co. v. Teleflex Inc. (2007), which led directly to the USPTO publishing its highly influential 2010 Examination Guidelines for Determining Obviousness. Under this framework, if element A is known and element B is known, combining them is not patentable if the combination yields nothing more than predictable results.
https://www.uspto.gov/web/offices/pac/mpep/s2143.html
So, general-purpose data network, such as USB or Ethernet was known (A), claims Haar-transform decoding was known (B) [DE69909880T2 / EP0955609]. Even if we don't find this combination in prior art, it is dubious that this is patent worthy based on USPTO guidelines (combine A with B).
But I see a lot of stuff that suggests prior art:
1. Noriaki Yamazaki and Tetsuo Tanaka (2007). “ZYWRLE: Enhanced VNC Codec for Motion Picture.” IEICE Transactions, J90-D(7), pp. 1790–1798. Brief summary: Adds wavelet compression to VNC’s ZRLE codec for transmitting moving images, reducing network traffic to 20–50% of standard ZRLE. https://search.ieice.org/bin/summary.php?abst=j&category=D&i...
2. Joshua G. Senecal, Peter Lindstrom, Mark A. Duchaineau and Kenneth I. Joy (2004). “An Improved N-Bit to N-Bit Reversible Haar-Like Transform.” Proceedings of Pacific Graphics 2004, pp. 371–380. DOI: 10.1109/PCCGA.2004.1348368. Brief summary: Introduces PLHaar, a reversible Haar-like transform supporting lossy and lossless image compression without increasing coefficient bit width, suitable for bandwidth-limited video and graphics hardware. https://www.osti.gov/servlets/purl/15014239
3. Joshua G. Senecal, Mark A. Duchaineau and Kenneth I. Joy (2004). “Reversible n-Bit to n-Bit Integer Haar-Like Transforms.” Proceedings of the Data Compression Conference 2004, p. 564. Brief summary: Introduces lookup-table Haar-like transforms that preserve bit width and reversibility, with applications to fast image compression and streaming hardware such as DVI channels. https://www.osti.gov/servlets/purl/15013668
Good luck synaptic!
Also possibly related:
Method and apparatus for increasing memory resource utilization in an information stream decoder, https://patents.google.com/patent/WO1999057908A1/en
And a paper by a PHD student in 2005, Real-time Rendering with Wavelet-Compressed Multi-Dimensional Datasets on the GPU, https://cs.ucsb.edu/sites/default/files/documents/2005-05.pd...
Displaylink's own abandoned patent application talking about sending patches of compressed/uncompressed graphical data to the screen over USB: https://patents.google.com/patent/US20060119602A1/en
Using reference frames with Haar wavelet transforms in video compression, https://patents.google.com/patent/US6272180B1/en
It would take much human time to search for prior art across all of human knowledge for existing patents and new patents being proposed. But GPUs are tireless, and can continually grind towards success. They build their case for dismantling all existing active patents as long as they are kept fed.
If I were a frontier lab, I would be finding the most valuable patents, starting the research agents up, and then shorting companies that would see enterprise value decline if subject patents were invalidated (think Hindenburg or Culper Research activist investors). Just one way of monetizing the fancy search engine and token dispenser.
After how many dollars spent? I mean, it is good that eventually bad patents get killed, but it still costs too much to do it, one at a time. We need a more streamlined process, maybe? Maybe an academic-review-like process at filing? Send the claims anonymously to a few experts in the field, randomly selected. If they all say it is obvious, kill it immediately and fine the company for wasting the patent office's time?
I think we could get the 7 grand together here.
My only concern is, what if some well-meaning person puts in a poorly constructed claim, and then that creates a counterproductive precedent.
I think ideally, someone like Software Freedom Conservancy or the Linux Foundation or the FSF or whoever would take charge of doing it, taking advice from patent attorneys in the process.
Maybe there is some UK patent attorney reading this who is willing to take it on pro bono?
The patent system itself has been a moat and we will see more and more reasonable submissions that will force a rethink.