5 ms·

The patent's in the trial are pretty bad. Half are just about devices synchronize state using a LAN. The other half concern the UI and stuff. Maybe these had a

The patent's in the trial are pretty bad. Half are just about devices synchronize state using a LAN. The other half concern the UI and stuff. Maybe these had a bit of innovation in 2003 but there is nothing substantial. I'm sure some pre-network audio equipment had to synchronize state and audio timing so doing it over the LAN is not (IMO) sufficient for a new patent. In any case, software patents should be very strict and last at most 5 years (which is more like 8 years from application to expiry).

tomaszs 5y ago
Especially if we consider over 50 000 patents Google holds. They even tried to patent ANS compression that was made free for everyone by Jarosław Duda.
newdude116 5y ago
Well, the problem is, like with proof in math, afterwards it is always obvious, if not inevitable.
- knorker 5y ago
  If parent post is correct, though, the "innovation" being protected is basically "we used to send signal over wire. Now that both endpoints have LAN, I'll send that signal over the LAN". That, like many other software patents, is not "obvious is retrospect". It's not even "obvious idea at the time", it's less than that. It's just inertia. It's a decisionless decision. It's patenting "not stepping in a puddle".
  - tremon 5y ago
    But aren't patents supposed to protect implementations of ideas, not the ideas themselves? So then "keeping state between autonomous devices connected via LAN" seems like a worthwhile implementation, regardless of whether all the individual components were already existing in other technology domains. Besides, I don't think that consensus-based asynchronous state protocols like Paxos were widely available in 2003 (and even so, implementing Paxos on a resource-constrained device like e.g. an Arduino might still qualify for a patent in 2021).
    - knorker 5y ago
      > But aren't patents supposed to protect implementations of ideas, not the ideas themselves? This is the problem with software patents. The border between ideas and implementation is not as clear. Strictly the answer to your question is no. Copyright protects the implementation, and patents protect the ideas. But where's the difference between an idea, and a design? For software, what's the difference between a design and the implementation? Sure, you may actually have a separate design (and not "the implementation is the design"), but in software the step from "design" to "implementation" isn't "manufacturing". Sure, even for physical goods, and their patents, the patent is often not sufficient to create a working system. For those holding software patents, they have to also list all ideas they have around that. Because there are patent trolls who check for new patents, and then file patents around that, creating a mine field so tight that you cannot even use your own invention. Analogy: you patent a laser pointer, and immediately someone patents all possible use cases for a laser pointer they can think of (point at things during presentation, aim weapons, exercise cats[1], project things onto a wall, stick the laser pointer in a handheld device, have a button on the laser pointer, non-chargable batteries in the laser pointer, chargable batteries, mains powered, capacitor powered, etc… etc… Anyway, a comment field is not sufficient to explore just why software don't fit well under the model of patents. [1] this is a real patent, by the way.
      - DoingIsLearning 5y ago
        > For those holding software patents, they have to also list all ideas they have around that. > Because there are patent trolls who check for new patents, and then file patents around that, creating a mine field so tight that you cannot even use your own invention. This is something I never understood. When someone submits those type of patents, the patent officer most certainly knows the 'prior art' original patent exists. So why is this not patent 'obviousness'?
        thfuran 5y ago
        Because the patent office now pretty much just checks if something is the same as prior art and let's the issue of obviousness or usefulness be settled in litigation.
szszrk 5y ago
Have you had a closer look at that synchronize state stuff? I can't see much in the article. > devices synchronize state using a LAN If that has anything to do with music latency, then it's not "nothing substantial". It's their core business and their flagship feature. What they do well, that others could not, is playing multiple speakers perfectly in sync. Either stereo, or multiple rooms, almost as many speakers as you want, connected via wired network, or wireless, or by their on mesh system. And each speaker is basically a separate computer. It's not connected by analogue cable. Anyone who were looking for Sonos alternatives with that functionality knows it's very hard to find and suddenly their prices become somewhat justified.
- cosmotic 5y ago
  Slim Devices' squeezebox had this feature before Sonos. I implemented it in high school using visual basic. The most non-trivial part is that if you try to synchronize an MP3 or other encoded stream instead of the decoded PCM stream, you end up seeking to mid-frame points which causes invisible adjustments to the seek point. In the end, it's as simple as 'seek to this point' and then check if the point is off by less than some threshold, don't do anything. I did a few other fancy things like not seeking until the 'heart beat' lined up a few times in a row (duration in heartbeat was same as difference in time since last heartbeat).
- KingMachiavelli 5y ago
  I admit I only briefly looked at each patent but even looking a second time, I can not find anything more specific basically the devices share clock information and calculate the difference between their clocks. How is this different than syncronizing transactions between database servers or how GPS works? I found a GPL license multi-room/device audio player that discussed this very lawsuit last year [1]. The developer raises the same point than anything like this will involve something resembling clock syncronization. Looking through the lawsuit itself, about 90% of it is just comparing the end result of Sonos' & Google's products in that both let you add/group speakers, adjust volume and playback of the group, etc. I cannot find any claim that Google's implementation is a copy of Sonos'. It seems to try to sugest that is the case by bringing up that prior to Google's devices, Google had worked with Sonos and had access to details of their implementation. [1] https://github.com/badaix/snapcast/issues/642 https://github.com/badaix/snapcast/issues/642
IshKebab 5y ago
One patent is just on how to set up a playback device by connecting to it directly and sending wifi details. It's lucky they wrote "playback device" and not "device" otherwise the entire IoT industry would be violating it.

tomaszs 5y ago

Especially if we consider over 50 000 patents Google holds. They even tried to patent ANS compression that was made free for everyone by Jarosław Duda.

newdude116 5y ago

Well, the problem is, like with proof in math, afterwards it is always obvious, if not inevitable.

knorker 5y ago
If parent post is correct, though, the "innovation" being protected is basically "we used to send signal over wire. Now that both endpoints have LAN, I'll send that signal over the LAN". That, like many other software patents, is not "obvious is retrospect". It's not even "obvious idea at the time", it's less than that. It's just inertia. It's a decisionless decision. It's patenting "not stepping in a puddle".
- tremon 5y ago
  But aren't patents supposed to protect implementations of ideas, not the ideas themselves? So then "keeping state between autonomous devices connected via LAN" seems like a worthwhile implementation, regardless of whether all the individual components were already existing in other technology domains. Besides, I don't think that consensus-based asynchronous state protocols like Paxos were widely available in 2003 (and even so, implementing Paxos on a resource-constrained device like e.g. an Arduino might still qualify for a patent in 2021).
  - knorker 5y ago
    > But aren't patents supposed to protect implementations of ideas, not the ideas themselves? This is the problem with software patents. The border between ideas and implementation is not as clear. Strictly the answer to your question is no. Copyright protects the implementation, and patents protect the ideas. But where's the difference between an idea, and a design? For software, what's the difference between a design and the implementation? Sure, you may actually have a separate design (and not "the implementation is the design"), but in software the step from "design" to "implementation" isn't "manufacturing". Sure, even for physical goods, and their patents, the patent is often not sufficient to create a working system. For those holding software patents, they have to also list all ideas they have around that. Because there are patent trolls who check for new patents, and then file patents around that, creating a mine field so tight that you cannot even use your own invention. Analogy: you patent a laser pointer, and immediately someone patents all possible use cases for a laser pointer they can think of (point at things during presentation, aim weapons, exercise cats[1], project things onto a wall, stick the laser pointer in a handheld device, have a button on the laser pointer, non-chargable batteries in the laser pointer, chargable batteries, mains powered, capacitor powered, etc… etc… Anyway, a comment field is not sufficient to explore just why software don't fit well under the model of patents. [1] this is a real patent, by the way.
    - DoingIsLearning 5y ago
      > For those holding software patents, they have to also list all ideas they have around that. > Because there are patent trolls who check for new patents, and then file patents around that, creating a mine field so tight that you cannot even use your own invention. This is something I never understood. When someone submits those type of patents, the patent officer most certainly knows the 'prior art' original patent exists. So why is this not patent 'obviousness'?
      - thfuran 5y ago
        Because the patent office now pretty much just checks if something is the same as prior art and let's the issue of obviousness or usefulness be settled in litigation.

szszrk 5y ago

Have you had a closer look at that synchronize state stuff? I can't see much in the article. > devices synchronize state using a LAN If that has anything to do with music latency, then it's not "nothing substantial". It's their core business and their flagship feature. What they do well, that others could not, is playing multiple speakers perfectly in sync. Either stereo, or multiple rooms, almost as many speakers as you want, connected via wired network, or wireless, or by their on mesh system. And each speaker is basically a separate computer. It's not connected by analogue cable. Anyone who were looking for Sonos alternatives with that functionality knows it's very hard to find and suddenly their prices become somewhat justified.

cosmotic 5y ago
Slim Devices' squeezebox had this feature before Sonos. I implemented it in high school using visual basic. The most non-trivial part is that if you try to synchronize an MP3 or other encoded stream instead of the decoded PCM stream, you end up seeking to mid-frame points which causes invisible adjustments to the seek point. In the end, it's as simple as 'seek to this point' and then check if the point is off by less than some threshold, don't do anything. I did a few other fancy things like not seeking until the 'heart beat' lined up a few times in a row (duration in heartbeat was same as difference in time since last heartbeat).
KingMachiavelli 5y ago
I admit I only briefly looked at each patent but even looking a second time, I can not find anything more specific basically the devices share clock information and calculate the difference between their clocks. How is this different than syncronizing transactions between database servers or how GPS works? I found a GPL license multi-room/device audio player that discussed this very lawsuit last year [1]. The developer raises the same point than anything like this will involve something resembling clock syncronization. Looking through the lawsuit itself, about 90% of it is just comparing the end result of Sonos' & Google's products in that both let you add/group speakers, adjust volume and playback of the group, etc. I cannot find any claim that Google's implementation is a copy of Sonos'. It seems to try to sugest that is the case by bringing up that prior to Google's devices, Google had worked with Sonos and had access to details of their implementation. [1] https://github.com/badaix/snapcast/issues/642 https://github.com/badaix/snapcast/issues/642

IshKebab 5y ago

One patent is just on how to set up a playback device by connecting to it directly and sending wifi details. It's lucky they wrote "playback device" and not "device" otherwise the entire IoT industry would be violating it.