Interesting that in the beginning it says 'We know that many TVs use automatic content recognition (ACR)' but according to answers none of those manufacturers do.
So which TVs do the ACRs, and why Verge did not send those questions to them? Hisense?
So which TVs do the ACRs, and why Verge did not send those questions to them? Hisense?
They definitely do, I have a little snippet on their image recognition ACR from when I hacked my Roku TV here: https://blog.ammaraskar.com/roku-tv-philips-hues/#reverse-en...
I think they're using "collect" to mean something specific and "store" to mean something specific, neither of which fit our definition of collect and store.
Here's the APK in case you'd like to take a closer look: https://files.zajc.tel/unlist/6361bb6b86f2ed10e2e75a0deb3269...
That the Television Business transformed from a) an advertising conduit to b) a propaganda conduit to c) a live-surveillance technology tells you all you need to know.
Whether the Corporations answer your questions or not, their claims of doing The Right Thing are untrustworthy.
Literally, LG's answer in TFA says they do, despite the mealy-mouthed-ness around not capturing screenshots.
It is fundamentally a citizen data security and data privacy issue, which only regulation and the law can solve for (i argue, based on all available evidence and information). But asking nicely or trusting corporations and management? lol, rofl even.
(i am joking but the cognitive dissonance about telemetry is so stupid. you are watching TV on a subscription service, someone already knows who is watching what.)
you are running jellyfin on a device, like a NAS. it reports telemetry. docker reports telemetry. you download jellyfin, microsoft gets the telemetry. you update jellyfin, it gets the telemetry. you download radarr, docker and microsoft get the telemetry... all of this stuff reports telemetry. github, you can see your own repositories' usage and telemetry! graphs/traffic. it's all there...
Etc...
But there is a difference between the company whose services I use knowing about the service use, and the television mfg knowing about all of my activity.
Conflating that with actively spying on the content of my television programming is wrong.
you pay for a subscription, the subscription service knows what you are watching. 90% or so of people.
you download a torrent, bro, your IP address is shared in the swarm that anyone, without authorization, can see.
you are in a private tracker, every profile is open!
you visit some website for a stream. the website has tracking.
somebody, somewhere, somebody big and knowledgable and sophisticated in all the scenarios, not some little adversary but all these 1,000 or so ad-tech and analytics companies, knows what the content of you, you specifically, you are watching.
But my television - bought and paid for - doesn't provide me any service beyond displaying images. It is a commodity, and TV manufacturers are trying to get not just a slice (which would also be too much), but their own copy of the whole pie.
Jellyfin Server packages are hosted on repo.jellyfin.org, so the Jellyfin organisation gets the data that you're probably still running Jellyfin and want updates, not Microsoft. I will accept that if you use the Dockerised version, Docker or GitHub will get that data instead. None of this is "telemetry", e.g. detailed usage information. It's quid-pro-quo; you want to know if there are updates, you ask the update server, it records that your IP looked for updates.
Jellyfin Server collects data and stores it locally on your device in logs and its database (e.g. viewing data, login sessions). This data doesn't leave your device.
What Jellyfin Server _will_ share a lot of, is data about your collection, to its metadata providers. You chose those when setting up your libraries. In return, they hand back nice metadata and pictures. So TheTVDB.com, OMDB, MusicBrainz and others you've configured, get queried about "Film name (year)" or "TV series S01E01" or "Band name Album name" and thus learn everything that's in your collection... but not whether or when you've watched it.
Not if you run one of the "good" Linux distros and FOSS tools like Podman. Then there is no telemetry at all.
The AOSP (Android core) does hardcode some Google services like a fallback DNS. I patched that out of the source for an x86 Android build that runs on my TV box. So there can be zero or very little telemetry if you actively avoid it. You mostly need to avoid Big Tech stuff.
Just because you took the easy way out doesn't mean it's hopeless.
These companies get away with these things in part because they can point to users accepting some agreement by using the service where clause 12 subsection 4 addendum C says they can do it, using some terminology nobody reads or understands.
The EULA terms I did agree to likely had a clause allowing for returning the TV to not be a contract of cohesion. As for why I consented to ACR, I guess I have a lot going on and privacy (for, say, location data) in a contractual way is difficult.
Supposedly the LG TV does not send ACR data out without agreeing to the EULA. It gets into non-falsifiablity to just say “duh, of course they do.” Well, it’s what they say and there’s no evidence they do if the EULA isn’t agreed.
Now the fire stick and its services have their own EULAs.
The libertarian contractual ideology obviously falls apart in real-world conditions. I had to agree to a lengthy contract to pay for parking after I had already parked. I had to rely on unconscionablity or some practicality (I.e. assuming I didn’t sign over one million dollars in the parking EULA) to get through my day.