I RTFA, I know this an arduino-based DIY hobby project. But if something like this ever comes out for real, the ability to doze off and pause your TV show will be a trojan horse...
Sure, downloading the Netflix pause-your-stream-when-you-fall-asleep app is comfortable, but it also provides a treasure trove of audience response data. Forget focus groups, now you have the real-time emotional response of many thousands of people A/B testing your original content in real environments.
And this ain't old-media Nielson, this is biggest-user-of-AWS technology-first Netflix.
Oh its a lot more sinister than that, and its already here. An app I tried to run on my Android phone would run a video ad before it started, it used the front facing camera to determine if someone was looking at the phone. I looked away and started typing on my laptop and the ad paused, I looked back and it restarted, I looked away and the ad paused. I looked back and deleted the App.
This is why I don't use Android. On iOS, you install the app and at some point you see a prompt "This app would like access to your camera. Allow?" and you can click "Don't Allow" and keep using everything else. Versus Android, which at install time asks "This app wants access to these 37 different things. Allow everything and install, or would you prefer to kindly fuck off?"
It isn't a fanboy war. The Android permission system is terrible, and users cannot get a sense of what apps are able to do. I've used Android since Galaxy I. I've never used an iPhone.
Given that the number of users for whom it's "fixed" rounds to 0 (and the proportion of apps targeting the v6 SDK probably also rounds to 0, so even if you are running Marshmallow it won't work for the vast majority of apps) I think that specifying that it's "fixed" is a tad overstated.
Well, fortune favours the bold. You can take that risk or you can be cattle for all the apps that want you to actually watch the ads they support themselves with. Make your choice.
On the AppStore, apps are required to gracefully degrade where permissions are not granted, and you can be rejected for not doing that. That seems unlikely to be enforced on the Play Store.
Can anyone confirm if that is the case with uber? On Android, if location isn't enabled, you can't use uber.
And before anyone claims location services are necessary to find the user and stop ddos attacks - I've ordered uber before for people in other parts of town, and I'm a signed in user with a track record of making payments.
As much as I respect not wanting to share location, the average user tends to be quite un-tech-savvy. Say that on average users without location services enabled makes 5x more wrong requests - that can result in hundreds of thousands of lost driver hours.
Yes, and if that were a problem, we would not have Uber for iOS. There are whole classes of apps that cannot be made for iOS (it used to be the case of custom keyboards & ad blockers, for example). Because security in general, and privacy in particular, are non-negotiable on iOS.
I haven't used Android, but culturally it seems like iOS developers try to deliver as much as they can with the permissions you give them. A voice recording app would probably just refuse to function without microphone access, but a video app will typically simply record a silent video if you give camera access but deny microphone access.
I don't know if this is enforced by Apple's review process or whether it's an emergent custom.
Actually, part of the problem is that they don't have to ask for 37 permissions... they are grouped together in such large groups that you don't know what permissions they are really after. Cyanogenmod has Privacy Guard that I think can help with this privacy nightmare.
"Facebook wants to send Sms" seems really odd until there's some feature like inviting someone that would obviously require sending Sms. Contextually asking for this permission then is much more reasonable, especially since the app can be used otherwise. Clearly certain permissions like camera may be the entire reason for the app, but the more sinister stuff like location tracking is often not the primary purpose.
Fandoy war? This basic privacy/security feature has been in iOS since forever. It is as common sense as you can get. Not having it is actually outright hostile towards users. I've used both Android and iOS for some time and I can tell you without a shadow of doubt that iOS is far superior.
In the meantime... talk to me when Android gets this basic feature in 6.0.
Granular permissions have only been in iOS since version 6.0, which came out 3 years ago. The first version of iOS was 8 years ago. For a feature that's "as common sense as it gets" we sure got along fine without it for half a decade.
I wouldn't say we were getting along fine without it. There were some pretty big privacy breaches[1], which helped move Apple forward. I think Google and Android are great, but on this topic they have held out for as long as possible, it seems[2].
its not really that simple, even on iOS. I'd like to allow facebook to access the camera so I can take a picture to send to someone directly from the app. But that means I've given it permission to access the camera period at that point. Not just permission to take a single picture. Similarly I'd like to give an app permissions to save pictures to the os level photo db. But the permission isn't 'save' it's 'access your photos'. Which means once I give it permission it can look at all the photos on the camera.
I'd personally prefer more fine grained permissions. Permission to take a picture (and read just those pictures the app took), permission to save pictures to the OS picture db. Permission to read pictures from the os picture DB the app didn't take.
I actually don't know if that would be any kind of solution. Another would be if the only way to access pictures was an OS level ui. That way the app couldn't access the pictures directly.
at the same time that would limit useful things apps can do. For example the fb messenger app showing half the display as a live feed from the camera.
Yes, that would be a possible evolution of permissions. If an app lets you choose a picture from your photo library, iOS swaps you to your camera roll where you select the image and then click "Share this image with app". You then get redirected back to the app and iOS sends the app only that image.
Currently, for things like camera, microphone, location, etc. you can always revoke that permission after granting it via Settings.
The solution here is to clearly tell those apps to fuck off.
Though, the latest version of android odes allow you to "allow" or "don't allow" features like that. Its the only reason that I finally installed the facebook app.
This is exactly what I need. I've made peace with the fact that Google knows much about me because I use their services and they made the OS and major apps for my phone. I have Facebook purely to get the occasional message or invite from a friend, and I hate how exposed to them I feel for needing this minimal functionality.
I do indeed... Though, I am on the new android, so its mostly a moot issue for me now. I also was using the mobile website before, so this may not have provided much for my use case. Thanks for the link though!
Even such prompts are disingenuous. They amount to a salami attack against the user's privacy, because the overwhelming majority of users will authorize Yes on a per-app basis just to take a look, without really considering the potential consequences.
See also: server admins who accept SSH host keys without verification.
The people who mindlessly click 'Yes' are generally the ones who don't care about the privacy implications of enabling such things though. I know I care, and I have uninstalled several apps after they've started asking for ridiculous permissions to enable non-essential features that I don't care about.
Sure, but we need to be thinking beyond ourselves. You sound pretty pleased with your own approach to software hygiene, and good for you, but it's easy to be rigorous when you understand the implications. Educating end-users won't move that needle much. What I'd like to see is a mobile device stack I could recommend to my dad, with high privacy assurance that he can't inadvertently compromise.
Smart pause is a feature you have to explicitly enable. Just like smart-something that prevents your phone from auto-locking after a certain period of time if it sees you're still looking at it.
Good point, but any app that runs video ads before starting up is not worth keeping in any case. For that matter, any app that forces you to watch video ads, period.
I can say I've had a lot of moments where I run into a situation in my actual life that makes me feel like I'm living in an episode of this show. It's a modern-day Twilight zone, and although I wouldn't say it's always a hit, when it's good, it's really good.
That is a great show that demonstrates sci-fi's greatest strength. Its ability to speak about our current time via metaphor, allegory, and extrapolation.
I thought this was a standard feature on Samsung Galaxy phones? It's not for ads but rather to just pause the video when you look away. It will pause ads if that's what you happen to be watching.
once again, a "Black Mirror" episode got it right. In "15 Million Merits", they pause the ad when you close your eyes or look away and will keep saying in a stern voice to look back at the screen.
I can imagine a world where the eventual primary use for type of technology is for for ad-based content where the screen pauses when you stop paying attention to the ad. No worries about that with Netflix, at least.
There's a start up named Affectiva which does exactly that. Emotion measurement from video which is used for, among other things, targeting ads and marketing campaigns.
Amazing point! Getting that data will be incredibly important. Adding play/pause/next options to your watch to control your smart tv would also be great too. Also allows another reason to download the app for the watch.
Climate change is incredibly important. Net neutrality is incredibly important. Growing wealth inequality is incredibly important. This kind of data is a business model, nothing more.
Just to clear something up - Nielsen might be an old company but they have plenty of technology that can be deployed - just depends on what their business customers are looking for.
That really only became possible when they decided to auto-play new episodes, which is a bad feature in general. There are quite a few shows with really nice music videos in their ending credits. Shrinking the video to the size of a postage stamp and kicking me into a new episode is a rather annoying.
They also have a terrible habit of skipping the opening credits. The credits are often funny or beautiful. Every time I watch Futurama, I need to start the show and then rewind back to the beginning to see the opening joke. It's like VHS all over again.
Opps. I'm ranting. You know, I never noticed how much the interface annoyed me until I stopped to think about it.
It depends on the show. I think they only ever skip opening credits right at the start of the video, but most shows tend to have a scene before them, so it rarely applies.
They also use the mechanism to skip "previously on" sections, which is probably a good idea during a marathon, since they're usually more spoiler than reminder in that situation.
It definitely skips the opening credits on the shows my kids watch (Octonauts, Cat in the Hat, etc.). For better or worse, both of these also have another song in the episode near the beginning (the same song, every episode).
This is on a Roku 3, not sure if it's a different experience for other devices.
Auto-whatever interfaces in general need to hit an incredibly high "success" rate in terms of doing what the user wants to avoid feeling frustrating. I don't remember the study, but the number was in the high 90s, percent wise.
For example, I can go a whole afternoon some days without it auto-pausing, while others as you said, you can go two episodes and it will pause on you.
I suspect Netflix tracks window focus via JavaScript at minimal, and maybe other metrics too to try and guesstimate user idle-ness.
Interestingly when I am in a full screen game (and Netflix on another screen) it "never" pauses. It only pauses when I am not doing anything on the machine and just watching (i.e. I effectively get punished for paying too much attention).
My theory (based on 0 facts) is that they will ask you if you're still streaming on shows that experience a high load and want to save the servers streaming that particular episode.
As someone who works for a CDN who does streaming, I don't think this is the case. Popular videos would be streamed from multiple servers, and it is actually easier to stream the same video to lots of users vs a lot of videos to only a few users each.
they might have a "binge-ness" score for every show. If they know people usually don't binge more than three episodes of Scandal then they can safely assume that the TV has been left on when the fourth episode kicks in.
They may even construct segmented binge patterns where they know episodes 3 to 6 of House Of Cards usually gets binged where as episodes 1 and 2 usually do not.
Building and training these segments would appear random to the end user but after awhile should be good at guessing when the TV has been left on or the user is binge watching.
Perhaps they don't want Netflix to be used for public screens (laundry, hairdresser) and randomization is necessary to check for the presence of a human.
The same thing happens on the Sony Blu-Ray player I watch Netflix through (via "apps" or whatever). Usually after more than 3 episodes maybe, with no interaction from the remote.
It has been way less than 2 hours on Chromecast, but that isn't a scientific test. My anecdata is that they seem to do it when you least want them too, yet keep playing episode after episode when you turn things off and Netflix didn't get paused for whatever reason (see this on Roku as well which doesn't seem to detect that the HDMI receiver is switched to standby/off).
I've noticed that they seem to start prompting more and more rapidly. If I'm binging on a new show, I can start Saturday morning and not get prompted until shortly after noon, say, but then I'll get prompted at 1:30 and then 2:30 and then after the next episode (numbers made up, but you get the idea).
This is really annoying, and the opposite of the sort of smart adaptive behaviour I'd like to see. It'd be great if Netflix could detect, 'this guy seems to have regular work hours (because he normally watches Netflix at a regular time), and appears to be binge-watching a TV show; let's not bother him.'
But then, the Netflix app is still terribly primitive compared to their website, so…maybe I'm asking for too much.
I might be making this all up, I'll need to actually test it out now. I swear when watching Netflix on a computer, you start an episode and your cursor is auto-hidden. At some point, maybe even when the second episode starts, I swear the cursor becomes visible again on the screen. I then instinctively giggle my finger on my touchpad to make it hide again. After watching couple episodes in a row and doing that, if I don't I feel like it pauses and ask if I'm still there within 20 or so seconds.
Has anyone else noticed that? My theory is that they reveal the cursor after every new episode to try to get you to move it and hide it. If you do that each time, they could assume you're not paying as much attention.
You know, I was actually the same way and instantly started searching for ways to disable it to no avail (aside from a simple setTimeout grease monkey script). After I've had it for awhile, I've come to actually enjoy it because a lot of times I've been on the verge of passing out while watching TV, wishing I could turn it off without moving and breaking out of my close-to-falling-asleep state.
<rant>That pisses me off to no end. What's even worse is that now some of their movies start playing as soon as you select them from the main screen. I'm still reading the description, trying to understand if I want to watch it, and while I'm reading, it fades out and starts playing anyway. Of all the dumb things they could do</rant>
See, this has been a solved problem for literally decades and this is the only comment in the whole thread that has gotten anywhere near it. The idea that I need socks that can talk to my TV for _any_ reason is just nuts.
In Thomas Pynchon's novel V, there is a fellow who has a switch on his skin that detects whether or not he is awake (by the resistance? I forget), and so controls the TV. Judging by the Wikipedia article, this is Fergus Mizolydian; I don't have a copy lying around the office.
Instead of flashing an LED, the socks could warm up. That way, if the socks start getting warm and you're still awake, you can react to that, but if you're asleep, your feetsies will be nice and warm.
Sounds like the hardest part would be making sure it isn't a fire hazard. That being said I don't like my feet being warm while I sleep, so this project misses for me by being a sock. So close.
Pretty much no one moves around during deep enough sleep, it is just when your sleep cycle lightens up that you move around in your sleep. The same tech still works, it just might not be able to detect all the stages of your sleep as effectively.
I'm one of those people. Falling asleep on the couch under a big blanket, and waking up every hour or so to catch a few seconds of some Star Trek TNG is a pretty profound sort of comfort for me.
It reminds me of watching old James Bond marathons with my dad when I was a little kid, and falling asleep during them.
Huh, my dad has sleep apnea. What about waking up to Star Trek makes you think that, though?
What I mean by waking up is like: do you ever wake up in the middle of the night to the dog wanting to go out, or a noise outside or something? You come awake just enough to process what is going on, and then go back to sleep right away.
For instance: sometimes I'll wake up to the really faint sound of somebody jingling keys outside the kitchen door, but I can figure out that it's just our roomate, and to go back to sleep. Like a very low thread priority or something.
The way you described it sounded like you just randomly wake up in the middle of the night, which _can_ be related to sleep apnea. If you're waking up that easily it _could_ also mean that you're not going into deep and REM sleep because your brain needs to keep you at a certain level to wake you in case you can't breathe.
Edit: or you're just a light sleeper, but generally waking up several times a night is not common.
The website, http://makeit.netflix.com/ asks for us to submit our own ideas. Most of the things I think would be cool to do with Netflix involve having some form of API access to its catalog or apps. IR socks are not Netflix socks.
I suspect something more accessible to the general public could be done with the heart rate sensors and accelerometers already in Android watches, Apple watches, and Fitbits.
What about socks for truck drivers to detect if they are dozing? Have them set off a phone alarm, etc. Better yet, can FitBit, etc. detect dozing and use the vibrating alarm to alert?
While I think an alertness sensor is probably a good idea for drivers, I don't think an accelerometer is a useful way of going about it. These don't really notice that a user is asleep until they have entered a fairly deep sleep, which is fine for pausing a TV show but far too late for alerting someone who is operating a motor vehicle.
With Netflix if the accelerometer doesn't notice you've fallen for 5 minutes before pausing, that's totally fine. If the user is operating a motor vehicle and has been asleep for 5 minutes it'll be a miracle if they're still alive.
It's solved in a much nicer way in modern vehicles. Where they detect # of line crossings (dozing makes you more prone to cross the lines on the road), and how often and for how long your eyes are shut when blinking.
Together they indicate tiredness, and can recommend that you take a break.
Atleast that's how I remember it working on Mercedes cars
You don't, you just knit new ones, download the pattern, do a little programming, solder some electronics and configure your remote and VOILA! New SOCKS!
It's cute, but is it a joke or something more serious ?
I don't see any use for it, except perhaps saving bandwidth.
Surely it must takes some time for the device to find out that you're actually sleeping, then you anyway have to rewind back to the point you stopped watching, so I don't think it makes a big difference to go 10 minutes or 1 hour back.
This is part of Netflix's "makeit" projects[0]. People can submit ideas for DIY projects that incorporate Netflix in someway, and they pick ones they like. So I believe this would fall under community engagement.
It's probably hopeless generally, but thought you might like to know that Huxley himself rejected BNW later in life and offered a new take on things in Island. So it's not really fair to the author to use Huxleyish as a synonym for BNWish. Sorry I don't have a cite, read this in a biography of Huxley that had a cite. The wiki page for Island hints in this direction though: https://en.wikipedia.org/wiki/Island_(Huxley_novel)
I thought title meant some socket library or socks proxy? Came here to mention how, out of context, people would be confused wondering what Netflix foot underwear has to do with networking.
For a moment I stopped to wonder what kind of mind reading capabilities their clients have, if even their sockets could tell when the viewer isn't receiving :)
Please don't use Netflix and other paid streaming services. By doing this you support Digital Restrictions Management development and finance copyright lobby.
If you want to watch their shows too much, download them illegally via file sharing services. They can arrest a very limited number of people, and by engaging in such activity you lower other people's chances to be persecuted.
I would argue that you can choose to watch these streaming services and support DRM, or you can speak with your wallet and refuse to watch their shows. Pirating the shows is pretty unambiguously unethical IMHO.
My position is that I don't really care that much about DRM on something that I'm not buying in the first place. With Netflix, I'm not purchasing TV shows or movies, I'm just paying a flat monthly fee to have access to view their entire catalog whenever I want. So to me it doesn't really matter if there's DRM, because it works, and when I want to watch it again, I can do so for no additional cost. I just have to pay the monthly subscription fee.
For something I buy, it's another matter altogether. I don't want something that I own to be limited in how I use it, like only playing on certain devices, or a certain number of times. So as a result, I'm not real wild about spending any money on Blu-Ray discs, which had strong DRM and can only be played on certain players or with certain software. (AFAIK, you still can't watch Blu-Ray movies on Linux, which to me is pretty much a deal-breaker since I use my laptop for things like that constantly, including watching Netflix.) Worse, because of DRM, these limitations could change in the future, which is entirely unacceptable.
I know it's a bit of a gray line, but I just don't see subscription services the same way, because I don't actually own anything there, I'm just buying access to a store of online content.
You've seen those charts where people use their smart watch to record their heart rate during the game of thrones finale? (No? Here you go: http://blogs.wsj.com/digits/2015/08/13/what-game-of-thrones-... )
Sure, downloading the Netflix pause-your-stream-when-you-fall-asleep app is comfortable, but it also provides a treasure trove of audience response data. Forget focus groups, now you have the real-time emotional response of many thousands of people A/B testing your original content in real environments.
And this ain't old-media Nielson, this is biggest-user-of-AWS technology-first Netflix.