Hacker News Reader

Comment by fidotron | Hacker News Reader

This is a neat idea, however, I fail to see either why it requires so much funding ($1.7m according to TechCrunch) or any server backend whatsoever.

View on HN

The real product here is almost certainly an analytics service. From Cover's privacy policy at https://www.coverscreen.com/privacy:

  Cover records the following items: 

  The list of apps installed on your phone, including which apps are
  on your homescreen and dock

  The number of times you run an app, including the time and place 
  where you ran it (specifically, the cell tower or wifi network you 
  were connected to, or that you were located within a geographic area 
  you identified such as home or work)

  The number of times you turn on your phone (i.e. view your Cover 
  lockscreen), and what you do subsequently: launch an app, slide open 
  the cover, or simply turn the screen back off (e.g. just glance at 
  the clock).

  Information from connected Bluetooth devices, such as the Bluetooth 
  device ID of your car.

  Information from the sensors in your phone, such as the 
  accelerometer, gyroscope, and GPS (though GPS is used exceptionally 
  sparingly in order to save your battery).

  The system starts to collect data as soon as you install Cover. The 
  data collection continues until you remove Cover from your phone. 

  The data is uploaded from your phone over a secured connection to 
  our servers.

Cover is tracking every app you launch on your phone, when you launch it, where you are when you launch it, and even what angle you're holding your phone when you launch it. Privacy concerns notwithstanding, sounds like valuable data.

I'd suggest that a portion of the funding go towards a re-write of the Privacy Policy and Terms of Service. I'm by no means an expert, but I'm not one to mess with having holes in legal documents that could later bite me in the ass.

It's glaringly obvious they did not hire a lawyer to write their Privacy Policy or Terms of Service. I understand they mention "human readable" in the beginning of the Privacy Policy, but does this really hold up in court? In the Terms of Service, they make mention to "You" as opposed to "Client", "Customer", etc, with no definition.

Hold up in court to what?

Yeah, I was excited until I got the point about "our servers". Why would they need any data on their servers just to know when I need to read email?

But it's a good idea, hopefully easy to copy and release as an actual app (something I define as an executable that runs *on' a device, not a glorified json interpreter and exchange system).

There is probably useful analysis that they can perform in aggregate across all of their users. Perhaps people who read email are more likely to do <x> at a certain time and that signal is only noticeable with more data than you as an individual generate.

Seems that type of analysis is not really helpful in the long run. It should be catering to me, not what it thinks I need based on other people's habits. If my habits are completely different than everybody else's, then the results it provides are almost always off. Then it has to figure things out again solely using me, which it should have done in the first place.

It would be interesting to read the app permissions on Google Play when this app launches.

It sounds like it requires everything.

I could build you one of those without privacy intrusion, and add the ability to switch from home/work/car automatically using GPS.

I'd be happy with 1% of what they raised , anyone interested? :(

Judging by the reaction here, I think people would.

Do it and I'll donate money to make it happen. I am being serious.

... and I could help by suggesting that you shouldn't use GPS for trying to figure out home/work/car, but rather a combination of WiFi and accelerometer data. Using GPS often will drain a battery faster than your users will uninstall the app.

Don't need GPS - WIFI connection can cover home and work. I personally don't suggest any automation for car mode. It should be manually started / stopped for safety sake unless you have an accurately-made car dock.

Kickstarer campaign?

We could do it all together on assembly - I can fast track through to assemblymade.com/ideas

(Disclaimer: founder)

Dude, can I just say I wanna make love with the idea behind assembly? I just saw the website and video and it makes so much sense!!

I agree with you. I like the idea and concept, but I prefer that apps like this actually let me do some of the teaching manually. Perhaps try to define the apps I want to appear at certain times of the day.

One of the reasons I stopped using Aviate was that I felt like their scenes just didn't really match what i was doing at the time.

Just out of curiosity how long did you use Aviate for. I would assume their algorithm just keeps track of how many times you use an app at a specific location and then the premise would be after some amount of time it would reach a general consistency. To speed up that process It would be nice to be able to define a list of apps up front which have a greater weight and then hopefully it will reach a list of apps that you actually use over time.

I'd assume you prefer to do it yourself is because you haven't found a system that does it well enough for you to let go of control? Or are you just the kind of person who tends to like to do things the harder way?

I wouldn't blame you for being the latter - it's kind of like people who prefer to drive a manual transmission to automatic.

Or perhaps it's faster, easier, and more accurate to place your own five icons that to wait for an algorithm to try and figure it out.

Reply on news.ycombinator.com