Very cool animations on the UI side. I'm one of those users that wouldn't trust to share all of that data though. Like others pointed out, maybe if there was the option to run it on your own infrastructure...
I second this opinion. We're getting ready to enter the era of being able to host our personal data on our own infrastructure. Let's start making that available with all offerings, instead of making these centralized hosted solutions that all require different types of access.
It's taking a while to connect to everything, but I'm really pumped to see my data in this interface. I've been working on something similar — an automatic journal based on Moves/Instagram/Dropbox etc. One reason I'm scared to publish is data security — are you storing any of this data, or pulling from 4+ APIs every page load?
I made the April Zero site by myself, but Gyroscope has been a massive undertaking with many people involved.
@ericflo built most of the backend and API powering this, and some of our other friends like @mirashii and @wingfield also helped build some of the features. Our investors have also been a huge help, from @semil getting us great deals on our hosting last week to @jyri helping us with legal issues and @hiten being a brilliant source of product inspiration. It's a team effort and everyone has contributed so much in the last 6 months.
This is a product for tracking everything you do, and tries to accomplish that by pulling data from other services that track everything you do.
Is Facebook uniquely untrustworthy in this category? Is there anything about Gyroscope that makes it uniquely trustworthy, if only they gathered your data directly?
It isn't about being untrustworthy: I don't use Facebook, I use many other things but not Facebook. If he has the code to handle Facebook logins, he has the code to handle Google logins as well.
The main issue I have with Facebook is that their goal is diametrically opposed to the concept of the Internet that I would subscribe to. They are actively working toward attaining a monopoly for online social activity. In a sense, they want to become the "social wrapper" for everything that is done on the Internet.
I completely deleted my Facebook account a few weeks ago and quickly am realizing how far-reaching their monopoly already is.
This point alone - which doesn't even include the many, many privacy issues FB has - is enough for me to say I will not use a service that requires a FB login.
Let's not just assume the parent of your comment thinks Facebook is untrustworthy; how about if someone (like me, for example) just doesn't have a Facebook account?
I would like to use it but not reveal who I am. I think this is more than fair enough reason, and probably the direction these guys should move towards.
IIRC the friends list is in the default app permissions, and calling the graph API only returns the friends on your list who also gave the same permission to the app.
I don't like Facebook but I have an account because my mom and wife do too. So I choose not to connect my account with anything. Am I an unacceptable weirdo?
Even if I am in a tiny minority, I don't have a Facebook account and I would prefer to continue not having one
(motive: snobbism, more than privacy concerns).
So if a service is only available through a FB login I decline to use the service.
Common plight these days. I've considered making throwaway facebook profiles for these sorts of things before, but there's got to be a better solution.
Why wouldn't you make a throwaway Facebook profile? My Facebook account was initially created as my legitimate account, but now I treat it as essentially a throwaway. It makes logging in to most random little sites extremely handy, and I see no measurable downside.
If you want to remain any kind of anonymous, it means logging into each facebook login-enabled site with a new, fresh fb account. Using one trash account might stop them from linking things to your 'real identity' (such as fb perceives it), but it still lets them link all of your activities together and build a profile. Persona linking from there isn't hard, based on usernames or email addresses used (see pipl and similar for examples of casual, publicly available version of these tools). But even if there's no persona/pseudonym linking algo going on, you're still being 'tracked'. It's not fully anonymous.
I really hate it when websites override Safari's scrollbar and substitute a 5px one that may fit the theme of the site, but is virtually impossible to grab because its hotspot is only a tiny bit larger than the edge of the window.
Nowadays, I find the only interaction I have with the scroll bar is to look at it to determine how far down a page I am. I always use the scroll wheel on Windows or the two finger scroll on the Mac trackpad to scroll a page.
That said, I would pay money for a license to run a personal instance of this. The data aggregation you need to make it work would be much less scary that way.
Yes, I don't know why all this data has to be stored in the "cloud" and can't be stored locally. That way if you really want "the cloud" you could just use Dropbox or Google Drive or any of the other billion cloud storage services out there.
Absolutely beautiful design. I wish I had the skills to design something like this.
As a user, it would be fun to look at your journaled life years from now. However, the skeptic in me thinks advertisers must be drooling at that much data. Also wondering what impact could it have on potential health insurance providers. Or perhaps one day big G or FB would swoop up all my data.
All that aside, congratulations on a fabulous job. I'd definitely give it a try someday. Please enable sign-up without FB.
It's crazy to see how far @aprilzero has come with this. Actually bringing his idea to life, and turning it into a profitable business. There's a lesson to be learnt here in regards to working passionately and finishing your side-projects.
Designers who can engineer are a step ahead of everyone else IMO.
They have 133 customers paying $7/mo. I don't think that's quite enough to be profitable after paying for taxes, web services, and the team's salary; I'd be surprised if it could cover everything. It's a nice start, though!
I'm glad this is finally out! I signed up for the notifications about this back when it first started, but never got an e-mail about Gyroscope... so I'm glad I saw it on HN today.
I hope there are plans to integrate data from Google Fit. I know my watch tracks my steps and heart rate, so it seems kind of redundant to get Moves for the former and not be able to use data I have for the latter.
Interesting. I don't mind using Moves for that, but I didn't see anything under Vitals that indicated that it would use my Moto360 heartrate measurements.
Looking forward to playing with a real account once I get through the waitlist!
Also, I'm glad and impressed you guys are responding so quickly (I asked about FourSquare/Swarm on Twitter a little while ago).
Question: Moves (which I already use) tracks cycling and runs - why do I also need to link Runkeeper/Strava/Fitbit in addition? This appears to be redundant.
Other question: How do you get back into the "follow these steps" page once you've clicked away? The link on the front page has been replaced with a notification that i'll be emailed when my account is set up. (http://gyrosco.pe/setup/)
Moves does theoretically track that stuff, but it is not reliable enough. For example, walking quickly or being in a car often ends up being considered cycling for half a block.
On the other hand, if you turned on Runkeeper or Strava and say you're going for a run, that's probably 99% reliably what you actually did.
Also I think the page you want is the dashboard: https://gyrosco.pe/dashboard/ - we are working on building a more intuitive global nav so people don't get lost.
I also tried Moves and found it really bad if you're actually interested in accurate tracking your location. Runkeeper and Strava seem to use a lot of power because it's GPS only. Is there a GPS+celltower combo app you've found? I found GPSLogger for Android tries both GPS and cell tower, and stops when one of them returns with an accurate enough coordinate. So it gets amazing accuracy and uses very little power at the same time, since 95% of the time it uses cell tower.
Sorry about that. I don't think many people here are getting straight in, I imagine they're commenting on the homepage and the pages that other users have made publicly viewable.
We haven't automated the invite system yet and still have only sent out a few, but we plan to accelerate that process now. Scaling is hard so we want to make sure we get it right and everyone has a good experience.
I have been working with a couple of friends on something like this for the last couple of weeks as a fun/side open source project.
It sucks when you see that there is no point on working on anything because it is or will already be made by someone else. I guess we will carry on just for the sake of doing it, learning and having fun.
Any advice on how to deal with this constant struggle with my motivation?
By the way, congrats to the guys, it is an awesome piece or work!
A pre-cursor was launched almost 8 months ago [1] but perhaps you missed it. I can imagine looking at such a refined design and feeling like you can't measure up. But if you are really passionate about whatever your are developing, you'll find a niche. Moreover there's always room for more than 1 player and competition is good to shape up the landscape for everybody. If you share your open source project, you might get a lot better feedback and perhaps more contributors.
Yes I missed it, thanks for the link. In any case we are very far from being able to compete ha! We all have our day jobs and it is hard to find time to build our thing. However, I hope one day we can post here a link to a working version no matter how simple it is ;)
Regarding the opensource aspect, we are already working on Github but it is too early and hacky to be shown to other people.
Having used both, I personally prefer Strava. I've found it far better at motivating me to run. You can compare your performance against other runners along pre-defined segments (e.g. some hill you regularly run). The two people I've recommended Strava to are now hooked.
Their engineers make cool side projects [1] and release interesting new features regularly.
Can you consider integrating Nike running into gyroscope? I believe they recently introduced new API for that, and I already have hundreds of km logged in
Yeah that makes sense. We've started to switch more towards things like Rescuetime for that information, since the github auth options are very all or nothing.
I wish there were more granular scopes so we could use just the metadata or timestamps, which is all we care about and not nearly as sensitive.
I had only thought about this for Twitter before, but it would be nice if a trustworthy third party could act as a mediator between some of these APIs to give developers the ability to request less information or for users to restrict/filter what personal data is given away. That would help with this github all-or-nothing problem.
It says there are 5 steps, with the first four steps containing free apps, and the 5th step containing an integration to RescueTime, which appears to cost money. I don't particularly want RescueTime. Am I just going to be on the "waiting list" forever, then?
Also, please provide a way to cancel the account. :)
Rescuetime is free. They have a paid upgrade but you don't need it.
And you don't need to have all of the steps complete, just a majority of them. There isn't an exact threshold, we just want people to have a good experience and not start off by landing on a mostly empty page.
We've split the backend into a service layer written in Go, a data fetching and processing layer also written in Go, and a web layer written in Python/Coffeescript. We use Kafka to connect a lot of the pieces together, and data is stored in PostgreSQL and Cassandra.
@ericflo will probably pop by and answer a little bit more, but the major pieces are a golang powered API stack, a Django app that consumes that API stack and does all the template rendering and such, and a lot of background workers based around Kafka/Storm to pull the data from the third party services. @aprilzero can certainly talk more about the frontend stack.
He wrote a couple of articles about how he build the service for himself and it used Python with Django so it's safe to assume he stuck with that stack https://aprilzero.com/journal/making-of-aprilzero/ (he's whole site seems to be down now, perhaps the HN effect)
Dear @aprilzero, I'm loving the dashboard. Hope you can add in more integration with other services, like kanbanflow for productivity (it comes with a nifty pomodoro timer) and miband for fitness.
In Linux Firefox the 3d rotations are aliased and ugly. Could you just give us a normal web page? This web site has all the problems of bad flash websites but built in JavaScript.
It actually shows more than health data, it's like a personal dashboard for your activity, may it be physical (fitbit, jawbone and so on) or digital (instagram, twitter, foursquare)
Ahh amazing! We met one of the creators at Famo.us earlier in the year. He was truly inspirational and the css work he has done on April Zero is mind boggling.
We haven't fully launched yet - we've started sending out invites as we slowly scale up to everyone. But we just updated our site with more details about Gyroscope and are starting to accelerate that process now.
Scaling is hard so we want to make sure we don't break everything, but we're really excited to have everyone start using it as soon as possible and doing everything we can to make that happen quickly.
We're also doing multiple passes, starting with people who already have enough data connected, because we know they will have a great experience. If someone only has a few photos connected there and not much other data, we still have a lot more work left to do to give them a great experience.
Look up CSS3 3D transformation and transition tutorials for the basics. Then, a 3D tilted stack effect would go something like this:
Apply 3d perspective and rotate the parent/container element about the X axis ".container{perspective:400px; transform:rotateX(-45deg);}" (or maybe a positive 45deg)
Next, push the children forward along the Z axis ".child{transform: translate3d(0,0,100px);}"
Then, add transitions to the properties being changed, z-translate different amounts for different items in the stack, add some interactivity, and you're set.
Now I can finally quantify how little I do of everything that I should be doing. Before it was vague; now the guilt really can come to life with exciting (or rather, flat?) graphs and visuals.
I'm not going to argue for keeping address details on here that you don't want up. However, publishing a business address isn't really "doxing" surely - business addresses should be public information I feel.
