Hacker News new | ask | show | jobs
by johnvschmitt 4634 days ago
This is why I'm really hesitant to use popular libraries like Flurry.

I'm making a kids game & really want to respect the kids' privacy. I can't hand the keys to all that data & possible backdoors to some "free" third party library & just trust they will play nice.
1 comments
gtufano 4634 days ago
After giving some thoughts on your same issue (customer privacy and security), I moved on to use Piwik (http://piwik.org/ a well known opensource analytics server) on a personal server. There are also native libraries for mobile usage (I use PiwikTracker https://github.com/mattiaslevin/PiwikTracker for iOS and OS X). Less bell and whistles than Flurry, but definitely a more controlled environment. ;)
link
vdaniuk 4634 days ago
Piwik allows site owners to track and uniquely identify visitors using their IP adresses. If I were concerned about my privacy I would trust Google more than a random website owner. Just a point to consider.
link
gtufano 4634 days ago
You can always identify your customers though your application without any external library or effort. The point is that (your analytics provider) can track you through all the apps using the library. This is a significant difference, IMHO.
link
josephlord 4634 days ago
The developer can obviously trust themselves and the user has trusted that particular developer (and the app store approval process) as they can already use the phone features allowed to that app.

If you use Flurry/Crashlytics/... then that is an additional entity (of many people) that both the developer and the user are trusting.

link
aw3c2 4634 days ago
Piwik allows you do not log any number of bytes of an IP address. So you can eg. make it log and display 123.213.x.x instead of a full IP. This is vastly better for privacy than sending all your visitors details to Google in the USA.

What kind of trust are you talking about?

link
krapp 4634 days ago
To be fair, every server logs an IP with every request anyway.
link