Facebook get its share amount of bad PR (some are well deserved), but we shouldn’t dismiss amazing engineering work because of those. This is a technical piece that highlights solutions to problems not many out there get to solve.
Uber's technical blog posts some amazing stuff - their work on using satellite rssi data and GPS almanac info to detect multipathing in gps signals reflected from highrise buildings to more accurately determine which side of the road a phone is on is _magnificent_.
But they, like Facebook, deserve every bit of bad PR they get for being truly awful corporations.
The thing is, this is clearly for marketing purposes in the end. The approach is complex and sophisticated, but certainly not novel. A good architect in any decent interview should get to a similar solution. This must be read through the lines: this is an attempt to try and change their bad PR cycle.
I'm not saying the parent comment is absolutely essential, but I'd rather see that instead of people using this to change the conversation from central flaws in Facebook's data practices, business model, and ethics in general. This may sound a bit harsh, but at a certain point you have to really consider what giving publicity to a group gives them versus the inherent value of the paper.
