A commenter on the post just suggested auditing all the apps/sites with which you've used FB auth. This hadn't occurred to me, and frankly is even more scary IMHO. In theory, anyone who sufficiently compromises one of those apps would be able to access any of the permissions you've given the app.
Sound advice. If you used someone's side project on Hn 3 years ago, and they e.g. didn't patch Rails in January, they're rooted now, so your Facebook account is effectively rooted to the limit of all privileges you've granted. (That assumes you trust the Facebook security model for app permissions.)
I found this to be a great post on common startup misconceptions about SaaS/Enterprise sales.
We learned a lot of these lessons the hard way at awe.sm, not necessarily because of misguided principle or bad advice, but more because we didn't properly understand and prioritize what we needed to do.
A lot of startups seem to be enamored by the "consumerization of the enterprise" trend and think they can be the next Dropbox. IMHO if that's how you want to build your business, then you must also make sure your product and market are conducive to the dynamics that made Dropbox successful (i.e. a very simple product that could be deployed in a decentralized manner and still deliver value).
The coolest thing (IMHO) about what we're doing is that we've built it as infrastructure for others to plug into their apps. Everything (and more) that can be done through our GUI can be done via our APIs (see http://developers.awe.sm).
FWIW, I have seen mentions in some of the coverage that they plan an Android version (if that indeed would help you). Given the deep integration with the phone dialer and other device hooks as well as the in-app purchasing model, I can understand why they wanted to start with a single mobile platform (whatever it may be).
We're also looking for a great Platform Evangelist to own http://developers.awe.sm, extend our tools for developers, and help get the word out about what can be built with our APIs.
