FB: Facebook is committed to both Android and iOS, and you’ve now seen what we can do. Stay tuned. (There are things in the Android app that aren’t in the iOS one today, like mentions in posts and comments, photo multi-upload, event creation. Since version 1.9 the Android app has tested faster than the iOS one, but FB-iOS 5.0 obviously changes the game.)
Android presents some unique challenges for developers, especially those with a large user base, but we do and will power through them. A lot of time is spent dealing with device-specific issues and limits, and you really have to fight with the toolkit to get iPhone-smooth interactions. Some vendors have a different HTTP stack (!), none implement the Camera APIs consistently, and reliability of hardware acceleration is…imperfect, GC pauses are terrible, lots of the toolkit insists on doing real work on the UI thread and allocating recreationally. On iOS you can test on 5 devices and basically have the market covered. We have to test on many dozens to get to the top 1/3 of our users, and then the tail starts to getreally long.
Q: Do you know when we could expect a native app too? I realise you probably can’t give away too much but are we talking weeks? Months?
A: Nobody is more excited about the state of our current development version than we are, and we will get it to users as soon as we can. One of our awesome PR people is standing next to me (10,000 miles away) with a gun (frowny face), so I can’t say more. Also, I have been doing software 20 years too long to make estimates in public. Experience and quality determine the time. I am utterly confident that you’ll find it worth the wait, and I wish I could give it to you today.
Q: Pretty much everyone I know has problems with the app even loading anything at all, and that’s before complaints about performance issues. This has gone on for ages, how come something as important as this has been unresolved for so long?
A: There are lots of reasons that people can experience problems, and we work through (and fix) different ones all the time. We have pretty detailed metrics on different aspects of performance, stability, load-time, load-error, etc. We can see them getting better in meaningful chunks, but that spreads out across 130M users in a way that isn’t to anyone’s satisfaction. (This is one area in which neither the inherent characteristics of the Android webview nor the OEM-specific tweaks that occur are our friends. Really, they aren’t even cordial.) We’ve been on fixed-date release cycles since 1.9, and we’re now down to every 4 weeks (where we’ll stay); this was a shit-ton of work for a large number of people, but it means we can get improvements out to users faster even while investing in longer-term features or architecture changes.
Q: What’s the best way for us, as users, to make it known that something isn’t working right with the app and make a difference? As of right now I could make quite a list of things that don’t work with the app, however it would feel a bit like a drop in the ocean for all the feedback you probably receive and feels like it would go unnoticed…We want to help :)
A: There’s a bug reporting mechanism either via the website or the app that someone on my team reads and rolls up for the developers weekly. My internal build has another mechanism for it, so I embarrassingly can’t tell you in more detail right now. :-/
Q: Is it true that Facebookers have been forced to use the android app to get a feel for how bad it is?
A: Neither the assertion nor the implication are true to my knowledge, and I am virtually certain that my knowledge on this issue is complete.