Chet Haase (ex-Googler and the author of the book “Androids: the team that built the Android Operating System”) talks about Android history and the team that built it, mentioning Be and Dianne Hackborn.
Some interesting parts:
“One of the main things in most projects on ‘why did it work?’ is about timing, like if you’re not doing that thing at the right time it doesn’t matter how good or bad it is, it ain’t gonna work because nobody is looking for that thing.
…
There were other elements that really made it work as well, and one of them is obviously the team. The right people were there by which I mean they hit the ground running. The people that Android hired at that time for the most part were domain experts in platform development, and there’s not a lot of those people.
Some of those people were at Be writing an operating system that I think of as a cult classic. Like if you’re a consumer outside tech you have no idea what I’m talking about right now, but the people who are into operating systems know Be. It was a really interesting multi-threaded operating system, multi-threaded UI and graphic stuff, really good for Media stuff. And they continued to not succeed for about a decade until they finally fell apart. Palm acquired the IP of Be and those people then got spun out of the PalmSource.
And this meant that people who came to California or the Valley to work on operating systems for the sake of working on operating systems were now part of the Android team, and got to apply their expertise and their passion into this new operating system possibility.”
“Running up to Android 1.0 the Window Manager was one of many things that Dianne did on her own, which is more impressive the more I think about it. She was responsible for a lot of framework code and a lot of the more gnarly bits.”
Chris Simmonds (the host): “Dianne we’re talking about here is Dianne Hackborn who was one of the early members of the team, and who did so many amazing things including of course the Binder, which is still one of the unique features about Android.
I still can’t quite get under my head, why it is that everybody doesn’t use Binder. Why does it not scale beyond Android but it is one of the things that makes Android applications interoperate with each other seamlessly?”
Chet: “There were debates at the time, and there probably continued to be debates on whether that is the right solution for everyone for an operating system. I think Dianne won that discussion.
Binder was an open source project that came out of some of the early work that some of those people had done, and then they expanded on it to make it do the right things for Android.
But there were people there who preferred a sort of a lower level more focused approach. Binder is language independent thing. You send these messages through, and I think it was blocking at the time, and then they found a way to not make it [blocking]. I can’t remember, but there were like various debates about how the thing should work. But there were people on team that were like: “I just want a socket, just give me a socket, and I’m going to pass my own messages, I don’t need a whole framework thing”.
So, to some extent it was a more complex solution than some of the people wanted, but the framework team essentially won that particular discussion, and Binder it is.”
p.s.: I’d have loved to have Binder in Haiku, but I think that ship has sailed.
I’ve said this a few times, the spiritual successor to the BeAPI is the Android API. There are so many parallels. Loopers, Binder, the way Intents are basically BMessage in disguise. The more you dig in to Android the more obvious is is. Unfortunately, Android has become more and more locked down and opinionated about security and multiprocessing, that it is no longer pleasant to use te API.
My own two cents: I feel that Android killed Google’s excellence. It was the first non-next-gen non-original google product, it was rushed out to catch up to iphone ios. It has innovated since then, sure.
