Friday, May 13, 2011

Thoughts about Honeycomb and the new course of Google

We all are pissed. Yeah, we are. No Honeycomb source in the foreseeable future.
This is hard to swallow. It is even harder because it was confirmed by Andy Rubin himself (and later by Jean-Baptiste Queru on Twitter, one of the leading Android devs at Google, responsible for AOSP)

Google doesn't make honeycomb open source for one simple reason:
So china-tablets can't ruin the reputation of Android tablets in general.

Just remember what happened last year: A sheer overwhelming amount of easily affordable but cheaply built and low performing Android tablets flooded the market. While the tech-blogger community posted one hands-on video after another, looking for an "iPad-killer", the tech world was almost certain that Android would never succeed as a tablet-OS. But this wasn't Google's fault. Android's full code is and was back then freely available for everyone, so if you built a shitty product, everyone just saw "the Android OS" as a failure, not that you built a shitty product.
Android's reputation took severe damage during this time and Google had to react.

They had to find a way to keep Android open but also repair the damage that has been done to the Android Ecosystem. So they came up with Honeycomb, specifically designed for tablets, but only made open to a small circle of companies that had to apply for access to the Honeycomb code.

It is like Google would say: "Only tablets running honeycomb are original Android experience tablets." It is a quality seal.

Google chose that path to get better reputation on the tablet market and to gain share against the Apple iPad and BlackBerry's playbook. They knew they would piss a lot of people off by doing so and that they would be criticized for their (temporary) course.

The official reasons for the closed Source are quite stupid, but most people swallowed them (also some certain tech-bloggers I don't want to name here)

It sucks for the development community. It really does. But what would be the alternative?

If you are a software engineer and want to build an application specifically designed for tablets, you have to evaluate if the effort/revenue ratio pays off. Building for tablets means additional effort, so you are looking for a stable, well maintained platform that shows continuous growth. The platform has to be "mainstream compatible" to show significant revenue, what means especially, that it is (relatively) frustration-free.

A big point is the hardware that the platform will be running on. So Google had to gain trust of the hardware manufacturers first. They did that by making an exclusive deal with Motorola for building the XOOM. It should be the flagship... a good visible invitation for other major companies to hop on board. And this strategy was a success. We're seeing a lot of high quality, powerful Honeycomb tablets surfacing on the market and this is just the beginning.

The new course, as controversial as it might be, doesn't change much for the Open source spirit of Android if you take a closer look. They fulfill the GPL by open-sourcing all the GPL bits (you can find them in AOSP right now) but this is less than 10% of the original Honeycomb code.
It is kind of disappointing for us, the dev. community, that some devices which should in fact deserve a working Honeycomb port like: The nook color, The Adam Ink our belover Galaxy Tab, the GTab and others, won't get a clean build so soon.

Honeycomb might or might not ever be fully released into AOSP, but they  will definitely (that IS confirmed) fully publish every single bit of Ice Cream Sandwich in Q4 this year (which also contains all of Honeycomb merged into this release)

Ice Cream Sandwich, or Android 4.0 is a merge of all platforms and releases that are now in separate branches:

  • Google TV (Android 2.x with full Chrome Browser)
  • Honeycomb (Android 3.x especially built for tablets but lacks some smartphone functionality/optimization)
  • Gingerbread (Android 2.3.x including NFC capability and ADK thingy-stuff)

And I dare to speculate that Ice-Cream Sandwich will be also containing the infamous new big thing called "Android@home"/Project Tungsten. A platform making it possible for Android to function as a hub for all kind of appliances and also as home-automation system.

Merging the platforms has some practical reasons. 3 different source-trees are hard to maintain. Every major change has to be merged into each branch and maintaining intercompatibility becomes more and more difficult over time. This might be what you can refer to as fragmentation.

[And dear Captain Crunch: Different user interfaces do not mean fragmentation, it means freedom of choice, that is something GOOD. Like if I want to have several ways of commenting on a blogpost, instead of being forced to use shitty facebook for authentication.]

I know it is a shitty long wait, but it will be worth it. 
In the end it will pay off for us, because if the manufacturers and customers all over the world gain trust in the Android Platform as a good, solid, well supported (and most important an open) OS for Tablets and other devices, the dev. community gains strength and support too.

Furthermore the open spirit of the Android platform gains more momentum by better acceptance among the users and major hardware developers worldwide.

So, even if it might seem like a punch in the balls for us Galaxy Tab owners, we will profit from it later on.
I know this is hard to understand and you can be damn sure that I was raging when I heard that Andy Rubin himself said that Honeycomb will remain (to large amounts) closed source until further notice, but... what the hell. I'll donate all the money I can give to developers like XDA's Spacemoose [link] that work on inofficial Honeycomb ports, based on the SDK's Google has published and enjoy a great, unofficial Honeycomb SDK port until those Mountain View people release the next big thing. And it will be all open and awesome.

If you haven't seen the keynote yet, head over to youtube and watch it today.

As a little apology for their unfriendly decision to let us wait for so long, Google is working on an agreement with the hardware-manufacturers to guarantee 18 months of updates to the latest version of Android (if the hardware supports it, of course). And stop hyperventilating, because every single time some stupid techblog vomited something about minimum hardware-requirements, it was just made-up bullshit.

So support spacemoose [link] and Dip7 [link] with all the spare change you can give.. We can build something awesome too, with or without the support of Google.
They have their reasons to temporarily close the development.

But WE are the community and WE will get our Honeycomb, it is just a matter of hard work, time and brainpower.

We will succeed. :D

spacemoose's Honeycomb CDMA port for Samsung Galaxy Tab -> here
Dip7's Honeycomb GSM port for Samsung Galaxy Tab -> here


Andrew said...

Great post! It's obvious you put a lot of thought into it.

My views are basically yours, although I won't forget the Honeycomb source blockage anytime soon to be sure.

I'm most excited about Google TV, it combines Google's two largest platforms (Android and Chrome) into a single box. Well, potentially. Even though the screenshots released at I/O show a new icon, I haven't heard a peep about it supporting the Web Store.

Florian Rohrweck said...

@Andrew yeah Chrome-Webapp-store for Google-TV was one of my predictions for the google-IO... But I guess they postponed the idea because they want to create a specialized category or even a separate market that ensures that the webapps will look good on a TV too :)

Andrew said...

Ah OK, that makes some sense. Still, I'll be disappointed if I can't deploy .crx files to it.

Florian Rohrweck said...

@Andrew: I bet you will :D But I guess it is just an additional category on your developer dashboard... as a little reminder "hey, you better test if it works on GoogleTV and if it doesn't look like crap" :D Wasn't there a session at the I/O about Webapps for GoogleTV?

Andrew said...

Yeah but it was only pure webapps, not Chrome apps.

Florian Rohrweck said...

@Andrew: Bummer... Huh, I suspect it will take until summer before we can see something happening... but hey it is chrome :D chrome without extensions and webapp market? I'm quite sure it will come :)

Post a Comment