Android is not open.
Android is open.
Both of these are true statements, given a certain definition of open. The question is whether either tells us anything particularly useful about platform outlook. The primary utility of debates about the relative openness of a given platform seem to be as soundbites in service of agendas for or against the platform in question.
Besides its central role in vendor marketing efforts or disputes between fans of competing platforms, however, it is unclear what the practical implications of openness are in a mobile context: unlike traditional software markets, openness has not been predictive of user adoption and/or volume opportunity. Depending on how you define open source, anywhere from 63% to 98% of the mobile market belongs to closed platforms. Historically then, the correlation between developer interest and platform openness in mobile has been weak. Not least because there have been effectively no relevant open source volume platforms in mobile prior to Android.
Consider that, as one friend observed, in 2005 two well capitalized companies picked the Linux kernel as the basis for a mobile platform. Maemo (which was subsumed by MeeGo) has to this point failed, Android has succeeded. Widely regarded as being a more open project, neither Maemo nor its successor MeeGo’s relative advantages in codevelopment were sufficient to offset other market factors which favored closed platforms.
With respect to questions of open and closed, no project is under as much scrutiny as Android. Scrutiny that has been earned, however: Google’s outspoken defense of the project’s governance has predictably set high expectations for source code availability and the ability to participate. The benefits of this approach are debateable: leverage of the term open appears to offer marginal soft benefits in mobile against a hard cost of negative public relations and developer sentiment.
Ars Technica’s Ryan Paul speaks for many with the following:
Sadly, those promises were never fulfilled and the dream of an open mobile ecosystem around Android never materialized. In reality, Android has become an insular platform developed almost entirely behind closed doors in an environment that is hostile to external contributors and is mired in a culture of secrecy that serves a small handful of prominent commercial hardware vendors and mobile carriers.
If we accept those charges for the sake of argument, however, the question remains: what does it mean? Does it matter if Android is open or closed?
Free software advocates would likely argue that it does, as would developers and enthusiasts wishing to apply open sourced Honeycomb code to a wide variety of platforms from handsets to tablets. And developers are, remember, the New Kingmakers.
But while developers are unquestionably and understandably disappointed, there is little evidence to suggest that a less than open Android will have a material cost in developer traction associated with it. Apple’s iOS, a platform that is not open source, has immense developer traction with over three hundred and fifty thousand applications available at the moment. And Android itself, increasingly attractive to developers as it represents the volume shipping platform, has in fact been developed in a closed fashion previously. Precedent for the decision not to open source Honeycomb immediately includes the initial code drop, which was developed behind closed doors.
In most software markets, open source is an excellent strategy for both developer recruitment and achieving broad platform adoption. In the context of mobile platforms, however, its track record for success is limited. Not that this implies permanency: there was a time when open source software within the enterprise was rare. But the short term forecast for mobile platform development does not include a similar wave of open source software. Absent more open projects that represent legitimate competition, then, it is difficult to build the case that Android’s governance will read on its outlook.
There is a cost attached to a lack of openness, as Google and Android enthusiasts are both aware. If the current market trends are to be believed, however, that cost is not likely to involve platform traction. Given that, Google’s messaging approach with the project is open to question. Google correctly believed that Android would need to distinguish itself from iOS. What’s becoming clear is that openness was a suboptimal choice for that differentiating feature.
If nothing else, however, it’s given us some funny tweets.
March 30, 2011 at 10:01 am
I suspect that for the commercially driver side of Google the important pieces of "Android" as it's viewed by consumers are the Google Marketplace, and the various Google apps, none of which have ever been open. The OS itself has been vital in getting these various components pre-loaded onto a wide variety of devices – saturating the market, and reducing the risk of Apple controlling the market and taking users way from Google's services, but it's a means to an end, not the end result.
Now that RIM and others are proving that the Android OS itself is relatively unimportant in running Android apps, with the Playbook tablet shown running Android 2.3 apps, it'll be interesting to see if Google make their own marketplace available on these other platforms. If they don't, the run the risk of Amazon or someone else capturing the grey-market of Android apps not running on Android.
March 30, 2011 at 10:31 am
I suspect that for the commercially driver side of Google the important pieces of "Android" as it's viewed by consumers are the Google Marketplace, and the various Google apps, none of which have ever been open.
The OS itself has been vital in getting these various components pre-loaded onto a wide variety of devices – saturating the market, and reducing the risk of Apple controlling the market and taking users way from Google's services, but it's a means to an end, not the end result.
With the new Playbook from RIM shown running Android apps without change, and I'm sure other suppliers not far behind with their own Dalvik environments, it'll be interesting to see if Google makes their Marketplace available on non-Android devices – otherwise they'll be giving up this grey market of Android apps on non-Android OS devices to their competitors like Amazon and GetJar.
Restoration of Pace | Techrights says:
March 30, 2011 at 10:39 am
[…] brands. There are already many brands in the Linux world, including WebOS and Android (which are decreasingly participatory, much less than Ubuntu for example). They have their own plans and their own promotion methods, so […]
Openness is Irrelevant to Android’s Growth « rand($thoughts); says:
April 1, 2011 at 5:16 am
[…] concerns RedMonk’s Stephen O’Grady tackles the question of whether it matters if Android is open or not. O’Grady writes: But while developers are unquestionably and understandably disappointed, […]
Boot up: Twitter’s Biz Stone on the $500m Facebook offer that never was, and more | Trade Bloggers says:
April 3, 2011 at 6:34 pm
[…] Does it Matter if Android is Open? >> tecosystems […]
April 6, 2011 at 4:49 am
I wrote this post http://meedabyte.wordpress.com/2008/12/17/democra… something like more than two years ago.
The point is that openness and democracy aren't key for quality and success in the mobile world. The important thing is understanding the market and the value chain.
Android is extremely well positioned from this point of view >> http://meedabyte.wordpress.com/2010/12/29/how-and…
Why open code is irrelevant to Android’s success « Android based Tracking System for Moving Objects says:
April 8, 2011 at 1:29 pm
[…] of openness RedMonk’s Stephen O’Grady tackles the question of whether it matters if Android is open or not: But while developers are unquestionably and understandably disappointed, there is little evidence […]
Go Droid – Google Buying Motorola says:
August 21, 2011 at 9:14 pm
[…] as possible People are worrying about Android closing up. It’s already up to some debate about how open Android is. There’s a larger debate to be had about how much “open” a company really needs now-a-days: […]