Microsoft Rumored To Integrate Android Apps

phmadore writes "Windows Phone has been struggling for market share, largely due to a serious lack of developers willing to invest their time in what one might consider a niche market. Statistically speaking, Android has more than 1.1M apps to Windows Phone's 200,000+. Well, according to unnamed sources informing the Verge, Microsoft may soon integrate/allow Android applications into both Windows and Windows Phone." This follows the recent debate over whether Microsoft should try to fork Android. Peter Bright made the point that doing so would be extremely difficult, and probably not worth Microsoft's time. Ben Thompson has an insightful post about how Microsoft's real decision is whether to focus on devices or services.

Numbers aren't the story

[jbmartin6]

So MS has 20+% of the apps that Android has, that doesn't sound horrible. How many of those Android apps are garbage? The numbers aren't the whole story, if the 200k are much better quality than most of the 1.1M the Windows phone would win. I am not saying that is the case, just saying that comparing the number of apps in a store isn't useful information.

right,

[roc97007]

...because, this strategy worked so very well for Blackberry.

That said, I can't think of a reason why Microsoft should not integrate Android applications, provided the results gives some reasonable user experience. I suspect that "supporting" Android applications where the user has to put up with significant numbers of crashes and hangs, rendering errors, screen geometry issues and so forth would actually hurt the platform further.

Re:right,

[fuzzyfuzzyfungus]

The trick with 'Android compatibility' is that it's really two different problems. One is merely engineering ('merely' in the 'may actually be quite difficult; but there are engineers that are quite smart, trying giving them money' sense) and one is strategic:

'Android' as in the ASOP is a mixture of GPL and Apache. Exactly how many man-hours it takes to get ASOP running on your platform, or Dalvik and friends running on your non-Linux kernel is an open question, and may end up being quite a few if you want it to work well; but there is nobody to stop you, and you just need suitably skilled software people.

Trouble is, much of the good stuff in 'Android' (and stuff that Google doesn't exactly discourage developers from using) isn't ASOP, it's Google Play Services, a set of proprietary applications, libraries, and Google-backed web servcies that can be bestowed or denied to your device at the power and mere pleasure of Team Mountain View. They tend to ignore indie ROM-cookers and two-bit pacific RIM clonemongers who quietly pirate GPS; but if a company large enough to target, or ambitious enough to try to cut deals with major carriers in markets Google cares about, tries to distribute GPS without Google's blessing, it's world-of-hurt time.

At a greater or lesser cost in software engineers, you could get an ASOP-compatible Android compatibility layer running on QNX, NT, OSX, whatever. However, how much that helps you is increasingly limited.

It worked well for OS/2

[jader3rd]

When OS/2 was struggling for market share, IBM decided that they could bring along more customers by allowing Windows programs to run on OS/2. So they put a whole lot of effort into it and the result was a disaster. The few programs that used to have an OS/2 version no longer did. The program maker didn't see a reason to make an OS/2 version if their Windows version ran on OS/2 too. And customers saw that Windows programs ran better on Windows than on OS/2, so why buy an OS/2 machine if all of the programs you want to run, run better on this cheaper Windows machine?