Project Astoria and Android app compatibility dead in the water?

Published by at

There's an interesting piece over at WindowsCentral overnight, talking about the possible shelving of Project Astoria, the Android emulation layer announced and demoed for Windows 10 Mobile back at the start of 2015. In contrast to Project Islandwood, the developer bridge to get iOS applications compiled for Windows 10, which was a very different beast, it seems that Project Astoria was too complicated, too confusing and too costly. I can't say I'm altogether surprised.

From the WC article:

Astoria is an Android emulator meaning devs do not have any real work to do. Instead, the phone ran the Android APK file directly. Early Insider builds of Windows 10 Mobile had this layer in the OS letting consumers sideload APK files directly. Not only did this project make app piracy easy, it undercut Windows developers who would have little reason to make native Windows apps anymore.

Astoria was never released to developers openly. Instead, they had to apply to test out the tools, which were still under development. Developer feedback and their experiences would be gathered in a closed forum found at bridgeforandroid.windows.com. At the time of this report, the Astoria landing page to apply was still active.

Project Astoria was a brute-force solution to a persistent problem for Windows and Windows Phone. However, it may have been too brazen even for Microsoft. Windows Central is now hearing from multiple sources that Project Astoria is on hold indefinitely, and maybe even shelved completely. Although Microsoft is not publicly — even privately — stating Astoria is cancelled, they are not openly talking about it anymore, or even privately discussing it with developers.

...Additional evidence supporting this conclusion comes from various sources besides the ones we spoke to, including:

  • The Project Astoria forums have gone silent since September with developer questions unanswered by Microsoft, including inquiries about the project's future
  • Recent Windows 10 Mobile Insider builds have had the Android subsystem completely removed including build 10586 (commercial shipping release)
  • Microsoft is no longer openly talking about the project even to those under NDA

There could be a few reasons as to why Astoria has been unsuccessful. Some of the people we have spoken too did not know the exact motive for the delay, although they did have some ideas, including:

  • Pushback – Developers were very unhappy about Astoria
  • Technical – There are reports that the Android subsystem caused Windows 10 Mobile to slow down over time
  • Legal — It is not clear what, if any, legal ramifications there in such an approach

Update: One other reason we have heard since publishing is that the Astoria team was 60-80 people versus the 5 it took for Islandwood. In the end, it may have been economic hurdles as much as technical ones.

All of this rings true, Astoria was always going to be a kludge of gargantuan proportions, ruining the elegance of a 'pure' Windows 10 running true universal applications. For a parallel, look at Blackberry 10, whose later builds included similar Android .apk (application) support, again in a very inelegant fashion. Today, Blackberry 10 is all but dead and the company has started making dedicated Android smartphones.

We're now seeing developers produce stunningly good universal applications for Windows 10 on desktop, tablet and mobile (e.g. Box, just last week), plus the Windows 10 Mobile app canon is at least well stocked, if not completely filled out, so the pressing need for shoehorning in Android binaries does seem to now be quite a bit less urgent.

Universal apps

Source / Credit: Windows Central