From WikiContent

Jump to: navigation, search

Compatibility is a better term for Fragmentation

Write Once Port Everywhere

How many compatibility axes are important to developers

Comes down to the way Google innovates, different than most others. Software is never finished, hard to maintain installed bases.

If we drive too much compatibility we are going to end up with three vendors making three devices for three carriers

Huge problem with varied chip providers and handset makers, hard to get uniform experience

Carriers further fragment (ie sideloading, Verizon limits Wifi, etc)

OS capabilities (1.0-2.3)

Google showcase phones have inside tracks

Unsupported/non upgraded phones

Developing for different Profiles Architecture/OEM for non showcase is more difficult Neither good or bad, but the way Google builds its ecosystem

Creates a splash, launching a new OS with a new Showcase device

Dave: 1. Carriers limit the os

 att - atrix cant sideload

2. Hardware limitations

3. Software release issue

4. profiles/platforms fragmentation

 phone, tablet, tvs


 google wont release source code till the new versin is on a showcase phone
 then takes 3-6 months to port, and 3-6 months to test
 thats why it takes so long to port a new versin to old phones

Lots of 2.1 devices, 25% of phones per the marketplace 4% have lower than 2.1

To get in android market pass cts

 sign licensing agreement
 And a document

Dave is being used as the canary to test for low level hardware testing for google

should it be compatability, not fragmentation.... YES

Soon we will get a a point where hardware will be so good, that OS's won't need a showcase device

Another bit of fragmentation people [carriers] fork OS releases [nature of opensource] History Repeating Itself!

Go into markets uncontrolled

CTS Compatibility Spec/Requirements License agreement

Personal tools