Your browser (Internet Explorer 7 or lower) is out of date. It has known security flaws and may not display all features of this and other websites. Learn how to update your browser.

X

Navigate / search

The darkside of Android fragmentation – updates!

In general terms, fragmentation of Android is a good thing. You can visit your local phone shop and choose your perfect phone from a huge range of shapes, colours and prices – all running a powerful operating system with a huge number of apps available. Manufacturers can produce phones and devices with specialised hardware, like the outdoor proof Defy + (mine recently survived a washing machine cycle), and specialised software, like the Amazon Kindle Fire, but still utilising the advantages of a common operating system and app ecosystem. Even manufacturers producing their own skins is, in principle, potentially a good thing.

However in reality there is a darkside to fragmentation. Just before Christmas my Motorola Xoom was updated to Android 3.2, released in July. To be fair the update was available elsewhere earlier, but my wifi only Xoom, with minimal obvious manufacturer customisations, was updated 6 months after a release. Why did it take so long?

Motorola have recently released information on their plans to update selected devices to Android Ice Cream – the keys stages are:

1. Merge and adapt new release for device hardware (including update to custom software)
2. Stabalise and ‘bake’ the result to iron out bugs
3. Submit upgrade to carriers for certification

Obviously the manufacturer needs to make sure the updated software works with a device hardware. What is not clear however is how long is spent customising the software. Personally I choose a phone manufacturer based on the hardware – it’s specification, reliability and manufacturer reputation – not on their custom software. Whilst HTC Sense adds some value, Motorola Blur is at best an annoyance. Each manufacturer also seem to have their own clock and alarm, calendar, email client and texting interface – most of which are inferior to alternatives in Android Market, yet cannot be completely replaced. Each software update requires each of these, unused in my case, features to be updated before a release.

A slightly worrying trend is that manufacturers are starting to include bloatware, usually the domain of network providers, in updates. In the latest update for my Xoom (wifi version, so no carrier) was included ‘Citrix Receiver’ – probably quite useful for business use, but useless for a freelancer. This app cannot be uninstalled and is currently nagging me to grant it the following permissions: “Your location (fine)” and “Services that cost you money”. Errrr no I don’t think so!!

It is also perfectly acceptable that carriers need to certify an update – after all a phone needs to make calls! Motorola give an estimate of upto 2 months preparation to enter a carrier lab cycle, which then lasts 1-3 months – upto 5 months in total!! Perhaps I am thinking too simplisticly, but does it really need to take that long? I am pretty sure iPhone updates are not delayed that long for carrier certification. Perhaps, as with manufacturers, some of this time is spent adding custom branding and bloatware the consumer does not want. I choose a carrier based on what their network infrastructure is like and their price plans, not on their software!

There is of course an issue with consumer expectation – if you bought a phone with Android 2.1, should you really expect it to be upgraded to Ice Cream Sandwich? How many updates should you get over a two year contract for your budget Android phone? Phone development is moving at a fast rate, and phones that were top of the range a year and a half ago can barely run (if at all) some of the latest software and apps, and I imagine the update cycle is not cheap. Personally, beyond bug fixes and security patches (which you should get for the life of your contract), I expect one Android update for my phone (although I have a higher expectations for the Xoom), but manufacturers need to be open and upfront on their intentions for delivering (or not) updates.

Fragmentation should be a good thing, but as it stands there is a darkside littered with issues hindering Android.

Manufacturers – concentrate on your hardware… only add custom software if it adds value to the consumer and you can deliver speedy updates. Be open and honest to consumers on planned (and budgeted) updates.
Carriers – concentrate on your network… all I want from you is a good signal and half decent mobile internet. Streamline your update certification (not having custom software and bloatware will probably help)
Consumers – accept that not all phones will (or should) get every update.

Comments

Clayton McKenzie
Reply

I don’t think your comments re the iPhone are entirely fair to Android as an OS Platform or the multitudinous manufacturers that use Android as an OS for their hardware. You’re quite correct, it does not take Apple 5 months to carrier certify an iOS update… But you do have to remember that Apple have a comparitively tiny number of KNOWN different hardware types and configurations to deal with… Whereas an Android update coded to ‘generic standard hardware’ gets provided to a Manufacturer which then has to see if any of it’s existing ‘feet’ will fit into the newly styled ‘glass slippers’ and then whether the princess can still dance in them before it can be passed on to carriers to test their own custom modifications if they have them. The Manufacturer will have to go through this process with each model of Android device it has in order to figure out which it can ‘certify’ with the new update. You’ve effectively got 3 completely seperate companies involved in the delivery process for the development of one update to a user instead of Apple’s 1 and as you know, getting useful infomation to pass successfully between companies is a ‘mission of nightmare proportions’…

Blackberry RIM are in the same boat as Apple. There will be no direct mobile handset OS to compare Android to until (unless) WinMo7.5+ suddenly becomes popular…

Everything else (including the conclusion) you’re bang on with as usual though..

Leave a comment

name

email (not published)

website

CommentLuv badge