Search This Blog

Friday, October 29, 2010

The Real Fragmentation That Threatens Android

Fragmentation is a real threat to the Android ecosystem and user experience.  But not the way you're probably thinking.  The word "fragmentation" is probably one of the most (ab)used buzzwords that gets flung around when people start talking about developing for Android.  And also one of the most misunderstood.  (According to Ray Walters of the AndroidGuys blog, many of the C-level executives who spoke at AppNation don't even get it.)

The term "fragmentation" is used to suggest one of two "ills" that supposedly "plague" the Android ecosystem:
  • Phone manufacturers who employ Android are forking off completely new "flavors" of Android from Google's root code base.  (Much like the Linux environment currently looks.)
  • Application vendors are forced to create a new version of their applications for each version of Android they support (or for each hardware vendor's version of Android)
The truth is that none of those things are really happening, and the reality of the impact of multiple Android versions is far different from this portrayal (unless, of course, a given developer is unaware of how to properly deal with multiple Android versions).  There is, however, a fragmentation that is potentially a very real problem for the Android developer:  App stores.

In the past month, I've read accounts of several new app stores that are already in existence or are in the works:
  • Verizon's VCast Android Market
  • Amazon's up-and-coming Android Marketplace
  • Archos AppsLib
  • SlideMe
  • AndAppStore
  • Adroia
  • GetJar
While competition is generally a positive thing, the multitude of disparate app stores is actually a hindrance to Android application distribution, which is, in turn, a hindrance to Android adoption, at some level.  Here are just a few reasons I believe this is true:
  • Each app store can have its own terms of service, by which each developer must abide (and to which each application submitted to the app store must adhere).  One app store, in particular (Amazon's), actually reserves the right to modify your application in order to support the addition of their proprietary DRM.
  • Deploying application updates could be a pain (should a developer decide to submit their application to multiple app stores).
  • A different set of applications is available in every one of these app stores (potentially with different versions of the same application).  And that's assuming that the app store doesn't require that you distribute your application exclusively through their app store.
  • Can multiple app stores be used on the same device?  (Granted, this is very likely more a function of the specific Android device, which may be locked to a single, specific app store.)  If any device is restricted to the use of a single app store, that reduces consumer choice for the users of that device.
  • There is a potential for a disconnect between the app stores that will be preferred by consumers and developers:  
    • Developers will end up choosing to use the app store(s) with the most favorable terms of service.  (The store that takes the lowest percentage of proceeds, and/or makes uploading new application versions the easiest.)  
    • Consumers will choose the app store(s) that they are forced to use, or that have the best applications
  • Didn't we learn about the evils of onerous "approval processes" that are employed by other app stores (lest you figure out that I'm talking about Fruit, Inc.)?  Why would developers choose to partner with an Android app store that has just as ludicrous an approval process?
Overall, this is one area where the fragmentation that's developing is not a good thing.  It doesn't have to be a negative thing, though.  If, for instance, there were a single "gateway" where developers could register and upload their applications for distribution by the individual app stores (like the Android Marketplace), that could simplify the existence of multiple app stores for application developers.  This would be one small thing that Google could do to decrease the impact of the fragmentation that appears to be developing.

What do you think?  Am I onto something, or are all of these app stores the greatest thing since slided bread?  Let me know what you think!

Tuesday, July 27, 2010

Why Do Manufacturers Customize Android?

Those of us who follow Android blogs routinely are certainly (by now) well-aware of the rumors that Google's highly-anticipated 3.0 (aka "Gingerbread") release of Android will focus on revamping the Android User Interface.  Just like the article found here, many in the blogosphere are speculating that this is an attempt on Google's part to nudge handset manufacturers away from slapping a lot of customization onto Android .  There is a certain sect of Android user [that includes me] who only want their phone to get the latest Android release two weeks before it's made public.  As such a user, I would hope that Google could succeed in eliminating the "frosting" of the OS with all these custom UIs, widgets, and other tweaks, as the customization only makes me wait longer for my next serving of dessert.  But, I'm here to tell you why that will not happen.  Ever.  Why?

Well, you have to look at why they customize the OS in the first place.  And it's not because "stock" Android is ugly.  It all boils down to one word:  distinctiveness.  Consumers want the "best" (read "most distinctive") devices, with the most flash, flare, and eye candy.  Mobile phone vendors can really only add distinctiveness through more/better hardware (like an amazing display, faster processor, front-facing camera, a camera with a flash, or a mood rock skin), or though the software that ships with the device.  Given that all manufacturers have access to the same OS (Android) from the same vendor (Google) at the same price (free), this isn't a source of distinctiveness until it's modified.  Adding all of the best hardware to a single phone may produce a killer device, but it's price may be considered grounds for murder.  On the other hand, sprinkling a custom UI on top of Android may turn mediocre hardware into something to die for at little or no cost to he manufacturer.

There are a few other (potentially minor) reasons that it makes business sense for the practice of "frosting" Android to continue:
  • Economies of scale.  Once a company like HTC has made the investment into Sense UI, the cost to add that to any phone in their lineup is minimal.
  • Adding value.  Once a critical mass of consumer deems that Sense UI (or Motoblur) is "cool", this becomes a source of value to the manufacturer.  They can charge much more for the phone by "adding value" to Android.
  • Reducing dependence on Android.  As good a deal as Android is to handset manufacturers, it only makes sense to preserve future revenue options.  The first time we saw Sense UI, it wasn't running on Android -- it was running on Windows Mobile.  And, if Android dies tomorrow (or Google decides to charge confiscatory licensing fees for Gingerbread), HTC could spend the time to port Sense to another OS (or write their own) to preserve their revenues.  
  • The UI is part of the experience that the phone manufacturers offer.  It (can) provide some consistency between different devices (even if those devices run different versions of Android, or different Operating Systems entirely).  
  • Sunk Costs.  HTC and Motorola already have significant investment in their customizations.  They're not going to walk away from those investments without milking it for all its worth.  

Overall, as much as I'd like to see these custom interfaces go by the wayside, I don't see it happening anytime soon.  The only thing that will get a significant number of devices (read "more than one") that has great hardware under the hood, and is updated with the latest Android release as soon as it's available is for consumers to band together and demand it.  In short, consumers need to demand another Nexus One.

Update:  If you agree that we need more Nexus One-like devices, there is a petition here where you can let device manufacturers know that there is a demand for Android devices with minimal customization.  Perhaps a show of numbers may tip the scales in our favor.

Wednesday, July 21, 2010

A Brief Introduction

I think a brief introduction to this little corner of cyberspace is in order, as this is the first of what I hope to be many posts.  First off, this isn't just another Android news site, or a collection of phone/application reviews.  There are enough of those sites already, and I wouldn't have the time to give any justice to a venture of that nature.  Rather, this blog is dedicated to my commentary and analysis of the Android platform and ecosystem as a whole.  Inasmuch as this involves the discussion of a specific device or application, then we'll go there... but as I said, the focus is more on where the Android platform has been, is, and seems (to me) to be going.  Or, in some cases, where I think it should be going.

I hope that, in the course of my ramblings, you read something that piques your interest, or makes you think.  I hope for this to be as much a forum for discussion as it is my soapbox.  And, more than that, I hope you enjoy it!