Commentary: On Android Fragmentation

Question: Is Android fragmented?

Short Answer: Yeah.

Long Answer: Yeah, kinda, but who cares?  Shut up.

Longer Answer: Well, let's first talk about what fragmentation means and why Android is fragmented.

Operating system fragmentation is the phenomenon when multiple versions of an operating system are available to consumers at the same time and the differences between OS versions cause difficulties for developers.  For example, Rovio Mobile, the makers of the popular app "Angry Birds" recently blogged about their difficulty making the game work smoothly on older versions of Android.  The game works fine on phones with better hardware and the latest versions of Android.  However, if you are the unlucky owner of, say, a T-Mobile myTouch 3G, then your phone officially does not support Angry Birds (for now).

There are several versions of Android available—1.5, 1.6, 2.1, and, the current version, 2.2—and each OS is often customized with manufacturer's special user interfaces, such as Motorola's Motoblur and Samsung's TouchWiz.  Additionally, there are hundreds of different Android phones running these different, possibly customized OSes, each with different hardware specs.  Granted not all phones are currently for sale; some have been discontinued.  But not every consumer can afford to upgrade every time a new Android phone is released and most consumers are locked into a two-year contract anyway.

It seems like the simple solution would be for carriers and manufacturers to allow users to download the latest version of Android on their Android phones, but they don't always allow this.  One reason is that manufacturers need time to update their custom UI to work with the new version of the OS.  I don't know why they wouldn't just let users forgo the custom UI and download the uncustomized ("stock") Android OS until the custom UI is updated.  Another reason might be that the new OS requires too many resources for old phones with old hardware to handle.  I honestly don't know the answer and whether the manufacturers or the carriers are more to blame.

Other smartphones (i.e. the iPhone) don't have to worry about this.  There are only four iPhones in existence (original, 3G, 3GS, and 4) and the software is tightly controlled by Apple (and okayed by AT&T).  Apple has made sure that the last three phones can run iOS 4, the latest version of their OS, and even if it's a bit slow on the 3G, Apple's getting better results than Google (case in point: several users state that Angry Birds works fine on the original iPhone).  Apple only has one carrier to consult about visual voicemail and whether FaceTime should be allowed to run over 3G.  Hardware and software decisions about the iPhone are simplified by the lack of diversity.

This is why Android is fragmented.  Android is a free and open source OS and any phone manufacturer can put any version of Android on whatever phone they want, and the carrier can choose to let users update it or not.  That's why you can still buy a Motorola Devour from Verizon today with Android 1.6 on it (with Motoblur), which Motorola refuses to update 2.2 or even 2.1.

Because of this fragmentation, developers who write apps for Android either have to test their software on a plethora of different custom Android builds using a bunch of different hardware, or they can take the easy route and test on a few phones running 2.1 and 2.2.  Most developers seem to do the latter, leading to buggy apps on older hardware and older Android versions.

So, yes, Android is definitely fragmented, but is it Google's fault?  I'd say no.  It's manufacturers who release crappy hardware and put custom UIs on Android, preventing simple updates.  It's carriers (I think) who act as the gatekeepers to OS updates, preventing and allowing updates when it suits them.  What can Google do about it?

I suppose Google could better coordinate with manufacturers about minimum hardware specs for each OS, but this wouldn't necessarily prevent Android phone makers from releasing cheap phones.  They could prevent access to Android Market (the premier place to find Android apps) for older versions of Android or overly customized OSes, but locking down the platform goes against the spirit of free, open source Android.  I don't think there's much Google can do about Android fragmentation.

Additionally, we have to take into consideration that Android is a newer, faster moving OS.  It came over a year after the first version of iPhone's OS and is arguably being updated faster as well.  Andy Rubin, the "father" of Android, made the point:

"Older Android devices that can't be upgraded to newer versions of the OS or run newer apps are no different than an iPhone from 2007 not being updated to OS 4. It's not fragmentation -- it's legacy. If so, legacy systems are now aging faster than ever before, due to a rate of innovation that has never been seen before in history."

- Michael Gartenberg, "Is Android fragmented or is this the new rate of innovation?"

Because it's such a new, rapidly changing OS, some hiccups are expected.  Android will eventually mature and hit its stride, reaching a point where OS updates are less significant and farther apart.  App compatibility won't be much of an issue then.

Furthermore, one can make the case that, at least software-wise, fragmentation is not an issue.  As of November 1st, 77% of Android devices are running 2.1 or 2.2, the most current versions of the OS.

Lastly, since the main culprits in fragmenting Android are the manufacturer and the carrier, consumers won't have any problems if they simply buy an Android phone from a good manufacturer-carrier combo.  If you're buying a $50 Android phone or any Android phone from AT&T (who are infamous for locking down Android phones), then don't be surprised when you can't run the latest and greatest Android app.

So, yeah, Android phones are kinda fragmented, but who cares? Shut up.

Dadaist Answer: The beans left yodel for what it's worth when.


Update: A post on why iOS is fragmented, too, and why some fragmentation is expected; by Matt Maroon, a real, live app developer.

Policy: Warrantless GPS Tracking

On August 25, 2010, The U.S. Court of Appeals for the Ninth Circuit agreed that the DEA did not need a warrant to attach a GPS tracking device to the underside of a Jeep belonging to Juan Pineda-Moreno, an Oregon man suspected of growing marijuana.  The device was attached to the Jeep while it was parked in his driveway, a few feet from his trailer home.  Pineda-Moreno challenged the DEA's actions, arguing that the DEA would need a warrant to invade his privacy.  A three-judge panel argued that Pineda-Moreno's driveway was not private because it was open to strangers, such as delivery people and neighborhood children, who could wander across it uninvited.  And once a GPS device has been planted, the government is free to use it to track people without getting a warrant. (http://www.time.com/time/nation/article/0,8599,2013150,00.html)

Should it be legal to track a vehicle by GPS without obtaining a search warrant?  As an engineer, I feel the need to weigh in on the legality of this situation revolving around possible abuse of new technology.  The wisdom of the federal government's war on drugs is another interesting topic, but I'll leave that for other blogs.

The pro side is simple: using a GPS tracking device to monitor a suspect is no different from the perfectly legal act of waiting outside his home in an unmarked car and following him when he drives around town.  In fact, it's safer because police don't have to drive erratically to follow the suspect in a vehicle.   Police officers have a long history of attaching similar tracking devices to cars, such as radio transmitters ("beepers") that give off an audio signal when officers are close to the transmitter.

The con side is more complex because it requires one to be familiar with the technology, its limits, and its possible misuses.   Many articles have been written about this already and make several good points, including:

• "For starters, the invasion of Pineda-Moreno's driveway was wrong.  The courts have long held that people have a reasonable expectation of privacy in their homes and in the 'curtilage,' a fancy legal term for the area around the home. The government's intrusion on property just a few feet away was clearly in this zone of privacy." And by saying that his driveway was not private because people could wander onto it uninvited, they are allowing police to discriminate against poor and middle-class people who cannot afford to have electric gates, fences and security booths protecting the privacy of their homes.
  
  - Adam Cohen, Lawyer, Former Time Writer
•  Tracking a person's movements with GPS devices is way more invasive than the act of trailing a suspect in public:
  

  "We hold the whole of a person's movements over the course of a month is not actually exposed to the public because the likelihood a stranger would observe all those movements is not just remote, it is essentially nil. It is one thing for a passerby to observe or even to follow someone during a single journey as he goes to the market or returns home from work. It is another thing entirely for that stranger to pick up the scent again the next day and the day after that, week in and week out, dogging his prey until he has identified all the places, people, amusements, and chores that make up that person's hitherto private routine." 

  - Judge Douglas Ginsburg, U.S. Court of Appeals for the District of Columbia Circuit, on a similar warrantless GPS tracking device case
• A GPS tracking device is more invasive than other, legal devices used to track suspects without a warrant (like a radio transmitter):
  

  "Prior Supreme Court cases that allowed the warrantless use of tracking devices were based on the technological limitations of the devices available to at the time. In particular, the devices could only be used to supplement and aid traditional visual surveillance, and were unable to record data on a vehicles movement without human intervention. [….]
  
  "The protections provided by the Fourth Amendment, as the Supreme Court has often recognized, must change to meet new technology. Especially where the cases involve sustained and long-term surveillance of a targeted individual unrelated to any particular criminal action, no reasonable person would expect to be the target of such a massive police surveillance operation. Accordingly, because the use of these devices infringes on a legitimate expectation of privacy, the use of these devices constitutes a search which, absent the present of another exception, requires a warrant."

  - Joshua Engel, Lawyer, Legal Blogger; on warrantless GPS tracking

One point I haven't seen made is that even if the device is attached in a public place and it is legal to track a vehicle's movements by GPS, at some point the vehicle may enter a private residence.  If the police do not remotely shut off or destroy the device at that point, they are monitoring a person's movements on private property.   I think any judge would agree that this would constitute an illegal search.

The GPS device could be turned off remotely if the officers tracking the vehicle saw the vehicle enter a private residence, but it seems unlikely that officers would continuously follow the vehicle around, given that the entire point of the device is that police don't have to follow the suspect in person.

Another way to prevent the device from monitoring private property would be to program the device to automatically stop tracking the vehicle when entering a private residence.   This would require the GPS device to have a map of all private and public areas in a given region; the GPS device would have to differentiate from an (evidently) public driveway and the privacy of a homeowner's garage.  Although online maps are pretty comprehensive, no map like this currently exists (that the public knows of) and even if the government put the effort and money into making such a map, it would be quickly outdated when a resident decides to remodel his property.  In fact, such a map would surely be outdated before it could be completed.

At this point in time, it is impossible to prevent an illegal search once a GPS tracking device is attached to a vehicle (or person).  In fact, if law enforcement officers continue to attach these devices to cars, an illegal search is inevitable.  Because of this and the other reasons listed above, I would conclude that attaching GPS tracking device to a vehicle, person, or his personal belongings without a warrant constitutes a breach in the Fourth Amendment of the Constitution.

Wish List: Tablet PC

Multi-touch tablet PCs can be many things for many people.  For some, it's a netbook replacement, for others, it's almost a desktop replacement.  Some consumers just want an e‑reader that can also browse the web on while on the subway.  Some have no use for any tablet.

For new devices like multi-touch tablets, it's difficult for manufacturers to pinpoint what consumers really want, which explains the recent series of craptablets being released.  In this article, I will address features I personally want in a tablet, but I’ll also address features I feel I don't need.  Tech wish lists are pointless if you say you want a device that is has super great features as well as an impossibly low price.

Even if you don't want the exact tablet I shall describe, I think you'll agree it fits a niche that current tablets are progressing towards, but have yet to fill.

List:

Display: Super AMOLED, IPS LCD, Pixel Qi?
Multi-Touch: Capacitive
Form Factor: 10" x 8" x .5"
Processor: 1 GHz
OS: Future version of Android (Flash, small, real multitasking) or some Internet browser OS (e.g. Chrome OS, joojoo OS)
Storage: Small OS: 8GB, SD or MicroSD slot, Internet browser OS: Whatever
Battery:  5+ hours
Camera: Front VGA
Ports: USB, headphone jack
Buttons: Home, back, mute, volume, on/off/lock
Antenna: Wifi, Bluetooth, no 3G/4G
Other: Speakers
Price: Small OS: $500, Internet browser OS: $200

Further Description:

What I really want: What I would like is a web-browsing device that I can leave on my coffee table.  Something to look up YouTube videos on and browse IMDb while watching TV.  I'd occasionally browse the web or play mini-games on it on long flights.  In short, I want a big version of my smartphone.  I think this is what many people want out of a tablet, especially those who want a large (9"-13" screen) tablet.  I think the closer a large tablet comes to meeting these specs, the more that tablet will succeed.

Display: A Super AMOLED (Samsung Captivate) or IPS LCD (iPhone/iPad) screen would be a plus I'd pay extra for; both are very bright and colorful.  High resolution is also nice. But neither of these screen desires are a dealbreaker.  A regular LCD would do me just fine.  This won't be my main computer; it's just a tablet for casual use.  I'm not expecting a bright e-reader screen, and I don't need a 1,000,000:1 contrast ratio for the movie clips I'll be watching.  A Pixel Qi screen would be the best because its e‑Ink‑esque low-power mode would make the tablet a great e-reader in case I do end up reading books on it, but, again, it won't make or break the tablet.

Multi-Touch: No brainer here.  The reason for this tablet revolution is the easy-to-use, responsive capacitive multi-touch screen.  Resistive just won't cut it anymore.

Form Factor: Here is where I may differ from many tablet enthusiasts.  I already have a smartphone and it has a 3.5 inch screen.  I don't need another phone-sized screen.  Don't get me wrong, I've visited many a website and am currently reading 1984 on my tiny phone screen, but my entire reason for getting a tablet would be the large screen that makes viewing websites, e-texts and short videos more pleasant.  That's why I want something roughly the size of a sheet of paper.  A 8"x10" screen would be a beautiful 12.8" in diameter, slightly bigger than the 12.1" joojoo screen and quite a bit larger than the 9.7" iPad screen.  Pocketable?  No.  Portable?  Hell yes, as long as the weight was a three-ish pounds or less.  Something smaller would be fine as long as it's over, say, 9" diagonal.  Any screen under 7" is definitely useless to me.

Processor: 1 GHz seems to be just fine.  HTC Evo's Snapdragon, Samsung Captivate's Hummingbird and the iPhone 4's/iPad's A4 all seem like sufficiently speedy 1 GHz chips. Perhaps there's some minute difference in battery life, but I don't care.

OS: Here's the thing: I'm tired of iOS.  I'm tired of their App Store policies and I'm tired of them locking down the platform. I'm intrigued by Android and writing apps for free in Java on my PC rather than paying a developer's fee to use the ungodly abomination that is Objective-C (only on Macs).  That's why, although iOS is a good tablet OS right now, I want my tablet to have a future version of Android, when Android is tablet-ready.  Word on the street is that Android should be tablet-ready (support higher resolutions, etc.) either with Gingerbread (due by December) or Honeycomb (due early next year).

I also would accept a web-browser-based OS like the joojoo has, except not crap.  Google's Chrome OS should be released by December, which should be decent, given the excellence of Android.  However, it can be assumed that any tablet with a web-browser-centered operating system would do little other than browse the web, so this reduction in functionality would mean that the tablet needs to be sold at a cheaper price.  Tablet manufacturers should feel free to cripple the hardware (processor, RAM, graphics, storage, etc.) of a web tablet as long as they get that price down.

Alternatively, the extra functionality of Windows would be nice on a tablet, but Windows 7 just isn't tablet ready.  The next version of Windows might be multi-touch focused, but Microsoft never releases two good operating systems in a row.  Take my word for it: the first good MS OS that works well with multi-touch won't come out for another 10 years.  There will only be decent tablet PC skins over Windows.

Storage: This depends on the OS. Web-based operating systems don't require the storage of a tablet OS.  If my tablet has a tablet OS, I'll want to fill my tablet with music and movies.  If not, it needs just enough memory to store my web cookies and to stream video.  So for a tablet OS, I'd like 16 GB of storage, or at least a SDHC slot, so I could add 16 GB or 32 GB of storage.

Battery: This is the other big one.  Less than 4 hours seems to be the Android standard, yet the iPad promises around 10 hours of life. When Gizmodo used the iPad heavily, they still got almost 6 hours.  I think any phone or tablet should last at least the duration of a long flight, around 5 hours.  I don't think this is too much to ask, especially when you consider that the battery life will eventually decline.

Camera: A tablet is too large to use as a regular camera, at least any tablet large enough for me.  Holding a 8"x10" device and taking pictures with it would look and feel ridiculous.  So I'm not asking for a for a camera on the back (although, I'll gladly take one if it adds little to the cost).  I would, however, like a camera on the front for video chatting.  Just chatting doesn't require high definition video, so a VGA camera will do just fine.

Ports: Headphones, I think, are pretty standard.  Only USB is something that Apple sadly forgot to include in the iPad.  Sure, you can buy a $30 camera connection kit to transfer photos by USB (and SD card), but I want the ability to hook up cameras, printers, webcams, external hard drives, phones, and other peripherals, and I want it build into the device.  Sure, nothing will work with the tablet until peripheral manufacturers create special drivers for Android or whatever OS the tablet uses.  Eventually they will create the drivers and then the tablet will be infinitely more useful.  If it's just a web tablet, USB isn't a necessity, but it'd still be nice to have the promise of someday being able to print web pages from my tablet.

Buttons: Apple likes to make simple, easy-to-use devices.  Someone must've told them at some point in time that less buttons == simpler.  Actually, it's almost the opposite.  I certainly understand not wanting to connect a full keyboard to your device so that users won't be searching for the Scroll Lock button, but it really limits the functionality of the device when you have only one physical button (not including volume buttons, mute and lock).  Wouldn't it be easier to press one button to bring up the task list/iPod controls and one button to bring up voice commands instead of double-clicking and triple-clicking?

Android phones have it right with their four buttons (Home, Back, Menu, Search).  However, I'd settle for just a "Home" and "Back" in addition to the mute, volume and lock buttons.

Antenna: Wi-fi and bluetooth are standard, so nothing to talk about there.  But here's another item where I may differ from the crowd: I don't need 3G/4G.  This device will mostly be used at home and places where there is wi-fi (coffee shops, airports).  Sometime in the near future I will have a phone with free tethering and I won't need any other satellite data devices other than my phone.  Satellite antennas always increase the price of the device in addition to the monthly price plan required to get them to work.  I already pay $25 for mobile Internet and ?? dollars for my home Internet.  I don't need another monthly Internet bill.  I probably differ from most people on this, but as long as tablet makers release a wi-fi only model, I will be happy.

Other: Speakers.  For showing people funny YouTube videos.  Hide yo' kids, hide yo' wife!

Price: The final key.  $500 is a nice round price and the cost of a wi-fi only iPad with a nice 16 gigabytes of storage.  I don't think it's absurd for me to ask that the next generation of tablets to cost this much and have the additional features I've specified above.  By April 2011, the iPad will have been out a year and the decent-looking Galaxy Tab will have been out for 6 months.  I'm hoping tablet manufacturers can accomplish this by sometime around then.

If not, a web-only tablet is something manufacturers could definitely throw together by then, since it doesn't need the same awesome hardware and software as a regular tablet.  But it better only cost $200 or so.

Other tablet enthusiasts' concerns:

Wait, what about GPS/a-GPS? I have a phone for this.
What about HDMI? It has a large screen, it has speakers.  I don't need it to hook up to my TV, although that'd be awesome and I'd pay a little more for it.
You forgot to say it has to be a 16:9 ratio. I don't care about black bars on the screen as long as the display is big enough to show everything.  With a 12"-13" display, I'll be able to see everything just fine.  Plus, there still is a lot of non-16:9 video out there, like most YouTube vids, which would likely compose 90% of the video I'd watch on this.
Fast boot up?  Sure, but most phones' boot up times are okay to me.  This thing would be asleep most of the time—not off—so it's not super important.
Apps?  If the OS isn't Android or iOS, shouldn't you mention a requirement for apps and an app store?  I have a phone for apps.  I want a web browser; and for a full tablet OS, I also want a media player.  The web browser will have Adobe Flash, so I'll have plenty of "apps" on the interwebs.


Did I leave anything out? Hit me up in the comments.