Android is fragmented, not just because Android devices have different screen sizes and resolutions, but also because they have different input methods as well (some are fully multi-touch, some have back/home button, some have QWERTY keyboard, some have stylus, ...) AND are usually a couple versions behind (just 1.2% of Android devices run 5-month-old 4.0, compared to about 60% of iOS devices that run 4-month-old 5.0[1]). I guess a huge portion of OS X users are on Lion or Snow Leopard (if I had to throw a number out of my hat I'd say 45% on SL and 35% on Lion; but the important thing is that every new Mac sold in the past 6 months came with Lion, not Android 3.2!).

And back to the screen size fragmentation point he's trying to make, I think it's really not as severe on desktop as it is on mobile devices - if you fine-tune a FULL SCREEN mobile app for a 4.1-inch display, it wouldn't fit well in a 4.6- or 3.5-inch display, and will look like garbage. But if you fine-tune an app for an 11" MacBook Air display, it's still reasonably tolerable on a 13" MBP.

[1] http://www.macrumors.com/2012/03/23/developer-sees-quick-ado...

I think the point he is trying to make -- that Android fragmentation is not that big of a deal -- is not supported by his sarcasm. You can, as a normal user, upgrade your laptop/iMac/Mac Mini to the latest supported version of OS X. For many Android users upgrading to ICS is simply not an option. (At least without rooting the phone, something most people are loathe to do.)

He's not even talking about OS version, just about the "hardware fragmentation" among devices that run OS X, specifically in regards to the varying screen sizes.

But at what cost ? I have seen horrors faced by people trying to upgrade their core2duo or even nehalem based laptops to the latest version of OSX.

Besides, the post was about hardware fragmentation more than software fragmentation.

> I have seen horrors faced by people trying to upgrade their core2duo or even nehalem based laptops to the latest version of OSX.

You'll pardon me if I meet that with some skepticism. Both my work-issued and personal laptops are Core 2 Duos, both run Lion just fine. The same is true for everyone else in my office: same model MBPs, same OS, no issues in upgrading.

> You'll pardon me if I meet that with some skepticism. Both my work-issued and personal laptops are Core 2 Duos, both run Lion just fine. The same is true for everyone else in my office: same model MBPs, same OS, no issues in upgrading.

What do you use the machines for ?

I know atleast three people who tried upgrading to lion, found their system to be laggy, went back and did a clean install.

Why are we missing the point here? When people talk about Android fragmentation, they usually mean different screen sizes, resolutions, and GPUs. All of which are also fragmented amongst machines running OSX.

I don't care what most people mean (if that is what they really mean!). They can be as clueless as the author of the G+ post!

I call Android fragmented, for many reasons; including different screen sizes, different input methods & (most importantly) not being able to upgrade to the latest version without your phone manufacturer's blessing (which in most cases doesn't happen, or happens 6 months later).

Personally I feel humor falls flat when it misrepresents the problem.

Even more so when it becomes inadvertently true. Those are perfectly legitimate reasons not to write software for OS X. Compared to writing for a much more controlled target, desktop software seems like a misery. Consistent design constraints are wonderful.

I wish I had the capacity to downvote this. Anyone who wants consistent design just because it is a pain to you should not be a programmer / engineer / hacker. Because if you are, any one of them, you whine, but then find solutions anyway and move on with your life.

-- I had to say it. I don't have much reputation to lose. So downvote away!

Lovely True Scotsman, there.

But okay then. I'll just quit my job because you said that.

Upvotes for being funny atleast.

I chuckled when the point dawned on me, albeit a bit later than it should have. But the comparison breaks down when you consider the different ways Desktop OSes and Mobile OSes handle the relationship between display resolution and application resolution.

That is to say: we have come to expect (or have been required to accept?) that mobile applications operate on a full-screen basis only. This is logical for tiny screens, but causes problems when the definition of 'space' differs in the phrase "Use ALL the space!" This is not the case for desktops.

In Desktoplandia, we operate in a windowed (or tiling) environment that allows application resolution to be less-than-or-equal-to the display resolution. Some applications may choose to require a full-screen approach, but that is their choice. Other applications may actually place a maximum size on a given window, or not allow resizing at all! Hopefully the developers chose to limit their app for a good reason, but that flexibility is open to them. And while obviously a "maxes out at 800x600" application might be under-utilizing a 1920x1080 monitor, the environment is one wherein we have come to accept that this will happen from time to time. Furthermore, when it does happen, it's not always a loss in the usefulness or experience provided by the application.

So while the post succeeds at lampooning the dismay of mobile developers for a moment, I don't feel it fully captures the difficulty in being forced into an ever-changing, forced-upon-you full screen mode in an environment where people are less accepting of a "Well my app just happens to run smaller on your huge screen" approach.

> "Well my app just happens to run smaller on your huge screen" approach."

Android has adequate methods of dealing with that (density independent pixels, Relative positioning/Box model of layout) so I still think the source argument holds.

I found this quite funny. I actually hear similar arguments from iOS developers quiet frequently with regard to Android. They've gotten lazy and think that designing flexible layouts that work on multiple screen resolutions is a challenge that they shouldn't have to solve.

Don't get me wrong fragmentation of versions & hardware is making some tasks difficult for Android devs but I feel like some devs are regressing in their mentality towards software challenges thanks to iOS.

It's not just about screen resolutions. It's also about higher support costs due largely to the 50MB limit for Android apps, which apparently causes issues for some users, and lower proportion of paying customers due to implementation issues with Google Play/Android Market. I linked the piece by Mika Mobile that this sarcastic piece is in response to below, but here it is again, with much more detail of some of the major issues with the Android platform:

http://mikamobile.blogspot.com/2012/03/our-future-with-andro...

Sorry, I didn't assume the post was a response to an article I haven't read. I took it at face value and as I said I enjoyed his point.

I think they changed the 50 MB limit recently.

A windowing system does wonders to unify different screen sizes.

Try making an application that runs equally well in full screen on an 11" MBA and a 27" iMac. The fact that windows can be roughly the same size on both machines gives you some common ground.

The question is how many people will pay for it.

Not always. It also depends on weather the programmer actually wants to spend a lot of his/her time trying to fiddle with getting the app to run fine on all different OS combinations as opposed to adding new features. It's not the most fun part of the whole deal and can be frustrating at times.

Very true. I just wanted to bring up the economic aspect of this issue, as it was one of the main points in the announcement by Mika Mobile of dropping Android support for Battleheart, which I assume this piece is in response to.

http://mikamobile.blogspot.com/2012/03/our-future-with-andro...

They were spending 20% of their time supporting the Android port, while only getting 5% of their revenue from the platform.