Hacker News Reader

Comment by ryandvm | Hacker News Reader

As a side-time Android developer, I can tell you that the platform is anything but buggy. For a while Android antagonists were painting the diversity of handsets as a huge testing hurdle, but that's just not the case. Write sane code and don't engage in sloppy shortcuts and you can get by with testing on a single handset (the emulator is too slow).

My main app is a live wallpaper that pulls radar imagery from NOAA (http://www.appidio.com/apps/radar-wallpaper/) and I have yet to hear of a device that it doesn't work on.

View on HN

> "I have yet to hear of a device that it doesn't work on"

Other than those that can't access the store and those that won't be getting an update to 2.1?

"those that can't access the store"

...such as?

Pick a tablet.

You mean the Samsung Galaxy Tab? It has the Market. Perhaps the upcoming Motorola XOOM? Yep, it has the Market.

The only two exceptions I can think of are the Archos tablets and Notion Ink: Archos doesn't fulfill Google requirements, and I don't know what the problem with the Adam is. (Obviously, no-name Chinese tablets also don't get the Market, but that's not a huge surprise.)

I'd have phrased it the other way: the only two exceptions that I know of that can access the market are the Galaxy and Xoom. One of which isn't available and the other is almost a poster child for the uncertain upgrade path problem.

No-name Chinese tablets aren't a surprise, but the fact that they get casually included when people talk about ecosystem devices and sales is part of the problem.

besides devices that can't access the store are those with limited access. Like the HTC Wildfire and the other devices with screen resolutions lower than 320x480

The HTC Wildfire (and any other phone with a low resolution screen like it) can access the Android Market.

yes, they do, but it's limited, meaning there are a lot of applications that don't show up in their markets.

Someone should probably point out this is a good thing.

It's exactly how Apple does it with the iOS platform - some apps are targeted exclusively at the iPad, and the App Store filters on screen size. Some people say this is fragmentation, but I think a better term would be specialization. All platforms do it - the question is how easily the platform lets deal with it. I think Android does it pretty well (as does iOS).

If you want to target the low-resolution devices then you put it in your manifest and it will appear in the market.

I'm sure the same thing will happen with Google TV - some apps will be designed exclusively for Google TV, and so won't appear for phones. Doing anything else would kind of miss the point of the market!

(It's also worth noting that many Android apps have download links from a website, so if you really, really want to use an app that isn't designed for your device you probably can)

The nook color has no market place access.

The Nook Color is pretty different from other Android devices - it's locked down by B&N and they're planning on opening their own app store for it. It's also missing menu, back & search buttons, found on every other 2.x device.

In addition, it isn't being marketed as an Android device.

You mean like these guys http://androidandme.com/2010/09/carriers/v-cast-app-store-to... http://phandroid.com/2010/04/13/vodafone-opening-up-shop-wit...

I can't say I remember any of the Droid marketing mentioning running android either. It is a little different but not that different.

If you take a look at the Barnes & Noble site for the Nook Color, you won't find a single mention of Android - it's pretty clear that they do not intend to be a part of the Android ecosystem at all, unlike the Droid.

Look, downvotes, I know that all platforms eventually cut off older hardware. But with Android the inconsistency in support for the market and OS updates are a notable part of the fragmentation concern.

The fact that hardware under a year old has been cut off is a cause for concern. The fact that hardware is launching without the market and/or will be launching with third party market apps is a cause for concern. Device numbers are bandied about in the general sense -- x-gajillion devices per month! -- but no context is available as to how many of those can access the store or got the last OS update, let alone how many will get the latest. That's a cause for concern.

I'm not suggesting Android's a bad or flawed platform. I'm just not willing to pretend that it's all roses either.

We're honest when we say "Going iOS can make you a ton of money, but the gravy train is always a subjective-review away from coming to screeching halt."

So what exactly is wrong with pointing out that this fragmentation thing isn't something we can just hand-wave away?

Cause not developing for Android cause of non-Market accessing devices and the fact that some 10% of devices are below 2.1 is like not developing for Apple cause a certain number of people never update their phones and the looming app rejection. It is a real problem, but it is really kinda something you have to handwave away if you are going to be realistic about doing business on a platform.

Who said it was a reason to not develop? No-one's making that argument.

I'm arguing in favor of being realistic. And part of that includes admitting that there are causes for concern. And because of them, you might invest efforts into an app only to find yourself in 2012 facing the most-popular android devices shipping with only a carrier- or manufacturer-specific markets installed by default, each with their own hoops.

And, no, even that situation won't be a blanket argument against developing for Android. But it also won't mean fragmentation was something that is a non-issue so long as you have a good coder.

I feel like the lack of Android OS upgrading is the best planned obsolescence ploy ever. Not even Apple dares to pull such an heist, offering updates to most old devices.

Updates that vary in features and have a history of crippling devices. Trust me, every tech company is on the same page. Most obvious recent Apple example: cameraless iPad. Of course, Apple does a remarkable job of positioning themselves as customer friendly, so that helps lower the blowback.

What platform are you on? I found the emulator unusable on modern OSX and Windows 7 desktops and, to my surprise, perfectly usable on my two year old laptop running Ubuntu.

I find the emulator rather slow on Ubuntu... Best just to develop on a handset!

It's a pain to use on Windows. I had a pretty beefy machine that was brought to a crawl by the emulator. I hope Google is listening. Visual studio for Windows Phone 7 was not such a hog and I tend to spend more time working in it these days.

What SDK version are you using in the emulator?

I find 2.1, 2.2 and 2.3 good (on Windows), but the 3.x prerelease is pretty slow (it's unoptimized yet, so I guess I should expect that)

Do you recommend any particular handset?

I'm pretty pleased with the Droid Incredible. It's quite fast, so it makes a good dev handset. I'm not necessarily sold on the HTC Sense UI. I don't think it's necessary. Especially since I suspect it increases the amount of time it takes them to port new Android releases.

Wouldn't a slower handset be best to test on? That way, if it runs smoothly on the slower handset you'll know it runs smoothly on faster ones.

Reply on news.ycombinator.com