Hacker News Reader

Comment by tsunamifury | Hacker News Reader

I design and support apps for iOS (Tablet and Phone), Android (Tablet and Phone), Blackberry (Tablet and Phone), Windows Phone 7.5, and various J2ME (Nokia s40 etc).

Android is not the very worst one to support (that would be blackberry), but it is the worst within its own ecosystem. As a designer it infuriates me to no end to never know the Dimensions, PPI, or general experience of any given user. It makes any refined UI design nearly impossible. While I can't create great experiences on specific devices, for Android I have to make glorified web apps at best.

If I were a Android customer, I would be irritated that my platform has significantly handicapped my apps.

View on HN

I'm actually mostly annoyed at my phone manufacturer for not being on top of software most of the time. I finally got ICS for my Dutch Samsung Galaxy S2, but not after a bad update left me with a phone that turned off every hour and had to go to a service center. The service center in turn flashed my phone with a Luxembourg version of the software, which properly upgraded to ICS but I no longer have a "Dutch" phone or the very latest stability upgrade.

The article mentions video - that's directly related to either 3rd party ROMs or phone manufacturers not paying for/getting licenses for MPEG and/or other standards. Google APIs is another - there's various additional requirements that manufacturers have to deal with to get those. There is an alternative BTW to Google's maps - Mapquest (http://developer.mapquest.com/web/products/featured/android-...)

> As a designer it infuriates me to no end to never know the Dimensions, PPI, or general experience of any given user. It makes any refined UI design nearly impossible.

This doesn't make sense to me. Android gives you all the tools you need to scale a UI to different types of screens, so why would not having a set dimension/resolution/PPI value make it hard for design? The added challenge is thinking about how your UI may work gracefully on smaller or larger than expected screens, but those challenges can be overcome as they are on other platforms like the web.

I've been told the concept of scaling a UI is akin to the concept of "write once, run anywhere".

That is, it works fine in theory. It just hasn't ever worked as advertised in reality. And trying has often generated more pain than it's saved.

It's worth noting that much of modern web design moved to a place of fixed minimum widths and white-space taking up the remainder. Responsive design promises to change that a bit toward scaling UI, but thus far is just a mechanism by which designers can incorporate a few hand-tuned minimum-width variants.

Sencha touch does scalable UI across tablet and phone form factors. They even resize controls so they have the same physical dimensions on different dpi screens, and this is all in the browser. I'm sure when you go native the possibilities only increase.

Right from the start Android has been designed with a form of Responsive Design. With scalable XML layouts, and selectors to choose which layout is loaded for different DPI devices, one could argue that even early versions of Android are at least as design-friendly as modern HTML5.

That's not to say there aren't challenges. But the whole point of Android was always to create an industry standard for a wide range of devices. This requires - by definition - support for a variety of screen sizes, DPIs, hardware features, etc. When most people talk about fragmentation as a negative point of Android, they're forgetting what Android is. A common platform, running on an unlimited number of devices. That's by design, not accident.

The reality of the matter is, the more you scale and stretch, the less you actually design for any given sized platform.

Compare a Galaxy Note to a Droid Razr. They have different sizes, dimensions, pixel densities and various other factors. The UI doesn't just 'stretch' and fix all that. It stretches and looks terrible, feels too large or too small on some devices, and makes fonts et al a mess.

Then throw in a Galaxy Tab, a Kindle, and a more tablet/phone 'tweeners'. Its a total mess. Then throw in operating system versioning and various hardware support and you see why many developers build for iOS first, then consider an Android app.

Edit: I support roughly 200 apps multiplied across all platforms on 20+ app stores. (Roughly 1000 app builds). I see more problems that your average developer sees.

So there are added challenges, and added work, compared to supporting only four screen sizes.

How many different-sized images do you save out if you want a button to be crisp on all screens? Now how many if it's a compressed (gpu format) image?

as an android customer I am pretty glad I had the choice to buy the device I wanted to buy, instead of the one you wanted me to buy

I've commented on this before, but I am a soft keyboard hater. I have a hard keyboard on my phone and tablet, and won't buy a device without the option.

I'm incredibly glad that Android doesn't go with the dogmatic 'our choice is the best choice' approach. It means I can use things as I want them, not as Google wants me to.

I can't agree with you enough.

precisely. people like choice. some designers wish it weren't so, that we would just all get in line with their "one true design", but people like choice. that's why there are dozens to hundreds of models of tv, car, dishwasher, shaver, ice cream maker, flashlight, battery, dvd player, chocolate bar, and yes, smart phone.

this whole "fragmentation is a problem" argument falls on this one point alone. fragmentation isn't a problem, its a fact. as a pro android dev for a few years now, android is resoundingly the best sdk/os that I've ever used for mitigating such an insanely diverse set of hardware (tv, watch, dumb phone, smartphone, tablet, laptop, no problem)

I've scaled glitzy over-designed android apps from the smallest phone to giant Google TVs, it is very doable.

Reply on news.ycombinator.com