Christian Heilmann

These apples don’t taste like oranges – let’s burn down the orchard

Tuesday, February 19th, 2013 at 3:52 pm

When I see comparisons of HTML5 to native apps I get the feeling that the way we measure failure and success could give statisticians a heart attack. Take Andrea Giammarchi’s The difficult road to vine via web as an example. In this piece Andrea, who really knows his stuff tries to re-create the hot app of the month, Vine using “HTML5 technologies” and comes to the conclusion – once again – that HTML5 is not ready to take on native technologies head-on. I agree. But I also want to point out that a 1:1 comparison is pretty much useless. Vine is only available in iOS. Vine also was purposely built to work for the iPhone. In order to prove if HTML5 is ready all we’d need to do is to find one single browser on one single OS, nay even only on one piece of hardware to match the functionality of Vine.

sobbing mathematically

Instead we set the bar impossibly high for ourselves. Whenever we talk about HTML5 we praise its universality. We talk about build once and run everywhere and we make it impossible for ourselves to deliver on that promise if we don’t also embrace the flexible nature of HTML5 and the web. In other words: HTML5 can and does deliver much more than any native app already. It doesn’t limit you to one environment or hardware and you can deliver your app on an existing delivery platform – the web – that doesn’t lock you in to Terms and Conditions that could change any time from under you. Nowhere is written though that the app needs to work and look the same everywhere. This would actually limit its reach as many platforms just don’t allow HTML5 apps to reach deep into hardware or to even perform properly.

What needs to change is our stupid promise of HTML5 apps working the same everywhere and matching all the features of native apps. That can not happen as we can not deliver the same experience regardless of connectivity, hardware access or how busy the hardware is already. HTML5 apps, unless packaged, will always have to compete with other running processes on the hardware and thus need to be much cleverer in resourcing than native apps.

Instead of trying to copy what’s cool in native and boring and forgotten a month later (remember Path?) if we really want to have HTML5 as our platform of choice we should play it to its strengths. This means that our apps will not look and feel the same on every platform. It means that they use what the platform offers and allow lesser able environments to at least consume and add data. And if we want to show off what HTML5 can do, then maybe showcasing on iOS is the last thing we want to do. You don’t put a runner onto a track full of quicksand, stumbling blocks and a massive wind pushing in the opposite direction either, do you?

HTML5 needs to be allowed to convince people that it is a great opportunity because of its flexibility, not by being a pale carbon copy of native apps. This leads to companies considering native as the simpler choice to control everything and force users to download an app where it really is not needed. Tom Morris’ “No I’m not going to download your bullshit app” and the lesser sweary I’d like to use the web my way thank you very much, Quora by Scott Hanselman show that this is already becoming an anti-pattern.

Personally I think the ball is firmly in the court of Android to kill the outdated and limiting stock browser and get an evergreen Chrome out for all devices. I also look for Blackberry 10 to make a splash and for Windows phone and tablets to allow us HTML5 enthusiasts to kick arse. And then there is of course Firefox OS, but this goes without saying as the OS itself is written in HTML5.

Share on Mastodon (needs instance)

Share on Twitter

My other work: