Interesting to see deep comparison of Sencha Touch and jQuery Mobile frameworks. Both seems to be good enough for making real applications. I don't mean computational heavy applications like games or augmented reality, they require native code optimised for each platform. But for facebook-, or twitter-like applications HTML5 is okay.
I think this proves that what Sencha has done is incredibly effective link bait. It has not proven anything about HTML5. Sencha makes a great tool for building HTML5 apps. There are still plenty of issues with building HTML5 web apps.
Also, part of the reason HTML5 keeps "getting better" is Moore's Law. If we keep adding processing power and webkit keeps improving, then HTML5 keeps "getting better" which is not exactly true. On 2-3 year old hardware how does it run?
I like writing HTML5 mobile apps, but what Sencha has done is overblown. Good marketing, that's it.
HTML5's limitations are at the hardware/browser level, not due to a lack of Sencha development prowess.
This is a carefully constructed scrolling demo (which still shows evidence of lackluster behavior -- look at when he goes from vertical to horizontal!), which doesn't show much other than what Sencha (and any other framework) has been able to do all along.
I'll admit the demo was impressive but there's still noticeable lag in animations. So far the only web app that I've seen come even close to native is Hacker Web: http://hackerwebapp.com/
I don't know about the rest of HN but I'm become somewhat tired of the HTML5 vs Native chatter. I think we can all agree that native performs better. Nuff said.
We're a really, really small shop, developing a webapp. We would love a mobile app before we launch, but nobody here has ever developed a native app before - our experience is strictly related to web development targeting desktop browsers.
We could probably throw together an HTML5 app for Android and iOS (phonegap, etc) pretty fast. Or I guess we could try and outsource development of native apps for Android and iOS, except we're not sure of cost, or risks, or how to find a developer, or whether the result would be up to our standards (our core selling point, is basically, "this webapp is like the others you've seen, but with a UI that'll blow you away"), so a slow HTML5 app is bad, but so is a clunky and ugly native app developed by someone who doesn't have the same obsession with attention to detail that we do.
So...we've got a choice to make, and "native performs better" isn't enough. How much better? And what do you give up to get native? Does Sencha (or whatever) get you close enough for our purposes? How much harder is a good HTML5 app to develop compared to a cude wrapper around an existing site? These tradeoffs are not obvious, and for people currently evaluating options, I think they're interesting.
I get where you're coming from. As a web developer I'm all for the web in general. I suppose my frustration stems from how an evaluation of a new web framework leads to a frustrating back and forth on native vs. web. It's a hot topic.
I'm currently working on my first iOS app (yes I know I'm late to the game) using Trigger.io: https://trigger.io/
Still working my way through the documentation but so far it's amazing. Consider having a look.
I've been developing several native mobile apps for android platform during the last 18 months. Actually i'm developing on iOS too. I used the Sencha Touch framework (and their Architect tool) to build one of the apps. Here my experience:
- Sencha did great efforts but HTML5 is maybe supported best on iOS than on Android so that great differences still exists between the same HTML5 app executed on Android or iOS
- You have different (too much different) behaviors between android 2.3 and android 4.0 also building an html5 app
- Sencha tools are really (too much) bugged for android development (this is more related to Sencha and not HTML5, but we are talking of cross platform build so it should matter)
- Simply (at least for android), HTML5 app speed depends on os version but hardware too (quite obvious!): so you can have an html5 app running on a modern android4.0 device and still see lags (a lot of lags) or other problems related to HTML5...if you develop a native app, it is probable that you won't notice problems regarding performances on different android 4.0 devices, but also with not too old 2.3 devices.
My experience is based on an html5 app built using Sencha 3 months ago, released for both iOS and android. I tested it on iphone4, Samsung Galaxy Note 1 and 2, Galaxy S2 and S3 so i can confirm that it is impossible to tweak an html5 app to be well performant on all these platforms/devices (so if you want to do it, you have to spend so much time that you will surely ask yourself "Maybe i should have build a native iOS app and a native Android app: i would have spend maybe a little bit more time, but i would have had reasonable return on performances and coolness of my app!"
9 comments
[ 2.8 ms ] story [ 36.4 ms ] threadAlso, part of the reason HTML5 keeps "getting better" is Moore's Law. If we keep adding processing power and webkit keeps improving, then HTML5 keeps "getting better" which is not exactly true. On 2-3 year old hardware how does it run?
I like writing HTML5 mobile apps, but what Sencha has done is overblown. Good marketing, that's it.
This is a carefully constructed scrolling demo (which still shows evidence of lackluster behavior -- look at when he goes from vertical to horizontal!), which doesn't show much other than what Sencha (and any other framework) has been able to do all along.
I don't know about the rest of HN but I'm become somewhat tired of the HTML5 vs Native chatter. I think we can all agree that native performs better. Nuff said.
Maybe we can revisit this in a year's time.
No, not enough said. Performance isn't everything.
We're a really, really small shop, developing a webapp. We would love a mobile app before we launch, but nobody here has ever developed a native app before - our experience is strictly related to web development targeting desktop browsers.
We could probably throw together an HTML5 app for Android and iOS (phonegap, etc) pretty fast. Or I guess we could try and outsource development of native apps for Android and iOS, except we're not sure of cost, or risks, or how to find a developer, or whether the result would be up to our standards (our core selling point, is basically, "this webapp is like the others you've seen, but with a UI that'll blow you away"), so a slow HTML5 app is bad, but so is a clunky and ugly native app developed by someone who doesn't have the same obsession with attention to detail that we do.
So...we've got a choice to make, and "native performs better" isn't enough. How much better? And what do you give up to get native? Does Sencha (or whatever) get you close enough for our purposes? How much harder is a good HTML5 app to develop compared to a cude wrapper around an existing site? These tradeoffs are not obvious, and for people currently evaluating options, I think they're interesting.
I'm currently working on my first iOS app (yes I know I'm late to the game) using Trigger.io: https://trigger.io/
Still working my way through the documentation but so far it's amazing. Consider having a look.
EDIT: Hello from Auckland :)
My experience is based on an html5 app built using Sencha 3 months ago, released for both iOS and android. I tested it on iphone4, Samsung Galaxy Note 1 and 2, Galaxy S2 and S3 so i can confirm that it is impossible to tweak an html5 app to be well performant on all these platforms/devices (so if you want to do it, you have to spend so much time that you will surely ask yourself "Maybe i should have build a native iOS app and a native Android app: i would have spend maybe a little bit more time, but i would have had reasonable return on performances and coolness of my app!"