I think mobile apps are a very inefficient way of developing, specially if you want your product to work cross platform.
Arguably, it seems to be the speed issue that keeps most companies from choosing html5. When does the HN crowd think that web technologies will become the standard for developing across mobile phones?
The easy obstacles:
Performance. Two more layers of abstraction make browser apps clunky. This will go away when phones get beefier.
Security. In Wonderland apps ask for permissions, app stores check for vulnerabilities and abuse, users read permissions and allow. In Real-land, click-buy-install-use. Browsers can also ask/warn users regarding permissions and features.
Phone Features. Can the browser access the camera, local storage, gps, address book, etc....yes, the browser is a 'native' app, it can have super powers; and then let HMTL5 use them.
The almost-impossible obstacle: the market.
If you are old enough you'll remember that Windows grew like yeast b/c it had Office and that Mac kept its niche b/c it had 'designer software'. Fast forward 20 years and FF OS cannot grow as it wished b/c the phones do not run Whatsapp, the new SMS. Ergo: the apps available weight more in the purchase decision than HW or brand. So, the platform owners will do as much as possible to keep things native b/c if anything runs on anything, and if harware is mostly the same on all phones, what is the differentiator?
Did you know that only a year ago it was possible tu create a native app with basically two features: push notifications and a webview. Inside the webview you could run your web app. Now appstores reject these type of apps, they require x% of the code base/features to be native (I believe x is changing, maybe is 50 now?).
> Did you know that only a year ago it was possible tu create a native app with basically two features: push notifications and a webview. Inside the webview you could run your web app. Now appstores reject these type of apps, they require x% of the code base/features to be native (I believe x is changing, maybe is 50 now?).
I would bet never. yes HTML5 is great for making responsive and mobile friendly sites and apps, and you can use it (with various tools) to compile a native app for a app store however, HTML5 will never be able to match with Objective C, or Java in terms of what can be done in a application.
tl;dr never 100%, their will always be room for real native language apps, but we will see the ratio shift more towards html5 for your basic apps and simple games year by year
Depends on what functions your app is performing. We develop web based apps and would only need a native version to access a users contact list (Maybe there is a way to do this already?). Otherwise we're pretty content without having to deal with the app stores.
I don't think it will ever become the standard. I think that it will continue to improve and more and more types of apps will be built with it. For example Flappy Bird could have been built entirely in html5 as the clones have shown us.
Why should it? I think we'll just stick with native applications for most serious things. The web as a platform-independent abstraction layer is not that great an idea.
It's interesting that people forget the iPhone was originally HTML5-apps-only, there was no app store, there were no 3rd party native apps. The roots are there. They just need to be watered.
I've been tinkering with HTML5 apps for the last few years. If you're on the desktop, I don't think there is reasonably anything you can't do. On mobile, there are still a lot of problems with audio. That is very unfortunate for a host of different types of apps, but for a lot of others it is no big deal.
There are things that I'm doing with my current project, https://www.justwritedammit.com, that I just don't think would be possible without HTML5. Building a cross-platform experience has been significantly easier than Java ever allowed me to do, and I'm getting to hit more platforms. I can deploy changes NOW and I can do it 20 times a day if I want. And I have 100% feature parity between desktop and mobile.
JWD looks nice! Hope to see this evolve into a great option for online writing.
Just wondering: do you plan on supporting Markdown? If not: is that by design? Plans on any formatting syntax whatsoever (considering ePub is the main target export format)?
I've toyed with the idea of using Markdown. As long as the formatting doesn't get in the way of the writing is the key. People have a real problem with focusing on the work that matters. It's really easy to procrastinate by doing busy-work like making sure you're using the right type of bullets for your lists.
My original idea was to have formatting be a completely separate mode. But the more I think about the problem in general, I realize that at some point, it's impossible to 100% disassociate content from presentation. Markdown is probably the best compromise. Some other people have had success with writing full, going-to-print books through Markdown. It's certainly a better choice than a full WYSIWYG or GUI editor.
Formatting is also one of those features that people think they need when they don't. Currently, you write into individual chapters, and the chapters get concatenated correctly to have page breaks and headers. There are many different types of books that need no more formatting than that, novels in particular. I'd rather spend time working on getting full ePub docs implemented (I can write HTML files to ZIP files right now, but haven't done much beyond that, more just need to read the spec and get it done at this point), or getting some of the flow-encouraging mini-games made. Or finally building a real server (this is all 100% client-side right now) so I can let people publish their writing directly to their own blog.
I also need to spend some time in the marketing department. There is a lot to do. It's been just under a month since I threw away the work I was doing in Java (from a year and a half ago) and restarted in HTML5.
hi. bowerbird here, checking in with you two.
i haven't forgotten you, just been working on
smoothing out the latest wrinkle in my thinking,
which i do believe has made a huge improvement.
i'll get some stuff to both of you, quite pronto.
I'm sorry you're having trouble. Do you have version info available? I just tested it on Windows 7, Firefox 28 and it worked. I started building this in Firefox, though now I mostly use Chrome (IDK, Firefox has felt like a beast lately, and I get antsy with browsers).
Sometimes, not all of the javascript files load in the right order and an error occurs because of a dependency. I'm fixing it right now, but maybe reloading the page will help with your issue.
Security. In Wonderland apps ask for permissions, app stores check for vulnerabilities and abuse, users read permissions and allow. In Real-land, click-buy-install-use. Browsers can also ask/warn users regarding permissions and features.
Phone Features. Can the browser access the camera, local storage, gps, address book, etc....yes, the browser is a 'native' app, it can have super powers; and then let HMTL5 use them.
The almost-impossible obstacle: the market. If you are old enough you'll remember that Windows grew like yeast b/c it had Office and that Mac kept its niche b/c it had 'designer software'. Fast forward 20 years and FF OS cannot grow as it wished b/c the phones do not run Whatsapp, the new SMS. Ergo: the apps available weight more in the purchase decision than HW or brand. So, the platform owners will do as much as possible to keep things native b/c if anything runs on anything, and if harware is mostly the same on all phones, what is the differentiator?
Did you know that only a year ago it was possible tu create a native app with basically two features: push notifications and a webview. Inside the webview you could run your web app. Now appstores reject these type of apps, they require x% of the code base/features to be native (I believe x is changing, maybe is 50 now?).