Live data from Hacker News

The Decline of the Mobile Web

cdixon.org

191–200 of 209 posts

Re: The Decline of the Mobile Web

#191
This is both bad and good.

Good: It might stop the rot of websites turning their whole site into a mobile version with fat-finger buttons and massive text, if websites become the means of interacting for people with real computers again.

Bad: Apps are a far bigger security risk, and prevent casual interaction (i.e. following a link onto something, then going somewhere else) - having an app means you're in a selfcontained site and are kept there until you close the app and open another.

Re: The Decline of the Mobile Web

#192
post #108

I think Facebook is an instructive example here. They tried for a long time to make their native app (for Android, anyway) just a wrapper over the mobile web client, but they could never get the performance all the way there. They went native and have seen great results. They resisted as long as they could but eventually the products had to diverge. I'd say 80% of developers who choose native do so for performance, a…

Facebook's mobile website doesn't want permissions to access camera, mic, location, and SMS. That's why I use that and not their stupid app, especially since it lets me switch between other sites as I like as well.

Re: The Decline of the Mobile Web

#193
post #164

Earlier quoted context omitted.

Sure, native apps have better performance on mobile. But the same thing is true on desktop, yet the web somehow stays popular on desktop. That suggests that the behavior of platform owners (Apple and Google) is the real reason for the popularity of apps on mobile, and performance is just a distraction.

As lugg said, it took a long time for web apps to rival native apps in functionality, and there are still a lot of places where they don't (large files, graphics, etc). In ~3-5 years the class of web apps that are lightning fast on desktop now will be fast on mobile.

>and there are still a lot of places where they don't

This would be better phrased as "and there are still only a few places where they do". Google Docs, for example, is great for basic collaborative editing (as long as you don't mind the NSA reading it), but has a UI that it a total pain (especially the world's worst keyboard shortcuts, e.g. having ctrl-enter and shift-enter reversed, and completely lacks advanced features.

Re: The Decline of the Mobile Web

#194
post #77

It seems 'mobile apps' are like Flash was in the 90's, basically a way to take full control over the UX to achieve some particular goal. Given the preponderance of such apps it seems as if home screens are due for a make over that looks more like iPod music libraries than 'desktop.' Basically something which supports dozens if not hundreds of apps on a device.

The entirety of current web design practice is taking full control over the UX. Gone are the days when you sent a GET and got an html document back, the most interactive element of which was a textarea box. The other reason flash was a pain, incidentally, was because it was easy to bog down a client computer with overzealous flash stuff... just like the current tendency to use javascript to do all your rendering and…

"Gone are the days when you sent a GET and got an html document back, the most interactive element of which was a textarea box."

I love the irony that this accurately describes the very page we are reading this text on...

Re: The Decline of the Mobile Web

#195
post #77

It seems 'mobile apps' are like Flash was in the 90's, basically a way to take full control over the UX to achieve some particular goal. Given the preponderance of such apps it seems as if home screens are due for a make over that looks more like iPod music libraries than 'desktop.' Basically something which supports dozens if not hundreds of apps on a device.

The entirety of current web design practice is taking full control over the UX. Gone are the days when you sent a GET and got an html document back, the most interactive element of which was a textarea box. The other reason flash was a pain, incidentally, was because it was easy to bog down a client computer with overzealous flash stuff... just like the current tendency to use javascript to do all your rendering and…

It was also a pain because there as nothing you could do about performance problems. Nowadays, there is serious work going into optimizing JS, that was impossible with a closed platform.

One wonders if performance could be solve by open-sourcing flash. I liked AS3 more than JS.

Re: The Decline of the Mobile Web

#196

Earlier quoted context omitted.

Sure, native apps have better performance on mobile. But the same thing is true on desktop, yet the web somehow stays popular on desktop. That suggests that the behavior of platform owners (Apple and Google) is the real reason for the popularity of apps on mobile, and performance is just a distraction.

Isn't the whole performance discussion based on synchronizing versus not synchronizing? There's no reason to synchronize on a Desktop as one can assume there's wifi available.

I think that sync vs not syncing tends to be cyclical.

We all started using Gmail because synchronising to a mail server is annoying and PCs where slow. Ten years later machines have a fast internet connection, 1TB HDD as standard and, maybe an SSD. When you have to wait 10 seconds just to view an attachment syncing makes a lot of sense. But generating a thumbnail of that same PDF would be much better handled pre-emptively in the cloud.

The problem is believing that one approach is better for every function of an app. Few user machines will ever have a guaranteed large amount of bandwidth. This is just as true on desktop machines as mobile, but less acknowledged.

Re: The Decline of the Mobile Web

#197
post #106

Earlier quoted context omitted.

>Apple routinely rejects applications to their App store. Have you not read any of the numerous stories where app devs get denied? Yes, I am aware. And yes, you are right. But the gatekeeping on the app stores has nothing to do with the music, movie or even book industry. Could they be more lenient? Yes, but you can't denied that the game has indeed changed. >I'm speculating of course, but real world anecdotes would…

Sure, you have the freedom to install any apps you want. However, if I am starting a business, and most people don't know how to exercise that freedom to install an app from somewhere else, I am not going to get very far in my business if this is what is required for my end users to use my program. Therefore, I am not going to invest time in developing anything that requires that sort of installation, making it pract…

But you make it sound more like an education problem than a freedom problem.

The issue is not that people are not allowed to install apks, the problem is that people don't know how to exercise that freedom.

Re: The Decline of the Mobile Web

#198

Security could become a big factor in web vs native. For example, I don't trust the facebook native app. I've used both the web and native versions and the native version is smoother and better. But I don't trust facebook. There's a lot of personal information on mobile phones these days and I don't want facebook accessing that data. For me, using the web version is much safer. And random apps on the app store from u…

XPrivacy[0] is the way to go. Require root though. Should be natively included in Android imho

[0]https://play.google.com/store/apps/details?id=biz.bokhorst.x...

Re: The Decline of the Mobile Web

#200
post #108

I think Facebook is an instructive example here. They tried for a long time to make their native app (for Android, anyway) just a wrapper over the mobile web client, but they could never get the performance all the way there. They went native and have seen great results. They resisted as long as they could but eventually the products had to diverge. I'd say 80% of developers who choose native do so for performance, a…

> Facebook is an instructive example here. They tried for a long time to make their native app (for Android, anyway) just a wrapper over the mobile web client, but they could never get the performance all the way there. They went native and have seen great results.

Except that anyone who actually uses the Facebook apps knows that the performance is horrible in the native apps– jerky scrolling, lots of on-demand loading requiring you to wait for multiple RTTs on every click, etc. I would be hesitant use Facebook as an argument for anything other than the need to place UI performance on equal priority with marketing.

Post reply on HN