Hacker News Reader

Comment by criddell | Hacker News Reader

And then somebody in management asks why the new app is missing so many features on his brand new iPhone. In fact, all the C-level folks and board members are primarily iPhone and iPad users and none of them are happy that so many goodies are missing.

If you aren't worried about provided a first class experience to your iOS customers, then build for Chrome + Android. Although, that sounds a little like "build for IE6 + Windows" 15 years ago.

View on HN

> If you aren't worried about provided a first class experience to your iOS customers, then build for Chrome + Android. Although, that sounds a little like "build for IE6 + Windows" 15 years ago.

Except that Android is the open one that is implementing agreed upon standards, and Apple is the laggard left behind.

> Apple is the laggard left behind

citation needed

The article itself cites things like the lack of support for WebRTC and Service Workers.

I can't find a single iOS app I use that needs either of these.

I'm not sure if you're trolling or sincere here. Of course iOS apps don't need them, they are native. We are talking about the capabilities of the iOS Safari browser.

That's my point. I'm rarely, if ever, in my browser on either of my iOS devices (iPhone 6 or iPad).

If you're not writing a native app, don't bother. Facebook learned this lesson, I'm sure everyone else does eventually as well.

Yeah, because the browser doesn't offer the capabilities needed to render native apps redundant. I'm not sure why we're talking in a giant circle here.

Clearly, the "build for IE6 + Windows" of 2016 is "build for native iOS", which was the original point I was making.

iOS apps are written in languages that have the ability to do equivalent things.

Service workers seems like a performance and battery nightmare.

You learn to evaluate technologies working in a relevant sector or studying some information systems stuff. No citation needed IMHO.

Would you demand citation if I tell you that 'mov ax, data' in assembler moves data to ax or look it up if you don't trust your own judgement at the moment?

They'd probably ask whether we're working with AT&T- or Intel-style syntax ;)

http://caniuse.com/

> implementing agreed upon standards

Yeah, except most of your users' devices are still running 2+ year old software versions.

But unlike in iOS, the browser is a Play Store app which is updatable outside of major OS releases. So a two year old phone can run the latest version of Chrome.

Why Apple still doesn't do this mystifies me.

Telling users that they have to install a new browser before they can fully use your app is a lousy experience though. And then two months later they will uninstall Chrome because (as far as they know) they never use it.

That's not how it works. Most Android phones have come with Chrome preinstalled for a while now. That version is automatically updated, just like any other app.

It was enough of a problem less than four months ago that phonearena.com published an article about it:

http://www.phonearena.com/news/How-to-change-your-Android-ph...

I don't think Chrome is as ubiquitous as it should be. Do new cheap tablets (like the Kindle Fire) even ship with Chrome?

New phones from Samsung don't have Chrome set as the default browser (not sure if it's even installed by default).

It probably isn't on the Kindle Fire, no, but neither is the Play Store. It might use Android as its base, but it's not really an Android tablet.

And while I don't doubt there are many variations of Android device without Chrome installed, the devices that dominate the market (Samsung Galaxy and the like) have had Chrome preinstalled for a long time. From a brief Google it sounds like the default browser varies between phone, provider, etc. etc., so I can't say for certain. What I can tell you is that my analytics logs show the vast majority of Android users are running Chrome.

The Kindle Fire ships with the Amazon Silk web browser, a Chromium/Blink fork which supposedly tries to serve all content via SPDY tunnel to Amazon's CDN.

I don't think that really serves your argument - phonearena.com post a lot of stuff like that either as reference for people who Google the right question, or because the writing effort:ad revenue ratio is working in their favour.

Oh, ok. I really had no idea.

You can install new browsers through the play store, but I do not recall the stock web browser being updatable back when I used Android. Can it be updated on Android 2.3/4.x and how much effort does that require?

As for browsers installed through the play store, how many users can reasonably be expected to install them and of them, how many would update them?

My phone (from HTC) doesn't include a stock browser any more (Android 6). Chrome was the default, and I think it is for many Android phones.

4.x, yes. 2.x, no. But 2.x now only accounts for around 4% of devices out there:

http://www.statista.com/statistics/271774/share-of-android-p...

Not only does the Chrome app auto-update, on Android 4.3+ so does the WebView component. This makes Android pretty hospitable for hybrid apps nowadays.

That's a really good point.

Do all Android handsets and tablets ship with Chrome now? I'm guessing lots of cheap tablets don't.

And I know Samsung used to have their own browser. Do they ship with Chrome now?

KitKat was the first version to ship with Chrome as the default browser, but IIRC it didn't replace the system webview or become auto-updating until Lollipop. Wikipedia shows 70% of Android users on KitKat+, and about 35% on Lollipop+.

https://en.wikipedia.org/wiki/Android_version_history#/media...

I dunno what Samsung ships with, when I was still doing a bunch of mobile web work (early 2014) they used a forked version of Android Browser that was really terrible. Hopefully they've replaced it with something saner. A lot of recent Samsung devices have dropped Android entirely for Tizen, though.

My Samsung S4 came preinstalled with some off-brand default browser that was worse than terrible. Hopefully they have gotten their act together by the time I'm due for an upgrade this fall.

Who cares what it came preinstalled with? My S4 was automatically updated to Android 5.1 and comes with all the stuff that 5.1 normally comes with. I also installed Firefox as my default browser there.

Why bother to upgrade to another phone? The newer phones (except the S5) aren't any better. I love my S4. The only thing I'd bother upgrading to is an S5, since it's waterproof, and still has a removable battery.

I think this is more accurate:

http://developer.android.com/about/dashboards/index.html

Probably yes, although they seem to show basically the same numbers. I think the Wikipedia graph comes from someone scraping the official dashboard periodically; the text of the Wikipedia article indicates that the data source is Google Play Store and excludes non-Google forks.

No, they're running 2+ year old OSes, not 2+ year old apps.

Just like Windows 7 still runs the latest Chrome despite being a whopping 6 years old, so, too, does Android KitKat run the latest Chrome.

You still need to write apps to be compatible with 2+ year old OSes if you want them to be able to install those apps. You don't get infinite forward compatibility for free. There are many updates I stopped being able to get on Android 2.x after 4 was released.

The market share of <Android 4.0 is small enough now to ignore, especially if you are also ignoring iOS entirely. Definitely not enough to stop you from writing a webapp in the first place.

We're talking about websites here, not native apps.

So... no, you don't. Browsers already did the compatibility work for you.

> Except that Android is the open one that is implementing agreed upon standards, and Apple is the laggard left behind.

Given that many of these things are potential exploit vectors, I am not inclined to fault Apple for taking its time to implement them.

Also, we are at the point where the standards are driven by a select few companies. If I recall correctly, those are Apple, Google, Microsoft and Mozilla. Opera used to be there too, but I doubt Opera does much of anything there anymore and I suspect the same either is or will soon be true for Mozilla given its leadership issues.

Do you have any evidence for your claim about "potential exploit vectors", or is it just speculation?

I don't know if you can claim Chrome is insecure, simply because it supports more web standards.

e.g. look at the results of the recent pwn2own contest:

http://venturebeat.com/2016/03/18/pwn2own-2016-chrome-edge-a...

There are other variables that probably have a greater impact on the security of a browser.

In the past, changes to internet standards were about enabling better presentation of webpages. Now changes involve JavaScript APIs that access things traditionally reserved for plugins (which were a security nightmare) and native applications. Doing that can pose unforeseen security risks, such as the supercookie:

http://money.cnn.com/2015/01/09/technology/security/super-co...

Not to mention, I posted elsewhere in this discussion that WebBluetooth seems like a way that an attacker with control of a server can run attacks against bluetooth devices that he otherwise would not have been able to access without breaking out of the sandbox. Once in control of one of those bluetooth devices, it would be possible to attack the system from the compromised devices. It is like giving unrestricted network access to Javascript once a user clicks accept (and they almost all would), except with a different network technology where devices all trust each other not to do malicious things.

Given that security flaws can happen when new features are used in unforeseen ways, I am happy to see Apple take its time under the assumption that it is taking its time to implement things correctly.

Reply on news.ycombinator.com