As a web vs. app proponent I love this. I do have a few apps installed on my Note 4 but I prefer getting Twitter, FB, etc. as a web site.
Web standards like HTML5 are wonderful so let's use them and make the web awesome.
My brother and one friend each have a few hundred apps installed on their phones. I am sometimes surprised that their phones still function with so much cruft installed, never mind the privacy considerations.
Websites also can't spam you with push notifications the way installed apps can, so I strongly prefer them as a user. But I suspect this is one reason companies like apps.
I'm not sure whether it's my imagination, but this seems to have gotten a lot worse in the past year or two, and is no longer limited to apps from shady no-name companies. Apps from reputable companies used to try to have some plausible reason related to your actual use of the app for a notification, but lately Yelp and Hotwire have just started brazenly spamming ads unrelated to any activity. So I uninstalled those two, and am now much more resistant to installing new ones.
>Websites also can't spam you with push notifications the way installed apps can, so I strongly prefer them as a user.
On Android, at least, you can disable notifications on a per-app basis. Press and hold on a notification, touch the icon that appears, and uncheck the notifications checkbox on the next screen.
I'm not sure it can be done directly from the notification, but on iOS you can also disable notifications per app and specify how it should be displayed (pop up vs at the top etc.)
I've also noticed an increasing amount of apps that now also use the notifications for ads. Definitely not what I like to see on my lock screen.
>I've also noticed an increasing amount of apps that now also use the notifications for ads. Definitely not what I like to see on my lock screen.
Once again, Apple is lagging behind by several years ;) This kind of Ad used to be incredibly common on Android for a while, but Google eventually stepped in and banned these sort of apps from the store. Honestly, I'm a little surprised that Apple allows them.
While it's gotten rid of the worst spam (apps sending totally unrelated ads as notifications), unfortunately there are holes in that policy big enough to still drive a busload of junk notifications through, as long as they're somehow related to the app:
> Apps and their ads must not display advertisements through system level notifications on the user’s device, unless the notifications derive from an integral feature provided by the installed app. (e.g., an airline app that notifies users of special deals, or a game that notifies users of in-game promotions).
To me those are still pretty spammy. An airline app notifying me that a flight I have a ticket for is delayed would be one thing; that's a legitimate use of system-level notifications. But an airline app vibrating my phone, just to announce that AirlineName Has Great Deals To The Caribbean In Our End Of Summer Sale? That is not ok, but Google allows it. I assume this is also how Yelp is able to send that kind of junk notification without getting banned.
This can be done via a checkbox on the app's info page. It's nice to know it can also be done directly from the notification too, that's even more convenient.
There's a web feature for push notifications; I got a dialogue box asking me if I wanted to enable it for Facebook the other day. It was easy enough to push No (since push notifications were one of the reasons I got rid of the @#$%er in the first place), but many people will still get those.
The fact that you can choose no is precisely what makes the Web so good. The fact that you can block ads at all, that you can disable JS, that the Tor Browser bundle can be so simple to use, that you can easily get URLs to things, that you can copy text, are all huge benefits of the Web.
Ever tried getting a URL to some Facebook content on mobile? The only way that I've found is to "Send as Message" (Facebook for Android only implements sharing inside Facebook, there's no way to share via email or anything like that), choose myself in the contacts list, then go to Messenger, open the link in a browser, and copy the URL from there. It's infuriating.
Really? Huge difference? I would say about 20% (being generous with that number) of the good apps that I install try to misuse the notification. I don't find it that hard to disable the notifications for them. The benefits of notifications outweigh the spamming by a lot.
Unlike with iOS or Android, in HTML5 invasive APIs (including notifications) are all opt-in. Each permission is granular, instead of needing to choose between granting all permissions an app requires or none. IMO this is a massive improvement.
Although you could grant notification access to an app and then it could become a bad actor, I expect on the web you will be able to be far more choosy about granting access than with an app.
I believe Android is doing something to bring in granular permissions, there was discussion at Google I/O this year.
Push notifications are opt-in and granular on iOS as well. You are prompted to approve each permission for each app, push notifications, location services, access to contacts, etc.
I had to kill the My Fitness Pal app on my Android because the notification dispatch was inside some kind of infinite loop where my phonr would not stop vibrating even after disabling notifications in their UI.
>Websites also can't spam you with push notifications the way installed apps can, so I strongly prefer them as a user. But I suspect this is one reason companies like apps
poorly maintained or non-efficient backend that didn't take note of whether the user receive a push or not leads to this problem. It's not the app problem itself totall
You could argue that apps are worse since they can cert pin and prevent you from mitm'ing your https traffic - you have no clue what data is being sent.
With browsers (for now, Chrome on iOS may start/be cert pinning) you can install a cert and still mitm.
>My brother and one friend each have a few hundred apps installed on their phones. I am sometimes surprised that their phones still function with so much cruft installed, never mind the privacy considerations.
This is a common misconception (also concerning desktop OS "bloat").
As long as those apps are not running at the same time and competing for CPU resources, they could have 100000 apps installed and no issue at all.
Especially with sandboxing and automated resource management in iOS (probably Android too, but not too familiar with it).
Hmm, not necessarily. Unless the user is careful, it's possible that many of those apps would request data in the background, etc., which would result in performance issues.
I mean, Twitter only makes sense when a third party digests their API. The default web client is broken (sometimes intentionally) in many, many ways. Perhaps I'm misreading you?
Of course YMMV, but my go-to is the Mac native client that they themselves put out. Not the web client, but not third-party. I don't have trouble with it.
Another interesting aspect is app size. These days, when many complain that the entry level iPhone only has 16GB of storage, it's nice not to have to install ~200MB social apps.
Here's my question about web vs apps--is the privacy different?
It feels like using different apps is like using separate "browser profiles" for each website, which is possibly what I want--it's so difficult to compartmentalize on mobile browsers. So things at least seem isolated to a greater extent than using the same web-browser for everything?
As a user, I dont really care about it being an App or Mobile Web. It needs two things,
1. An Icon on my Springbroad, so i can easily view it whenever I want to.
2. It should be butter smooth and dont warm my pants after reading. ( i.e lots of CPU cycle. )
Native Apps does better in both department. And whenever users gets a choice most would likely prefer the App version. HTML5 or Web has definitely improve, and it could or will do better then the web if we dont have the Ads constantly loading in the background. ( Ad blocker is another topic )
I guess atavist already have their group of loyal customers it makes sense for them to change.
Personally I see the article mostly complain about the Distribution and Discovery Problem. And some App Store Approval process.
I agree with you, but I must also add that not every site needs to have an app. Mobile web is plenty appropriate and sufficient for many types of apps. Many news and blog sites are a great example of this. In their case, an installable app would make sense if they are offering a feature or experience that requires the privileges of an app. Lets not unnecessarily crowd up our phones and app stores. The World Wide Web is still a good idea.
Techmeme is a prime example of this — bookmark it on your homescreen and it's as good as an app. Their mobile design is, IMO, easier to parse than the web, too.
Financial Times does not have an iOS app - just a mobile website. They have been very clever in building the site so that it is very app-like, for instance they use local storage to store the days download stories.
But, was initially confusing because I couldn't find their app in the app store, where I was expecting it.
But, was initially confusing because I couldn't find their app in the app store, where I was expecting it.
In my experience this is a huge problem. Along with the process of adding to home screen - people just don't get it. Once they have it installed they're very happy, but getting them to that point is difficult.
2. That's only true if the web version is crappy and overloaded with ads and trackers. A decently designed
I prefer mobile web for publications, because it's just a link that doesn't impact my phone. The native integration usually only brings annoyances, like unwanted notifications or sucking the battery to fetch articles I may not read.
Apps are for applications. For content, nothing beats the web.
> Apps are for applications. For content, nothing beats the web.
Exactly. I don't want to read your article over your own dedicated app, I already have a perfectly good browser...
...BUT! I also hate applications made as webpages, because no one can perfectly hide the fact it's not native. In webapps, UI pretty much always lags somewhere, breaks down completely under spotty Internet connection, or does something weird because there's no way for webapp to replicate every aspect of expected UI functionality of user's phone.
The vast majority of non-technical users have no idea you can do this, and I've even heard many say that didn't know you could do it with any site - e.g. they thought they could only do it if the site had a popup that said they could do it.
My mother is 70 and did it on her own once her facebook app killed itself in a botched update (bad install or something). And then she complained that she was using the web version and couldn't find how to do some stuff hehe.
Edit: btw, she's a school/music/languages teacher, so she has very little technical background. In her own words, she "touch[es] everything until it works".
This is the biggest takeaway for me. "Wow guys who would have thought the web was good enough for us to post our text content to." This article isn't about why you should drop mobile app development. News + magazines? Sure. Just look at Flipboard.
Choosing the mobile website over the app is one way of keeping the relationship from going any further. A website can't ask you for your contact list like an app can.
That feature was designed but never built, so his point still stands. In my ideal world sites can do that stuff, they just have to ask, and users can always say no.
eugh, I hate it when sites do that. I'd be really angry if I found out that a friend of mine had been sharing my contact details with strangers without my permission.
While apps have more capability, the web can do pretty much anything you need with content and data. The only time you really need an app is for CPU-intensive work or accessing hardware.
If you want to do any sort of custom drawing or animation you basically can't do it with the web, the performance just isn't there. You must stay on the specific fast paths offered by the browser or you're screwed.
So no, the web cannot do pretty much anything with content and data. The web can do very, very little with content & data, it is just often enough coverage for passive things like articles.
I don't know what kind of custom drawing you are referring to.. the web can work with 2d and 3d drawing and animations just fine. For example, an entire first person 3d game built using WebGL - http://www.littleworkshop.fr/keepout.html.
<<the web can do pretty much anything you need with content and data>>
Not on mobile. Due to serious, oft-encountered performance issues, the web remains crippled on mobile compared to native.
By "native" I'm referring to apps where the UI is rendered natively as opposed to being rendered in a `WebView`.
React Native is bending the rules here but it is important to note that it is divergent from the web platform and convergent with the native platforms.
The issue I have, and I think others have, is that 99% of "apps" don't need the extra capabilities and performance. Or if they do, it's only because of problems they've created themselves. "We're loading 45 tracking and advertising scripts, and now our site is slow. We need an app!"
The article here demonstrates the problem perfectly. Mobile browsers are absolutely capable of displaying the text articles on the site. The browser on my flip phone from 1999 was capable of browsing a site like that.
I'd add that if you intend to maintain a web site then adding support for apps (presumably multiple since cross-platform is still hard) then you are throttling your bandwidth to actually improve features and stability. I'm thinking of certain food delivery app that had enough bugs on their Android app, then I fell back to their mobile site and haven't looked back. They could have saved themselves probably a few million bucks and just kept their mobile site looking spiffy.
I feel exactly this way about a certain image sharing site. After numerous bad experiences with their app, I went back to their mobile website and found it fully functional. For some reason, copy and paste didn't work in the app, which is unacceptable in a community based on copying and pasting cat image URLs.
Well, the most important thing you gain from moving your apps to the web is owning the relationship with your users. And that is HUGE.
With the Apple iOS App Store you are isolated from your users, it is very difficult and nearly impossible to build a list and keep in touch with them in a meaningful way. I am sure there are cases where this isn't true, but I think I can say this is probably the case for the vast majority of apps out there.
By being on the web it is your business, your platform and your users. Not Apple's (or whoever).
And, yes, depending on your niche, building an audience could be very difficult. Guess what? The vast majority of apps on the app store fail. I think we might actually be able to say success on the web might actually be easier.
As a consumer I can relate to this. I subscribe to many monthly magazines including GQ and Esquire. I enjoy the publication as it is meant to be viewed, ads and all. Of course, now they're piled up and so I want to switch to digital. Sounds like a great solution.
GQ has a digital app or I can use iOS news stand. This was not a great experience as each issue took at least 20 minutes to download a gig of content. Of course, once the content is there the experience is rich and fantastic.
Based off my experiences, I really think a mobile web experience is best for reading magazines. Only the page I'm interested in would load and I shouldn't have to take up a whole gig of precious space to read an issue.
As an Android user I'm very fond of native apps as long as they integrate into the OS well via activities, intents, content provider interfaces etc, all stuff you can't really do with web apps (though I know someone is going to want to chime and and talk about WebIntents or whatever other thing nobody actually uses because it just doesn't deliver the same sort of flexibility as what I'm talking about).
So I'm sticking with native apps where it makes sense.
Having said that, for the type of site this actual post is discussing, eh... web is probably just fine. Go ahead and use it... I have no issue with the linked article, but some of the people discussing this here and other places seem to be nudging this into a discussion about "mobile web is ready to replace all native apps", which IMO is a ridiculous statement.
But isn't this an "App" which shouldn't have been an App in the first place (I think they write something like this in the article).
That's really one of my pet peeves. I don't like installing apps just to read an article. A good mobile site is much preferable in my opinion. But I've started a little bit later than most to the smart phone game and I'm a little bit quirky - so that may be just me.
If only there was a single native app, preferably pre-installed on the phone, that was specifically designed to download content from publishers around the world and render it dynamically on a range of device screen sizes. It could be publisher neutral and allow bookmarking, and maybe let you have multiple pages open at a time. Nah, it'd never catch on. Native apps for publishing are dead ;)
Exactly, and that's why it makes that a publisher would use the web for its original purpose-- the transfer of documents. Especially now with the News app on iOS, I don't see many reasons why a publisher would need to go to the trouble of maintaining its own suite of apps.
I suspect the reason many think they prefer apps is the one-click launch and the automated bookmark created on your screen.
If a mobile browser behaved like an appstore (Search for website, See results, Click to Install, i.e. bookmark), you would have the same behavior, downloads, and most people would not be able to tell whether they are using an app, web app, or website.
Firefox OS has something like this. On Firefox Marketplace, you can submit a mobile website that serves a small manifest.json file and it will be indistinguishable from a native app. I've been praising this innovative approach so many times on HN, but I have to do it again---it elegantly solves a lot of the problems we're having with native apps, and it all mostly works and integrates fine with the OS (the idea could use some polishing but is good as a proof-of-concept). It's a shame that the OS itself never really took off.
Unfortunately, the two largest mobile platforms (iOS and Android) don't have native support for these specs yet. And I'm not sure if they ever will properly support them since they'd probably prefer to keep the free developer lock-in they get with their native app platforms.
In the mean time though, you can use projects like Manifold JS to create Cordova-based polyfills for those platforms:
I don't use android, but I remember hearing about it trying to blur the lines between apps & browser tabs. They might be more amenable to the idea of browser apps considering web advertising is their cash cow too IIRC
Chrome for Android has a feature where if you visit the same site a few times in a two week period (or some sort of heuristic like that), the site can have it prompt you to add it to your homescreen:
This is great! Mobile apps have taken our age back to the age where everything is platform dependent, the Internet was invented so that devices from multiple manufacturers could communicate with each other, but with Android and iOS we have created a new niche of platforms, now it is devices with specific Operating Systems, and they are pushing our age back!
There is a trend in Indian startups to abandon the web altogether and go "app only", among other reasons it defeats the entire purpose of the Internet as we know it, kudos Atavist!
It looks like the folks at Atavist did the right thing. As far as I can tell they created a publishing platform that lets folks format thousands of words of text and photos and videos, which is a very good use case for a web browser. Plus, contributors can add links, italics, subheds, captions, etc. This is not the best use case for mobile apps.
They have a beautiful digital magazine that doesn't need to live in a native app -- and had no compelling reason to.
When we started building our smart news app ( more info at https://recent.io/ ) we considered creating a non-native web app and took a few steps in that direction. But we rely on touch gestures a lot, require no keyboard input except for searching, update news recommendations minute-by-minute, and wanted to appear in the app stores, so we went down the iOS and Android paths instead.
the article does not mention this but publishing an app for iOS is drastically more costly in time and efforts than publishing for android.
It is the worst publishing process I have ever seen. sluggish certificates, keys, provisionning profiles, iOS lack of retro-compatibility, apple lack of communication on what the major versions are changing and apple lack of judgement when reviewing apps with guidelines that are changing every day.
Frankly I don't even understand how can developers enjoy working on developing apps for iOS it is lackluster in every part.
What do you mean by lack of retro compatibility? You can still ship apps that target ios6, you have runtime checks for APIs that are missing in one version, the IDE suggests what is available or not.
The major versions (and minors) are thoroughly documented in API diffs, the IDE suggests what to modify according to your platform target.
The review guidelines are basically changing for the major versions, accordingly to the new tech introduced, and most of them are dictated by common sense.
If something goes awry the review team is available to discuss with you. Good luck getting an Android app reinstated if it is taken down.
The certificates are a pain in the ass, no way around that, but the rest of the platform makes up for this, the tools and language(s) are way easier to use.
I have an iPhone 3G in the drawer, as an emergency phone.
The amount of apps, that this phone can download from App Store, whether running iOS 3 or 4, is exactly zero (yes, even those that were available in the past, cannot be downloaded).
My brother has for exact the same purpose the original HTC G1, running Android 1.6. That phone can download every single app from the Play Store, that was made available when the phone was supported.
If you're using it as an emergency phone, do you really need to install apps? Apple has taken the position that the older phones are not worth the trouble it takes to support them. The numbers seem to bear out this decision -- most in-market iPhones seem to be 4S and newer, and people upgrade their iPhones almost immediately when the new OS is released (this I can't understand at all, the new OSes frequently break things, and I always wait a few months to update).
Background: I needed to sync the contacts with Google. Meanwhile, Google disabled the ActiveSync support, the phone was reset, and I was looking for some app to do that. In the past, there were apps capable of doing so, but now, they were unavailable. Add to that inability to side-load the apps..
Regarding the retro-compatibility:
Nobody supports Android 1.x either. However, in the past, developers did upload their apps that did work on these devices. So why the users shouldn't be able to download these old versions?
And that's the difference. Apple actively removed old packages from the store, while Google just left them there. Just this alone makes the old android devices more useful.
The parent is probably too young to remember a time when you could buy a device expecting to keep it for the rest of your life, or maybe even pass on to your children.
First of all, good for atavist. They recognize their market, and are following the path of greatest efficiency and effectiveness. Sounds like great business sense to me.
Now, on the other hand, I have to say, this kind of thing always bums me out. For a news distribution service, it does not matter so much I suspect, but in-general, the trend of putting more and more of the functionality I rely on daily into a web browser is distressing. Web browsers, despite all the innovation and work that has gone on lately, are still an enormous attack surface with the most significant likelihood of security breach (especially given that the rise of javascript has led us all to the point where we are almost allowing arbitrary code execution from untrusted sources on every page visit). Not to mention that web browsers were founded on the notion of conveying content, not functionality; HTML is a markup language (with <canvas> and JS and HTML frameworks, that's not so concrete anymore), it was not meant to host alternatives to the power of native applications.
I often find myself wishing for the ability to use a text-based web browser and have "The Web" still work, but those days are mostly long-gone[1] till I finally get around to writing my own browser[2] (which seems like it should not be a necessary step just to be able to interact with a simpler Web).
I suppose I am just tilting at windmills at this point (and, as I said, this post is not really aimed at atavist since I think content-delivery is a perfect niche for being web-only). Ah well, a user can dream, I suppose…
Most apps are delivering content, so they also fail when offline.
Yes, there are a number of apps that cache some data locally, but I believe they're in the minority. Even casual games have to check in with a database and retrieve a couple datapoints about your progress, your in-app purchases, and what powerups are available to you.
It's becoming increasingly clear that the future of publishing—for content providers—will be to relinquish control and publish their content directly to third-party distribution and engagement platforms.
The thing about potentially having to wait two weeks to make an updated app available, and the control over your business that implies, makes my testicles suck up into my torso.
It's sad it's taken people so long to realize it, but at least they're starting to get a clue. Who'd have thought that an app that's essentially a web browser tied to a single site was a bad idea that nobody would want to use?
If desktop and laptop users access your content/game/social network/whatever from a web page, it's probably best to have mobile users access it that way, too.
And I don't buy their justification for going down the app route in the first place: "But by the middle of 2011, six months after we launched in our app, more than half of installed web browsers were still not HTML5 compliant."
Who cares? The app only targets iOS, so as long as mobile Safari rendered it, they were fine. And most of us here know that the app would be a pretty thin wrapper round a webview.
I suspect the real reason for going down the app route is because back in 2011 all the marketing types were screaming "WE MUST HAVE AN APP!!!1!!"
News International made the same mistake with "The Daily", which (probably) cost them over $60M.
Could have used React and tried to share component source between iOS, Android and Web.
I mean since you're going to fully go web, I don't see the problem in this. It's all JavaScript anyway, unless you want to start playing with the hardware APIs.
Yeah you would have to write some of the UI layer differently, but I think it isn't that much. You can wrap your own UI components to use Android/iOS/Web internally.
> as we pushed the design envelope further on our stories, we were constantly running into the technical limitations of the app approval process. While we could develop, deploy, and—if necessary—repair new code instantly on the web, any change to our apps required re-submitting it to the app store in question.
It's possilbe that if the design changes are only on the javascript side that the app could just download those instead of them having to be sent for review.
The biggest problem with native apps is that there are a lot of native mobile developers. One in 20 or so are actually good.
So you have all these apps that were built by someone who has no idea on how to architect them. The company has already spent so much just getting it out the door that they refuse to rewrite it.
Native mobile dev is great. You just need to have good mobile developers run your team.
Interesting that a smallish magazine is moving away from native mobile back to web while not too long ago I saw that some large retail company discontinued their mobile web site and require use of their native app (flipkart I think).
I don't think expecting a new audience just because you have an app is reasonable. I'd be interested in seeing how many new users they generate with just the mobile web site - I suspect that ultra loyal existing readers may take the time to bookmark their site and come back, but would be very surprised if their overall usage doesn't decline.
What's interesting about the Flipkart example is that they have a fully-functional website available which they serve to desktop browsers, but the website is blocked to mobile browsers. Clearly they have decided that they prefer their mobile users to use their app, not a website.
Sencha does this for enterprise apps too. For publishing, I wish RSS feeds had taken off. Would be great to visit a site and see that I could just subscribe to their RSS feed and see it in my own reader app.
The loss of RSS just still blows my mind. I still get my podcast with it but sad to see RSS to be relegated to obscurity when it started out so strong. The issue was the implementing it was "too much trouble" with the public.
Good. Using the app mechanism to deliver static pages was silly and inefficient. The only reason it caught on was because it came with a payment system.
Native apps are just a momentary crutch until Web standards on devices get better. Especially since most so-called apps are merely glorified Web pages in custom browsers.
As a web vs. app proponent I love this. I do have a few apps installed on my Note 4 but I prefer getting Twitter, FB, etc. as a web site.
Web standards like HTML5 are wonderful so let's use them and make the web awesome.
My brother and one friend each have a few hundred apps installed on their phones. I am sometimes surprised that their phones still function with so much cruft installed, never mind the privacy considerations.