Is The Web Dying? | Hacker News Reader

Is The Web Dying?

medium.com

56 pointsamasad67 comments

The article misses one point: people were playing games on their desktop machines even before smartphones were the norm. Just because people spend 20% of time on mobile devices "browsing" doesn't mean the web is dying.

The time-spent statistic fails when you think of what all users were doing on their desktops other than browsing. The open web was not killed by Microsoft Word, and it won't be killed by Farmville either.

Exactly. And 18% of the time is spent on Facebook. Does it really matter if Facebook users do it via the web or an app?

That depends on your definition of "Web". If you mean the proliferation of watering holes (including this one), then the old web is dying. I've seen a few new nuggets of gold on NeoCities, but people's proclivities toward building "websites" seems to be on the wane.

I blame search engines for this as much as anything. Back in the days of the webring, the only guaranteed way you could come across a site that's relevant to what you were browsing for was to see whether it was linked on the current site.

I.E. "Surfing" actually meant something.

The web is not an application.

Whoa, I'd almost completely forgotten about "webrings". These days, links and targeted advertising play something of a similar role, but not quite the same thing...

They're not completely gone. I think "Blogroll" filled that void for a bit, before that too was taken over by ads. The next step was public bookmark sharing, like Stumbleupon or Delicious. Still, those services feel a bit sterile for my taste.

Something nostalgic and hard to replace like a hand crafted HTML link on your own site carries a bit more weight with me.

I loved this feeling that I got when I found a web-page with good content that was hard to find. It was like opening a treasure trove. There were also a lot of different (rough) designs with personality making the experience much richer.

Now information is per page, accessible from Google. And everyone is mimicking other people's design. More and more people are consuming than creating and it changes the ecosystem. I would love to see more personal wikies that have traesures in them.

They're still out there, albeit harder to find. Geocities dying was quite a blow and Tripod is scarcely touched these days. Squatters, I fear, are moving to NeoCities, but here and there I find some sites that are diamonds in the rough. Individually hosted sites are rarer still.

I.E.

http://www.nakka-rocketry.net

http://www.mcleanmonocycle.com

http://www.neogentronyx.com

http://rimstar.org/sdprop

http://www.matus1976.com

http://www.restrainingbolt.com

I tend to browse forums (barely have the time these days) where these links are shared. Once in a while Google will get these pages or you might see a Wiki page created for them, but for the most part, they're really spread through manual sharing.

The proliferation of mobile makes web traffic go up, not down. Mobile is a bigger slice of the pie, but the entire pie is bigger. Desktop web traffic itself appears to be growing, not shrinking.

I've had to correct the distortion of these numbers before (https://news.ycombinator.com/item?id=5847337), and I'll do it again. Using Wikimedia as an example, as mobile accounts for a bigger share, overall traffic is growing significantly, and even desktop traffic is growing, not declining:

Wikimedia June 2012: 149,085 M requests - http://stats.wikimedia.org/archive/squid_reports/2012-06/Squ...

Wikimedia June 2013: 223,725 M requests - http://stats.wikimedia.org/archive/squid_reports/2013-06/Squ...

As you can see, Windows declined in share from 70.45% to 56.10%, but total number of Windows desktop requests grew from 105,035M to 125,520M. Similar pattern with Mac Desktop.

Mobile means more web traffic, not less, and desktop web traffic is also growing.

You make a good point. I think it's also important to keep in mind that mobile may be growing but it's also becoming increasingly difficult to build a software business on mobile. The top 20 apps on both leading platforms are gimmicky IAP games and users have been trained to complain about paying more than $2 for anything.

If I were starting a new software business today it's very likely I'd completely ignore mobile and focus on the more lucrative desktop/business market.

> users have been trained

I'm not sure about that, users are just cheap in general. That's why sales and discount mega stores are so successful.

Speaking for myself, the more apps I buy, the more I'm willing to spend on a well crafted or useful piece of software with the expectation that gimmicky ads will not be there and the in app purchases actually do add value on top of the purchase.

Two apps in the $8-$20 range that I regularly use right now are Gaia GPS & Paper by 53 (although I'm slowly swapping that out for Procreate).

The web application may be dying, but that is in large part because the technology was never well suited to the task in the first place.

When it comes to what the web does well, a collection of hyperlinked documents full of information, mobile devices are not necessarily how people want to consume that kind of data. Despite having a home full of mobile devices, I still reach for the computer when I want to dive deep into a subject.

I see the web as a collection of information and information processors (applications) and don't see how can one exist without the other.

An application has a more narrow definition than an information processor. An application must serve the user.

For instance, if you wrote a piece of software that converts LaTeX documents into HTML upon request, I would not consider that a web application. As I take it, it would meet the definition of an information processor, however.

If you built a full accounting system that presents its user interface in a web browser, I would consider that a web application. The scope is quite different.

That chart is bogus. There are way more than 2 million desktops in the world. Did you mean billion?

Moreover, most of the apps on your phone are native front-ends for web-facing APIs:

- Facebook

- RSS readers (Pulse, Flipboard)

- E-mail

- Chat

If you think of the web as strictly HTML documents, then yes, usage patterns may be shifting towards native. But, the core-concept of the web is really a global network of interoperable machines that anyone can enhance. It's a world-wide watering hole. If you take this broader view of the web, it's thriving more than ever before.

It's a bit alarmist to imply that this network is dying because HTML may be displaced as the primary interface.

(Incidentally, the biggest reason this is happening is because Apple's interest is in making proprietary apps awesome and WebKit good-enough. Mozilla, Google, and many other community players are working to bring native-quality APIs to the web to preserve its platform-agnostic nature. As long as Apple dominates developer mindshare, they'll have an uphill battle.)

Yes it's meant to be billion. I'll update the chart. I didn't invent this, it comes from KPCB[1]: http://www.kpcb.com/insights/2012-internet-trends-update

[1] http://www.kpcb.com/

Thanks.

The power of the web is that there's a single name for every resource in the world, and anyone can link to it.

Mobile apps break the second part, which seems inconsequential to those that grew up in it, but it's a large part of why the web has become as ubiquitous as it is, why the web has become as useful as it is, and why the web has the freedom to create is it has become.

Last para is Total FUD.

Why do people who have no actual knowledge feel inclined to comment on things they have no idea about?

Have you tried android webview and ios uiwebview? We'll let me tell you - android webview hasn't been updated for 3 years and has no features that you care about. Websockets xhr2, a fast canvas and I could go on. None of this is there. In fact googles webview is the main reason people are writing native apps. Do not get misled by chrome. Chrome and webview are totally different. Uiwebview is actually quite perform ant and modern. The only thing is it has no js jit but you cannot call them out on that when android webview is a joke.

Seriously try it before spreading FUD.

It's true. You'd expect, given Apple's emphasis on native apps and the importance of tie-in to its ecosystem, that they'd do everything they can to handicap both Safari and UIWebView. But in fact they're both superior to anything Google's providing with the possible exception of very bleeding edge Chrome Beta.

Google's neglect of Android's WebView doesn't really jibe with their claim of pushing the web forward.

Also, remember that in the early days of iOS, Apple was pushing developers toward web apps. Support for native development was added in response to developer demand, not because Apple wanted to lock in developers.

It's what I've been told by numerous non-Apple (and/or former-Apple) WebKit contributors.

Now you know better :)

Also, don't take my word for it, try it out yourself.

The web changes all the time and the market catches up. In the early 1990s, the web was about linked information. Today we have Wikipedia and Google. In the late 1990s, the web became about ecommerce. Today we have Amazon. In the early 2000s, the web became about social interaction. Today we have Facebook and Twitter. In the late 2000s, the web became an app platform (HTML5 + cloud computing). Today we have Google and Apple. It gets bigger every time.

Maybe the browser and HTTP are less visible nowadays, because they are more tightly integrated into the UX, but they have never been more important, and it's still only day one.

couldn't agree more.

It's just a cycle. First came the "native" desktop, then came the "web" for the desktop, and it started replacing most "native" use cases.

Now comes the "native" mobile, and eventually the web on mobile will take over, too.

Yes or no. Google is pushing the mobile web because it can index/rank web pages (PageRank work well with web pages and the links between them). It is much harder to deep link between native apps, hence Google cannot search in-app content. Without being able to rank in-app pages and even whole apps, Google cannot make money.

The other key question is access to the device native capabilities from java script. Google can at max, assure access to android OS API from java script, but even than the device hardware manf (e.g. Samsung) can create additional hardware capabilities which can be accessed only by native apps.

So to sum up, the whole "mobile web app" vision can only occur if Google could rank mobile app pages and show you native apps (or a link to the original web site) in the search results.

The question that popped in my mind reading this is: should apprentice developers do away with learning web app development and focus on iOS/Android development? What if new graduates go through hoops learning how to develop web apps only to find a market that only wants mobile app developers? I'm not worried much for server side development, that's the same for both types of apps, but for client-side developers and designers that might actually find themselves with a very shrinked market 5 years from now.

Web traffic, including desktop web traffic, is growing, not shrinking.

A growing share of mobile devices in addition to the growing desktop web traffic just means more opportunities for everyone.

One of the biggest criticisms of Microsoft missing the Web wave was treating the Internet as the set of protocols with real future belonging to apps. Microsoft (via Bill Gates' vision described in "The Road Ahead") believed that "fat clients" were the future. But they shipped a browser for all miscellaneous Web needs just in case.

It took, ironically, Apple to really bring idea into fruition. Yes, Safari icon exists and allows you to navigate some of them "sites", but why bother, when you can have a full-featured fat-client native-app experience with all bells and whistles.

In enterprise software land, now everyone brings their own device, web applications are more important than ever.

A company selling SaaS might hope to support HTML, Android, and IOS, but not Windows Phone / 8 / 8 RT, nor Mozilla Phone, and oh wow, different form factors for each device.

So if you are a bank, and you want to write an expense program for your field people, and don't buy a SaaS because politics, you will be writing in in HTML. This won't change for a decade.

The PC revolution is also on the skids in a more general sense -- feudal models of security, centralization of networks, etc.

There are many reasons for this, but one that sticks out to me as relating directly to this particular article is usability and to a related and lesser extent design.

Usability and design rule the market.

This is why Linux on the desktop is basically dead outside of a few enthusiasts and pilot projects, and why the PC is dying in favor of (feudalized) mobile platforms. These platforms, while less free and open, offer a vastly superior user interaction experience.

The paramount importance of usability and design is something that open source and open platform geeks have never grasped because they're the sorts of people who don't mind being confronted with a "# " prompt and don't mind fiddling with computers to get them to work. It's like asking car racing and tuning enthusiasts to design a car for the general public. They'd be likely to include all kinds of nerd knobs and weird features (that's the NOX knob, yo!) that would confuse or endanger the average driver.

Yes.

Mobile apps offer a superior experience. Reading on mobile is better (e-readers) and that was the main thing that pushed the web forward. Apps are also superior, not just the speed/polish, web will get that too, but because of the limited screen space, apps are forced to become more utilitarian, and hence better for the user.

Here's a talk explaining this http://www.youtube.com/watch?v=NjE_Or4VIlU

Mobile, expecially tablets, bring back many of the superior experiences, like books, 90s style cd-rom apps, physical newspapers, that the browser pushed aside for many people. 'Long form' longer attention span activities will predominate.

The losers in this - aggregators such as search engines (google - replaced by wikipedia app etc.) and private social networks (facebook - replaced by peer to peer mobile apps)

the winners - creators - writers, artists, productiviy software developers, and game developers

A new type of app - hybrid between book-documentary-game-todolist, will become particularly important (a gamification of work).

In time, mobile devices will reach the heights of current desktop computers. While held back by the batteries, it seems inevitable to me that in 3-5 years, the performance difference between the web & native mobile won't be distinguishable (except for games!). After that (or before), it seems natural that a popular or up an coming OS will open up its native APIs to web sites, if the user wishes to. It might be Firefox OS, and Ubuntu OS. Whichever OS does it, will receive a huge market share boost. That's my vision.

> it seems inevitable to me that in 3-5 years, the performance difference between the web & native mobile won't be distinguishable

If native stood still that would be true. The thing is that non-native is always a few layers above native. If non native gets fast native gets faster. If non native becomes energy efficient native becomes even more energy efficient. etc.

That's true, but as both get faster the difference between them becomes less and less significant. The layers between non-native and native impose a relatively fixed amount of overhead, which uses a smaller and smaller percentage of overall performance as devices get faster. that's why virtual machines, for both full environments and for the runtime layer of high level languages, are so commonplace now while a decade ago they were a serious disadvantage compared to non-vm approaches.

The consumer web might be dying, but the b2b web isn't.

We are a far cry from me being able to properly program or design anything without at least a laptop size computer.

Exactly. B2B also happens to be where most of the money is in software development, unless you can manage to hit a home run in the consumer space.

It's not dying. It's just getting up from the desk and walking around.

Interesting way of looking at it :)

I'm beginning to think that smartphones have as much to do with the general computing paradigm created by PCs as PCs had to do with the general video consumption paradigm created by TVs. Yes, there is overlap, but it is not the ideal platform and the new platform does things that are unique to it. I know a lot of people who replaced their TV and DVD players with a computer, but I know a lot more that only replaced their cable TV service but still pump the media out to their TV.

I think we're still figuring out what the hell smartphones are good for. When the Web first came out, it was personal websites versus TV shows. We now understand those two things to he independent, if not complimentary, things.

The TV is for large scale consumption, the PC is for large scale content creation, the smartphone is for large scale content aggregation. These can be seamlessly complimentary tasks, if we can get them out of the hands of vertical corporate structures who are only interested in using the concept to build ubiquitous brand exposure over value.

I feel like I need to do something about this. I see the marginalization of small sites and niche sites (self hosted empires) and the growth of large massively huge systems like facebook. The web is becoming a tool of multinational corporations and less a community of people. This makes me really sad and I don't know what to do. I feel like I need to something.

I think this article was very Very enlightening about what is going on with the "Web" at large. It's refreshing to see some numbers backing up what is going on. I work as an amateurish FOSS developer and a network engineer by day. These numbers tell me something i already knew: the desktop is now longer the primary way to browse the web. I have a smarty-phone and occasionally use it to look things up, but the "normal" web on mobile is atrocious. Looks like the days of ads and pop-ups filling the margins of pages might be coming to an end. Maybe we can start catering to the users looking for information instead of trying to work everyone over for advertising revenue.

My site browses on mobile just fine, and it's fully un-optimized wordpress.

I honestly don't see the difference, if you consider it from the the consumer point of view. The web is alive and thriving, the only thing that has changed is a smaller screen size and touch screen interaction. Weather its a screen on a desk or held in your hand is irrelevant.

I guess a better title would be "Is the web browser dying". It seems to me like the web is transforming away from the browser into standalone applications that use the underlying architecture. Sure it's still HTTP requests going back and forth from the Facebook app but it is different from the old way of building a website.

I don't think any major service has launched in a long time without a dedicated app for iPhone and Android as their main product, the browser experience is just too clunky on mobile.

It would be interesting to see statistics on desktop vs mobile in terms of actual hours used. It's probably to be expected that mobile devices will outsell PCs as they tend to be cheaper and people get a "free" smartphone every couple of years as part of their plan anyway.

OTOH I spend 8+ hours a day in front of a PC generally and significantly less time browsing on a phone unless I am in transit or something.

Cinema didn't kill theater, TV didn't kill cinema, the web didn't kill tv & so on. Some parts are overlapping, some respond to different needs. I think instead of asking, "is this & that dying" why not ask what space is occupying over the other, how's it different, how is it the same. The discussion is to broad as is and leads easily to religious arguments

If you can show declines in internet usage, number of connections and number of installed web servers then I'd yes, the web is in decline. I also don't see how native vs. html5 on mobile has anything do with the argument. On another topic, I note that I'm consistently disappointed by articles on medium.com

The internet is not the web. Many native apps use the internet, precious few use the web.

Quick litmus test: if a system does not have URLs, or the content at those URLs is not usable by multiple independently developed user agents based on open standards, then the system in question is not the web.

Aren't native apps just specialized web browsers with the url encapsulated in the settings and not visible? I'm not aware of a shift away from HTTP to some other private protocol. If the billions of smart devices coming online aren't universal web browsers does it matter? The growth of public facing hyper text documents may be in decline, but I'm not sure that's a bad thing. The internet is evolving towards more specialization.

In the same way that a lion is kinda like a horse. Serving private apis out over HTTP is an implementation detail for a native app. If it becomes convenient, they could switch to a different protocol for their RPCs without anyone noticing. That's not the web in any reasonable sense of the world.

Without links, it's just not the web. Archive.org can't preserve it, search engines can't search it, you can't reference it effectively in a bug tracker or a blog post or a hacker news thread, or a footnote in an essay, etc.

The things you list would seem to be severe disadvantages. I doubt many would go this route exclusively but have a traditional site in parallel. I guess Google and other search engines will let us know when the number of pages turned up by indexing starts to decline or let us know when general web browser installation starts to fall. I don't see it happening.

Last time I checked you still need a data plan to download and load your favorite apps music videos emails etc... Is The web changing? The answer is yes. Is the web Dying. The answer is a vibrant no. We are more connected than never. It is an interesting era for software developers with creativity....

No it is not dying.

The web is much more than browsers, which should never have been more than a way to display documents online.

It is everything than can be reached by network protocols, IP, TCP, UDP, HTTP(S), SMTP, IM, Jabber, you name it.

The web refers to HTTP and HTML. When you say web, and especially when you refer to IP, TCP, UDP, SMTP, etc, I think you more accurately refer to the internet.

Some of us already referred to the internet as web, before mosaic came to life.

You may be right, though I was around then and don't remember anyone referring to it as the web (information superhighway does come to mind though; ugh). In any case, as used in the post (and colloquial usage), I'd still stand by my comment.

Of course not (in the dying sense). It is a common mistake to conflate 'change' with 'expiration.' The network experiences are changing, we're not quite in the 'post web' era but I can see where we will be here in the not too distant future.

The phrase "you can't go back home" captures this notion of change. Leonard Nimoy used it on a talk he gave about Star Trek (which was before the first film :-) Things change, moments in time are lost forever. Programmers are often acutely aware of this when they try to re-create a product from a few years ago and its impossible to get all of the pieces back to a consistent state from the origin time of creation.

Chrome at over 40% marketshare? Everything else I've read puts it around the same level as FireFox, both browsers somewhere between 17 and 25% depending on the source.

> Chrome at over 40% marketshare?

43% of web traffic, apparently, though I haven't yet tracked-down the exact source of this story:

http://www.economist.com/news/business/21583288-what-googles...

Yeah, the benchmark listed is pageviews, not unique browsers. It's not hard to guess that users of certain browsers have dramatically different browsing habits.

This looks like the law of the eternal-return of nietzsche. welcomw back to the nineties, only its 2013 and "computers" are smaller.. back to te future..

No.

Once again Betteridge's law proves true.

Case in point..I am viewing this forum using hackbot android app. I prefer this emensly to the browser.

Do URL's matter? If they do then the web is not dying.

no it is not...the new technology and platform is opening the new way for the web

ENYO

Reply on news.ycombinator.com