29 ms·
Web vs. native: let’s concede defeat
- hmans 11y agoOk
- nailer 11y agoThe headline is a little misleading: the article isn't about not using web technologies in favour of native apps - Instagram, for example, uses HTML5 for most of it's UI (you used to be able to see the error pages) and has great UX - but rather a call to stop making 'native app like web sites'. > Our working assumption has been that all web sites should be app-like, and therefore tooled up to the hilt. I don't think that's true. Some people create horrible scroll jacking functionality for whatever reason - maybe they thought it was more 'native app like', maybe they just thought it was clever. But many other web developers already consider that bad perf is bad perf and aren't willing to put up with it.
- datashovel 11y agoThis post leads me to believe the author probably assumes there will always only be a handful of native platforms. I think, if we're lucky, the ecosystem will be far more diverse. In the future I envision, noone in their right mind would ever think native is the way to go.
- pdpi 11y agoAnswer me this: What is the point of a plethora of platforms, if nobody in their right mind develops native?
- jevgeni 11y agoThis was just quick and brutal. :D
- datashovel 11y agoIt seems we have different assumptions we're basing our arguments on. I get the impression that the huge diversity of linux distros is enabled primarily by the fact that more and more development these days is web and not native.
- pdpi 11y agoYou're missing my point. There's no purpose in having a huge diversity of linux distros, if they're just glorified browser shells.
- datashovel 11y agoI'm guessing they all (the linux distros), in one way or another, add value given each project's mission statement. Or at least they believe they are. Otherwise they probably wouldn't be doing it. And I don't think we'd have to fear the OS will become a "glorified browser shell" if it wasn't the primary "app development platform". Right now native app development is feasible. You have pretty much 3 platforms you really need to worry about developing for in order to have access to the largest amount of market share for the least amount of effort. Imagine even one more player enters the market and gains any respectable amount of market share. Instantly every mobile app development team's efforts need to grow or every developer on each team's responsibilities grow linearly with the total number of platforms they need to develop for. From an operational standpoint it's not sound economics to make business decisions based on approaches that don't scale well.
- jevgeni 11y agoThere isn't any great diversity in Linux distros interface-wise. You basically have a handful of distros that use X Windows and one of the four old window managers.
- datashovel 11y agoI appreciate that. I agree it's not obvious what, if any differences exist between alot of the lesser-known distros. However I would argue that a large enough sample from the most popular Linux Distros has something substantively unique to offer. But we haven't even considered BSD-flavor. Perhaps if we expanded to include BSD-based operating systems. To name a few: FreeBSD, DragonFly BSD, PcBSD, NetBSD, OpenBSD. All of the BSDs mentioned above have a very unique mission. It's easier for me (as consumer) to choose any one of those to run my desktop / laptop knowing that I'm going to pretty much get all the same access to apps that I would if I were to run Windows or OS X. Without any reassurance that I'm going to be able to do what I need to do on a BSD or Linux machine, I'm probably going to be stuck choosing something created and maintained by Microsoft or Apple.
- swah 11y agoWell, the cross platform support will get really good ;)
- ubernostrum 11y agoThis post leads me to believe the author probably assumes there will always only be a handful of native platforms. No offense, but I'd trust PPK's beliefs about the mobile ecosystem more than any random HN user.
- datashovel 11y agoNone taken. It's probably not worth arguing who is right or wrong since none of us can reasonably assume we know what the future will be, but I will say that in general (1) diversity in ecosystems is good. (2) Monopolies are bad. Given only these two basic assumptions I think it's obvious that there should / will be a push for more diversity in mobile, and thus a larger number of platforms to develop for. This can only lead to one result which is for developers to want a unifying platform on which to develop. I would argue that this exists, and is improving. It's called the web and web browsers enable this unifying platform.
- ascagnel_ 11y ago> It's probably not worth arguing who is right or wrong since none of us can reasonably assume we know what the future will be, but I will say that in general (1) diversity in ecosystems is good. (2) Monopolies are bad. Diversity is good, but too much diversity can be detrimental. A one-mobile-OS world is a terrible thing, since there's no need to push forward and improve. A two-mobile-OS world is better, as long as it's easy for users to switch to whatever is better at the current time. Once you get beyond three, you end up with a situation where it takes more and more development time to cover less and less of the market, and it becomes more likely that a single winner emerges (for example, the DOS-to-GUI transition in the 1980s-1990s).
- datashovel 11y agoIn my crazy future I hope there will eventually be 100's of legitimate mobile OSes. One of the reasons this would be possible is because the responsibility of app development theoretically wouldn't be aligned vertically with the company that developed the OS, or with some platform that company is responsible for developing.
- onion2k 11y agoUnless you're making something that needs really good performance (3D gaming for example) building a hybrid app wins every time. Something based on Cordova (eg PhoneGap, Ionic) with Crosswalk can get you to about 90% of the performance of a native app, with about 1/2 of the development time because it's all built with web technology. 90% is easily good enough for most applications. Cordova wraps a uiWebView in some boilerplate to make a web app effectively native to give you the 'icon on the homescreen' advantages (selling the app, in-app purchases, access to most phone APIs) with all the benefits of building for the web (easy development, lots of developers, etc).
- mahouse 11y agoSeeing you saying good things about Cordova makes me think you have never used Cordova.
- aliakhtar 11y agoI have, and I agree with what he said. One thing which isn't that great, are the plugins. I feel like there could be easier ways to access / write plugins. But overall, everything said was pretty spot on.
- frou_dh 11y agoIs a "win" for the developer necessarily a win for the users? Whose experience takes precedence?
- wolf550e 11y agoIf cost of developing app is halved, many more apps will be built. The choice is often not between "expensive native app" and "cheap hybrid app", it's between "why is it so expensive? but I can afford this (hybrid app)" and "no way that's too expensive (native app)".
- dottrap 11y agoIn theory, this might be true. In practice, I find that it is just an excuse to cut corners and ignore the problems. I recently had to help somebody with their public library and a particularly magazine publisher for digital issues. They require a web app to read the issues. They were trying to use their Mac, but the site refused to load. The public library staff had no clue why it didn't work ("well, it works for us", and their attempts to contact the publisher was your stereotypical tech support horror story.) Long story short, their website doesn't work on all browsers. They implicitly know that it doesn't work in Safari because they apparently wrote a native app just for iOS. However, they have no native app for Mac and never bothered to fix their website.
- Navarr 11y ago> Native apps talk directly to the operating system, while web apps talk to the browser, which talks to the OS. Thus there’s an extra layer web apps have to pass, and that makes them slightly slower and coarser than native apps. This problem is unsolvable. But what if the browser is the OS? I agree with the point though that we shouldn't be trying to emulate native to the T with web applications, that we shouldn't be stuffing them full of JavaScript. I think the Google IO 2015 web app [1] is a good example of what can be done on the web right now with minimal overhead. 1: https://events.google.com/io2015/ https://events.google.com/io2015/
- EliRivers 11y agoBut what if the browser is the OS? Then the user has chosen, for whatever good reasons, to run a pretty limited OS, and in doing so has chosen not to be able to take advantage of all the things that having a much less limited OS would offer.
- Navarr 11y ago> "run a pretty limited OS" This is only true right now and only if browsers were not working to make more low level stuff available. There is very little limiting about being in a browser these days. Drivers, perhaps. Kernel stuff, perhaps. But as far as a game is concerned? Or an email application? Or anything in the context of this discussion? Not too much limiting them.
- EliRivers 11y agoThis is only true right now It will always be true forever, or at least for as long as a web browser is a program run within an actual operating system. It's not a competition. Pick the right tools for the job. If the job is getting your eMail or playing a relatively simple game, and a web browser abstracting as an OS can support your needs, knock yourself out. If the job is something a bit bigger than that, use a more powerful OS.
- 11y ago
- pjc50 11y agoFour points: a) Tooling: yes, it's still much easier to debug a native app than a pseudo-native app. b) "Emulating native leads to bad UX": this section is absolutely right. Stop hijacking scrolling and building in cruft. c) "Hassle-free web": absolutely right. You don't need an app for that. Just give us the information and let us have the basic interaction we want. No, you shouldn't present your menu as a PDF either. d) News: this is where I disagree. Users don't go to news sites so much; they go to aggregators which link them to individual articles (reddit, facebook). It's hard to deep-link into an app. This is a case of the app existing to further the commercial needs of the content provider (increase lockin, ads and tracking) in direct opposition to the wants of the user. People keep trying to make the "magazine app" happen, and I've seen very little evidence of it working.
- tragic 11y agoDisclosure: I work on the website of a blue-chip dead-tree newspaper. But I agree. As someone else said in another comment, the basic requirements of a news site is, let's be honest, Web 0.5 stuff. Look, some text about stuff that happened! A picture of stuff happening! These are, shall we say, solved problems. If there's anything where you wouldn't expect significant value-add from the native or pseudo-native experience, it would be news. Certainly as a consumer I have no news apps installed - I go to the website.
- duncanawoods 11y ago>> If there's anything where you wouldn't expect significant >> value-add from the native or pseudo-native experience, it would >> be news I strongly disagree. I see "the news" is a perfect counter example because technology has changed what news is. Its not daily static content anymore, its now real-time, multi-media, interactive and interconnected. The types of media on a news page include: recorded video, live feeds, audio, pictures and embedded digital content from other services like twitter created by both journalists and amateur eye-witnesses. These are not just value-less decoration like cute animations on landing pages but real content from the event being reported. When news is data like election results, tax changes etc. interactive data visualiations are essential for digesting it. Interactivity lets us understand the impact of the news to us personally instead of single homogeneous national agenda. We expect all this and for it to be real-time and provide ways for us to be notified about changes, to share it and discuss it with friends and strangers.
- orthoganol 11y agoSo, the OP is adverse to the JavaScript single page app fad, because in his opinion it is overly complicated to try to get a web app to behave like a native app, and instead should embrace its "weby-ness"? Is he just thinking about developers who have to use the tools he does not like, or the consumer experience side of it too?
- KaiserPro 11y agoSingle page apps are in my experience a massive faff. They may look good, but they tend to break things like back/forwards have text that can't be copied and pasted. Some have embedded video, which eats everything.
- ianstallings 11y agoThis logic is flawed. Disabling back/forwards, copy/paste. and embedding video has nothing to so with a website being a single page app.
- aliakhtar 11y ago> they tend to break things like back/forwards That's not a requirement for building a single page app, there have been ways of emulating history in spas for a while now. > have text that can't be copied and pasted What do you mean? SI'm not aware of any sites whose text 'can't' be copied unless they disabled right clicking.
- landmark2 11y agorant of the day
- mozumder 11y agoThe web works best as a document platform. All the UI tricks that web developers work on all end up managing something related to just browsing/reading a document. It's amazing that modern browsers/HTML just don't have this basic user experience down. Why is it virtually impossible to read long articles on a web browser? After 25 years, why do web designers need to create their own page transition mechanics for long-form articles, for example, when the basic purpose of the web is designed for reading documents? Why weren't multi-page transitions built in by default? The same with photo galleries - after 25 years, why isn't there a standard user experience for viewing photos related to articles? Why are there a million JQuery photo gallery plugins? In fact, why do we need to write CSS to create a basic readable document in the first place? Why doesn't HTML/CSS default to a high-end and legible reader experience out-of-the-box for basic text? (proper fonts & spacings, limiting column widths to 8-15 words max, etc..) The reader-mode that web browsers are building in, that should be the default standard UX for HTML, and CSS should build upon that. But instead, it's a non-standard option.
- msutherl 11y agoLikely this hasn't been done because it would require many very difficult, "subjective" decisions to be made i.e. which font, line-spacing, photo-viewer UI, etc.? Better to make no decision at all than to spawn a few dozen endless committee debates. By contrast, corporations can more easily make these sort of decisions – by fiat – which is why native app frameworks are more "opinionated" / make more design decisions for you out of the box. This topic is as much the story of two historical forms: the operating system GUI and the World Wide Web hypertext project as it is about two organizational forms: hierarchical corporations and decentralized design-by-committee non-profits / multiple competing corporations.
- pestaa 11y agoDon't forget that you can specify related pages as <link rel="next/prev" /> and let the browser handle pagination. I recall when I used Opera, you could scroll down, hit space, and it would go to the next page as long as the website provided that information (forum softwares often do.)
- amelius 11y ago> Is it the web’s purpose to emulate native by inserting yet more features? Certainly not. In my opinion, the problem with the web is that it has become too complicated. In an effort to make the web simpler for the ordinary user, the web has actually become a more complicated place for the (advanced) developer. HTML, CSS and javascript have grown so complicated that they lack certain desirable properties, such as verifiable security, flexibility. For instance, a developer has no choice of programming language, but has to resort to javascript; we are stuck with the cooperative multitasking of the 80s (webworkers support only message-passing, no shared state); also, the developer has no choice of render engine, and to make matters worse, the developer has to write code that is compatible with about five different platforms (desktop/phone/tablet). And of course, the fact that those platforms are not identical is also due to the fact that the web is too complicated. I'm hoping that in the future, we can make the web more simple and elegant for the developer. We need simpler primitives, not the complicated and restrictive building blocks that we have seen thus far.
- to3m 11y agoLack of support for shared state is no disadvantage! When you write multithreaded programs in a language such as C++ or C#, you try to impose on yourself the same restriction, in the interests of making your life easier... (I haven't used web workers, so maybe they're in fact awful to use. But from what I've read, it looks like the right decisions were made, pretty much.)
- amelius 11y agoWhy be so protective? Perhaps somebody invents some programming tool that makes it easy to reason about shared state. This is exactly the kind of mind-set that makes it difficult to develop for the web. Also, shared immutable state is a useful (often perhaps even essential) property for building efficient code in a multithreaded environment, even when using the message-passing style.
- to3m 11y agoTastes differ. I don't have any problem with erring on the side of overprotectiveness when it comes to multithreaded programming.
- LunaSea 11y agoNative apps are by definition less portable than web based ones. As a developper I look at native app development the same way I saw the web in the early 00's. Different platforms, screen sizes, versions, programming languages, paradigms ...etc.
- cjslep 11y ago> Native apps are by definition less portable than web based ones. I don't believe it is that clear cut. I think you are ignoring the fact that native developers have frameworks of their own. Using Boost, GTK+, or Qt gives various layers-of-thickness over the OS so the native code is portable. Some, like Qt, also cover mobile operating systems as well as desktop ones. These frameworks are not young either; so while websites had browser-specific-detection-code (thanks to IE), native apps were already running cross-platform by leveraging frameworks of their own.
- LunaSea 11y agoI haven't seen anyone use Boost, GTK+ or Qt cross mobile / desktop.
- pjmlp 11y agoDropbox at CppCon 2014, "Practical Cross-Platform Mobile C++ Development" https://www.youtube.com/watch?v=ZcBtF-JWJhM https://www.youtube.com/watch?v=ZcBtF-JWJhM MuseScore http://showroom.qt.io/musescore/ http://showroom.qt.io/musescore/ Qt show room has many others - http://showroom.qt.io/ http://showroom.qt.io/
- robmcm 11y ago>As a developper I look at native app development the same way I saw the web in the early 00's. Different platforms, screen sizes, versions, programming languages, paradigms ...etc. This is because you are using tools and paterns to avoid having to think about the differences. Those tools are also on native platforms, and those differences apply to the web.
- pavlov 11y agoI feel we’ve gone too far in emulating native apps. Conceding defeat will force us to rethink the web’s purpose and unique strengths — and that’s long overdue. This, a million times. Mobile websites and mobile apps have completely different strengths. The current trend is to develop them both with the same HTML-based toolchains and make them as similar as possible, which ends up being to the detriment of both. Users don't want mobile apps that are simply a website packaged up behind the "icon on the homescreen". Those apps lack the essential benefits of native: fast and seamless access, smart use of local data, integration with device services like notifications... There's nothing more annoying to a mobile user than an "app" that takes 10 seconds to start up because it first loads a browser engine, then makes a hundred HTTP requests to fill up that embedded browser with content. (And if the 3G network happens to be clogged, the app may end up showing nothing after 20 seconds.) For developers, Cordova/PhoneGap-style tools are not a panacea either. It's easy to get an 80% solution done, but then you run into problems with mobile browser performance, browser differences between devices (even Android devices with the same base OS can have very different web view browser engines), etc. With all that, the last 20% of a Cordova app may well take 80% of development time, and that's rarely been budgeted in. Shameless plug: my startup Neonto makes a UI design tool that creates usable iOS and Android code from visual layouts. It's a great way to remove the friction in creating real native apps: http://neonto.com http://neonto.com
- soft_dev_person 11y ago"With all that, the last 20% of a Cordova app may well take 80% of development time, and that's rarely been budgeted in." To be fair, that could be said about most software projects. "80% done, now the last 80% remains." Not that I disagree with you, I had the same experience trying to make an Android app in Phonegap. It's lacking, like most cross-platform frameworks (for most types of apps, but typically not games) I have tried. They are only suitable for basic stuff where the native feel isn't critical.
- pavlov 11y agoTo be fair, that could be said about most software projects. Yes, absolutely true. The difference here is between "known unknowns" and "unknown unknowns", to paraphrase a certain renowned military strategist... With Cordova/PhoneGap, it's easy to get the initial impression that you're dealing with "known unknowns", just the typical web development project pitfalls -- "This is HTML, I know this stuff". When the "unknown unknowns" hit the fan, it's often pretty late in the development cycle, and so things blow past original estimates even though they were competently planned. It's just that the plan was for web development, not the shifting art of websites-in-drag-pretending-to-be-apps-development.
- thomasfoster96 11y agoPerhaps what we really should be thinking is that 90% of applications can, in theory, be made using either web standards, native or a mix of both. Unfortunately lots of websites and web apps are not very well developed. They serve bloated HTML that hasn't been minified, let alone compressed. The page is unresponsive or blank until a dozen JavaScript files have been loaded. Hardly anything is cached. jQuery is loaded to do select a single element and animate it when CSS would have done. Most well written, architectured and developed web[site|app]s will not be too different in UX for end users than a native app. Sure, you can't write a great 60 fps first person shooter game for the web, but there's nothing saying you shouldn't be able to create an RSS reader or photo sharing app or social network or anything else with either the web or native.
- castell 11y ago> Sure, you can't write a great 60 fps first person shooter game for the web you can, you know WebGL is hardware accelerated and there is Asm.js - even the Unreal game engine runs in Firefox and Chrome.
- pjmlp 11y agoBoth drive my desktop PC fan to full speed, let alone what they do to my mobiles when the sites manage to render, if at all.
- thomasfoster96 11y agoI doubt you'd get a very good, highly detailed game running at 60fps with WebGL and asm.js today. Nothing comparable to what you'd get running well on a console. Maybe in another couple of years, but not today, and by then we might have non-web platforms doing 4K 60fps with ease. But a lot of games do run quite well using WebGL. Just not the jaw-droppingly realistic ones.
- janl 11y agoI’d love to take this discussion offline with a number key people. We have the Web Technology Fest in Berlin (http://wwwtf.berlin http://wwwtf.berlin formerly http://jsfest.berlin)in http://jsfest.berlin)in September (leading up to JSConf EU & CSSconf EU) and last year we’ve already held and Extensible Web Summit (http://lanyrd.com/2014/extwebsummit-berlin/ http://lanyrd.com/2014/extwebsummit-berlin/) among dozens of other events. Here’s a rough draft for what I envision, but I’d need your help to shape this, so it’s going to be useful: https://github.com/janl/wwwtf.berlin/issues/3 https://github.com/janl/wwwtf.berlin/issues/3 Thanks! <3
- sytelus 11y agoWeb APIs are in essence becoming an abstraction layer between OS and applications. It's not just about browser anymore. Browser is just one of the host for these APIs. In next decade or so, we should have all the power built in to standards that eliminates pretty much any advantage that native application can possibly have - even for heavy duty 3D gaming and offline video editing scenarios. Web APIs will and should become the standard way of developing apps across platforms. There is neither a theoretical reason that this is impossible or any other viable alternative agreeable among OS providers. There is nothing to concede or regret about. It's just how things evolve. Get over it.
- aliakhtar 11y ago> In next decade or so, we should have all the power built in to standards that eliminates pretty much any advantage that native application can possibly have Agreed. I was just watching this talk the other day called 'Making web apps appy' froom Chrome Dev Summit: https://www.youtube.com/watch?v=QbuLq4f6DGQ https://www.youtube.com/watch?v=QbuLq4f6DGQ It talks about a lot of new incoming features, like push notifications from the browser, adding icons to home screen directly, offline access, etc.
- slightlyoff 11y agoThose features have all launched; see: http://blog.chromium.org/2015/03/chrome-42-beta-push-notifications_12.html http://blog.chromium.org/2015/03/chrome-42-beta-push-notific...
- pjmlp 11y agoYou mean, RPC, CORBA, DCOM, XML-RPC, SOA,....? As someone has written on his tweeter account "Web APIs are hipster RPCs".
- rimantas 11y ago> Web APIs will and should become the standard way of > developing apps across platforms God forbid. There are few lamer things than those Web APIs. Why do you think the gazzilion of frameworks exists, if not to work around all the things broken? And take a look at the web in 2005 and now, see what real progress was made. Yep, we got a bunch of half-baked APIs, which will never mature, because all the crowd already run forwrad to tick another checkbox.
- shawnreilly 11y agoIt does not matter what platform you choose (web or mobile), bad developers develop bad software. For every horribly designed, slow, bloated website you find, you can likely find a similar horribly designed, slow, bloated native app. I'm sure you've heard of garbage in garbage out. I view development as a craft and like most crafts, not everyone has the same level of skill. This does not reflect on the platform people choose. It reflects on the people themselves, their time investment, work ethic, priorities, approach, methodologies, etc. Like with most crafts, there is usually more than one way to accomplish something; some more eloquent than others. At a very basic level, both web and native apps share many commonalities; some sort of UI being rendered and interactions with data via API calls. As long as you can reach your goal, does it really matter how you got there? How about this; the debate is ridiculous. They both rock (or suck depending on your perspective).
- robmcm 11y ago> bad developers develop bad software I whish this phrase would die and be replaced by bad "teams" produce bad software. I have worked on some projects with fantastic developers, but awful or no UX, or a misguided vision, rushed to market etc etc. I have also worked on projects with awful awful code base but the end product is actually alright (so long as it doesn’t need any maintenance). Perhaps we should flip it on it's head and say, "good technology doesn't guarantee good software".
- blub 11y agoMany companies and teams had a "web hammer" and went looking for things to nail even when that made no sense. At the peak of the web frenzy, web apps were going to kill desktop apps, and one could feel the pity when web devs were talking about desktop and mobile apps. And now... someone finally stops and thinks that maybe it's not such a great idea to turn everything into a web app. When a group of individuals takes such a strong position and they are proven to be wrong, I personally think some crow eating is due :-)
- fizk 11y agoRemember Flash? Narrow the gap, add a bit of Steve Jobs and Boom! The web Won. Look at a site like YouTube today. All the tooling we've created and all the progress of the open web platform that has made that site happen is incredible. If we've just given up 10 years ago, saying to ourselves that the web should only be for documents, then we would be missing out big time right now. It’s not for every site to try and push the envelope. And mimicking native can often lead to bad results. But to go from that and say that we shouldn’t try. That’s just sad.
- pjc50 11y agoYoutube is an interesting choice of example, since in a way it's "just documents" - those documents happen to be video, and are surrounded by hyperlinks to other video's pages. The real achievement here was cutting through the intellectual property thicket so that everyone could have an in-browser video player in HTML5. The first site I ever saw that used Javascript to produce a useful application rather than annoying frippery was Google Maps. (Rhetorical question: why does everyone use youtube rather than hosting videos on their own sites? Unpacking this question will show the obstacles that "redecentralisation" faces)
- nivla 11y ago>why does everyone use youtube rather than hosting videos on their own sites? Unpacking this question will show the obstacles that "redecentralisation" faces Two reasons: Bandwidth and Speed. Google has contracts with most major ISP's for caching Youtube videos. This gives them access to the best CDN money can buy! No matter how awesome or powerful your current datacenter/CDN is, it still wouldn't be able to match the fetch-straight-from-the-ISP performance. Interestingly this also creates a hidden monopoly since the barrier to entry is costly.
- nostrademons 11y agoIt's more than that. There's also legal liability (very few people would be comfortable hosting the amount of pirated content that appears on YouTube, nor would they have personal bandwidth for DMCA takedown requests), discoverability (YouTube recommends content you've never seen before, intelligently...by definition a personal site can't do this), and search (which only works when you have enough content to be worthwhile to search over). All of the algorithms used for the latter two points require a good amount of data, so Google gets large economies of scale there.
- nly 11y agoIt's HTML and CSS holding 'web apps' back. It's like we're all trying to sharpen our pencils with spoons. I think the future of remote web-like apps lies in something like QML[0]. A completely native UI with full access to the desktop widget suite can be downloaded in QML in a few kilobytes. Stick a browser like loader on the front of it, and provide a decent (seamless) asynchronous RPC, and you've got yourself a replacement. [0] http://doc.qt.io/qt-5/qmlapplications.html http://doc.qt.io/qt-5/qmlapplications.html
- VLM 11y agoSpawn off a rdesktop or VNC session back to a cloud based server. Skip all the intermediate interference that just slows down and complicates things, and turn the web into a dumb graphics terminal. You can either intentionally use the correct tool for the job, or painfully slowly buggily reinvent it if NIH is a requirement.
- jlebrech 11y agoweb byte code
- pornel 11y agoTo me the worst side-effect of naively chasing "native" is permission prompts. The web platform wants to have powerful features that are dangerous and/or easy to abuse, but haven't found a good way to allow them safely. The "solution" we've settled for is to blame the user for clicking "OK" on permission prompts that have unclear consequences to non-technical users.
- Brakenshire 11y agoThat's exactly the same as iOS native permissions, isn't it?
- pornel 11y agoSimilar, but the key difference is that Apple can kick the worst abusers out of the walled garden. The Free Web can't do that (and I strongly prefer it to remain Free with low barrier to entry), so it needs a better approach, or needs to give up on having dangerous features available.
- rockdoe 11y agoThe web platform wants to have powerful features that are dangerous and/or easy to abuse, but haven't found a good way to allow them safely. The same problem exists on native.
- pornel 11y agoIndeed, that's why I think it's a mistake to copy native with problems plaguing it (when we don't know how to add these features without the downsides). I'd prefer the Web to stay safe and hassle-free, and use native apps—with the risk of letting malware in—only when I really need more powerful features.
- bobajeff 11y agoYeah, this is kind of what I've been thinking: Let data be data. Let features be features. And let user interfaces be user interface. So many websites mess this up in order to deliver a quick working site that only works on the desktop and sometimes only Windows. And now we have sites making seperate ghetto mobile versions or mobile apps missing features. We still have data that can't be accessed through URL with text that can't be selected or shown without JavaScript loaded. Unless you are doing something more complicated than content devilery (ie games or utility apps) there is no reason to override fundamental features of the web.
- archagon 11y agoIt's clear that the web has outgrown its original intent. Instead of just being a simple substrate for linked documents, many (if not most) websites are now trying to emulate applications, platforms, and services using the poor, ill-suited DOM. But I don't think the solution is to give up. Native apps are great, but there's a vast gulf between the binary running on your device and the server you got it from. Downloading and updating take forever. Connecting to other apps is barely possible. Everything is horribly siloed. The brilliance of the web is that it allows people to quickly connect to computers anywhere in the world and get at whatever stuff they might want to serve; native apps will never have that level of flexibility, barring some horrible proprietary ecosystem. Unfortunately, the web was designed primarily with content in mind, not presentation. The only way servers can provide interesting and slick user experiences is to hack up the content presentation system, which is as silly as using spreadsheet scripting to build enterprise software. So I think the best approach would be to revisit the design of the web. Instead of having the web be strictly document-centric, I think there should be a higher level above that, through which HTML, CSS, the DOM, and everything else are implemented. A sandboxed, scriptable, low-level canvas that only gives you the bare essentials, allowing you to create native-quality experiences without having to piggyback on the DOM or reinvent the universe via canvas or WebGL. A "Vulkan" equivalent for web software. Not making a document? Feel free to not use HTML — or only embed it in the parts of your web-app that actually present content. It should be possible for a user to go to facebook.com and immediately retrieve an equivalent to the Facebook app in their browser — smooth scrolling, gestures, and everything. (Or whatever the future equivalent of a browser might be.) But if a user navigates to a simple blog, it should still appear as it does in the web today, because this new web would be a superset of today's web. We've been interacting with other computers through a seriously constrained pipeline this past decade. Instead of bowing our heads and conceding defeat, why not blow it wide open?
- TazeTSchnitzel 11y agoWhat layer can you realistically have below the DOM? If you get rid of it, how can you realistically avoid reimplementing everything? I don't think this makes sense.
- 11y ago
- castell 11y agoThis article feels dated. If he wrote it in 2011, I would agree with him. Web apps don't try to recreate the native UI anymore - something that used to be the case some years ago in the iOS1-6 era. Nowadays the UI of Android 5+, iOS 7+ and WinPhone7+ look very simple and Web-alike. Simply avoid bloated outdated big JS libraries and use a minifier to shrink the JS and CSS files. I thought the trend reversed as many little commerce switched from offering a Android/iOS app to a mobile-friendly website. This may not be a global trend. quirksmode.org used to be great resource in the IE6 era. Nowadays http://caniuse.com http://caniuse.com and Mozilla's MDN have replaced it completely for me as it offers up-to-date information.
- TazeTSchnitzel 11y agoDon't just use a minifier: use something that can eliminate dead code, like Google's Closure Compiler in advanced mode. Don't ship 10MB to the client if you only need 1MB!
- drinchev 11y agoBut what should the author explain to Chromebook users? They can only have web applications on their desktop as an icon. Actually I think they're doing pretty good with that.
- pjmlp 11y agoI never saw anyone with one in Germany, besides computer stores trying to get rid of them with massive discounts and still they gather dust.
- stuartmemo 11y agoAmazon's best selling laptops - http://www.amazon.com/Best-Sellers-Computers-Accessories-Laptop/zgbs/pc/565108 http://www.amazon.com/Best-Sellers-Computers-Accessories-Lap...
- pjmlp 11y agoOn the US market. Additionally how many of those laptops are actually running ChromeOS, instead of being replaced by a proper Linux distribution?
- stuartmemo 11y agoOk, here's the UK market - http://www.amazon.co.uk/Best-Sellers-Computers-Accessories-Laptops/zgbs/computers/429886031 http://www.amazon.co.uk/Best-Sellers-Computers-Accessories-L...
- pjmlp 11y ago3rd place in Top 20. Still you haven't answered how many of those are effectively running ChromeOS. Looking at Germany, which was was I mentioned on my post, the scenario doesn't look like that anyway.
- brudgers 11y agoSure it's just me anecdotally, but I don't care about native apps versus web apps. I care about where the ship takes me not the arrangement of its deck chairs. I'll use the command line...or the address/search box...if it gets me what I want. Good user interaction and great functionality subsume widget chrome and when it comes to native versus web experience please feel free to do something else on my behalf whenever it means doing more of the things I care about and less time bikeshedding.
- kozak 11y agoThere's a big market for ships with nice deck chair arrangements.
- brudgers 11y agoYes, of course and too many lifeboats might scare the passengers. That's why one can find work in the deck chair arranging industry, attend deck chair arranging conferences, and engage in debates over whether it is best to arrange deck chairs using tabs or spaces. Yet somehow, users manage to navigate the web despite Apple's app store gatekeepers not setting policy. And really the debate about emulating native exists because of app store policy, not because the native standards are inherently better than what someone might invent left to their own devices...or even that they are necessarily better than the command line.
- gjvc 11y agothose who do not learn from history are condemned to repeat it. watching the mother of all demos once a year is a sobering reminder of that. https://www.youtube.com/watch?v=yJDv-zdhzMY https://www.youtube.com/watch?v=yJDv-zdhzMY some better-than-most discussion here: http://arstechnica.com/the-multiverse/2015/04/from-the-vault-watching-and-re-watching-the-mother-of-all-demos/ http://arstechnica.com/the-multiverse/2015/04/from-the-vault...
- whoisthemachine 11y agoTo a certain extent, I think this is true. Many mobile sites still need to be more engaging, but they need to do it by taking advantage of their strengths - server side rendering, a nice easy way to stylize (think CSS is bad? try styling your application in Android's framework, or WPF for that matter) - rather than going the route of the "SPA". Once a web app tries to look and feel like a native app, other things out of the web app developer's hands are going to make his app always look inferior. In fact, it may be some of those strengths I listed earlier.
- pjmlp 11y ago> try styling your application in Android's framework, or WPF for that matter I will take any of those over CSS, thank you very much.
- whoisthemachine 11y agoHmm what advantages do those have that makes you prefer them over CSS? I work in both of those daily and they're usually just an exercise in frustration.
- pjmlp 11y agoThey make more sense to my developer mind without designer background, don't suffer from CSS global namespace, inheritance side effects and browser specific behaviors.
- whoisthemachine 11y agoHuh I also have a "developer" mind, and very few people would call me a designer, and those that do should re-think what a designer is. However, from my observations over the years, the cost of "browser specific" behaviors is equal to or less than the cost of learning entirely new styling paradigms every time someone puts out a new UI framework. CSS also has the benefit that over time browser specific behaviors fade away as standards take shape, which is inherently not the case with UI frameworks. As far as the namespace goes - I guess, but there is nothing to stop you from coming up with your own namespacing scheme using nested CSS classes.
- smhg 11y agoThis discussion should be about discovery and openness. Not about functional/technological differences between 2 platforms. Those will evolve. - URL's are a great starting point for interactions (see The Physical Web). - App stores (in their current shape?) just can't be the future.
- mromanuk 11y ago> Technically, it’s simple. The web cannot emulate native perfectly, and it never will. Native apps talk directly to the operating system, while web apps talk to the browser, which talks to the OS. Thus there’s an extra layer web apps have to pass, and that makes them slightly slower and coarser than native apps. This problem is unsolvable. Really weak argument, it reminds me of High level languages will never replace ASM... Windows 95 will not replace DOS... Software is composed by abstractions, doesn't matter how much levels of it, better computers allow it. Thinking that hardware<->OS is better than hardware<->OS<->browser is an oversimplification.
- rimantas 11y agoExcept in this case we are not developing high level languages, we are trying to patch poor ASM. I see the height of the web in the early 2000s, when web standards, accessibility, clean markup meant something. Now it is just a pile of frameworks (as someone twitted: JS frameworks are standards fanfiction) thrown together. And not even for apps, for a simple text page :( I will repeat Tantek's definition: if it is not curlable, it is not web.
- blisterpeanuts 11y agoCurrently, best practices are web apps for light user interaction, native apps for heavy user interaction or complex graphical presentation, and hybrids (native app that is essentially a container for a webview) for best-of-both. My money is on hybrids for the near term. A native app with a webview gives you a lot of local control but still leverages the server power and advantages of centralized web apps. The Youtube examples mentioned in these comments are a parallel of this approach (I doubt they use webkit, probably just access the streaming protocol, but effectively similar). I'd love to hear some comments/opinions on these thoughts!
- wwwtyro 11y agoOne point I've not seen mentioned yet is that the web is a fantastic distribution platform. Given the option, I'd much prefer to have written my molecule rendering project[1] on a native platform, but I'd get only a fraction of the users I get when deploying to the web. The activation barrier for native apps is just so much higher because of the download/install/launch process. For now, the URL click is king, and it's hard to beat easy when you want people to use what you've made. [1] http://wwwtyro.github.io/speck/ http://wwwtyro.github.io/speck/
- cbd1984 11y agoThis page is so narrow, and the text so small, it's unreadable.
- forgotAgain 11y agoThe post forgets a lot of the history of why web apps were originally welcomed: they allowed a choice of platform. Going native undeniably gives a better user experience. However there is a trade off in developing for a single platform that cross platform tooling cannot yet remove sufficiently.
- joeblau 11y agoI just had this discussion a few weeks ago and it wasn't even about mobile applications; It was about mobile application testing frameworks. We were deciding between a hybrid approach to testing (Appium) vs native (UIAutomation). I don't think these types of debates are going to end soon because most of the decision makers in large companies have grown up with and love the web. For the most part, I've only seen the web get faster and better since the mid 90's, but that was limited to the platform of a computer. The theoretical maximum benefit of writing a program once that runs everywhere (watch, phone, phablet, tablet, laptop, desktop, tv) and is always up to date (computing is done in the cloud) is extremely enticing but has serious performance implications. What you always end up with is a developer friendly framework that contains cases for each platform and in most cases, doesn't have all of the features each individual developer needs (Bootstrap, Appium, Node.js, React, Angular, etc...). I want the web to win, but in most cases the DOM was not built and designed for the direction that native applications are going.
- virmundi 11y agoWhat about the benefit that the web brings to updating the user's install? I know that it is technically possible to push a new Java or .NET app out to users. I've done it. The one thing it requires is a lot of coordination. Notify the users to keep their equipment on the network from time X to Y. Notify the infrastructure team to schedule a push. Have the dev team on standby in case the push goes bad. Have the software protect itself when inevitably someone's install is corrupted or skipped. Compare that to the web. User opens Chrome; they get the new app. No push. No real notification of any team. Hell, Ops probably doesn't have to be notified if they automated web app deployment to servers. The whole thing is very easy. Even if you have the added complexity of a Chrome-app wrapper, still pretty easy. Given the ease of deployment and the mitigated risk of data corruption due to version issues of the app, the web looks pretty good. Given the advancement of hardware and software, UX can be close to native. If not native, app-y for lack of a better word. If it lacks UX for native, don't target native. Target app-y. This is between native and document. It might become complex to manage local state, but that shouldn't be a deterrent. Tangentially related, to the general question: does the user want to have a link to your product on their desktop? The answer is probably no. They want easy access to your product, but the link conflates how this occurs with a sense of permanence. Your app is not the product. Your app is tool to get the product: information. UX and installation are both just means to the end.
- eloisant 11y agoTo be honest, Google takes care of all the complexity to have apps updated on Android. It's even better, because you can do a stage rollout really easily (i.e. just deploy to 10% of users, then 20%, then 50%, etc.). It's possible to do on a web app but you'll have to do it manually. You still have to push the apk and stay alert for the feedback, but you have to be careful with web apps update too anyway. I'm not sure about Apple because it appears the auto-update process is not that smooth and people don't upgrade their app as often as Android users.
- frik 11y agoAuto-updates are great except in the rare cases where they aren't at all. Picking up my phone in the morning to recognize that it lost half its battery in the night due an extensive auto update and being left in the cold with a 10% battery (instead of 60%) isn't that great. Beside that I think virmundi above has his points, and I agree with him. App updates are not instantaneous.
- jpswade 11y agoThe best argument I've ever seen is that it's ridiculous to think that there will be an app for every business, website and device. Instead, I see a future, like we're starting to see with devices such as the Synology NAS where the interface is a website served on the device. Things like the ESP8266 make it possible to add Wifi capability to pretty much any device. The future isn't about an app store nor about search, it's about discovery.
- wtbob 11y agoI prefer the experience of native mobile apps (and native desktop clients, too!), but I prefer the privacy and sandboxability of pure-HTML apps. If a site is going to add so much JavaScript that it's unusable without whitelisting it all, then I almost might as well just use a native app. If Android had ever-so-slightly better permissions, then the clear answer would be to stick with native apps, period.
- UUMMUU 11y agoThis is probably going to be an unpopular opinion but I think the current push to make native feelings apps in JS is awesome. Playing with an app like Untappd and then realizing it's written by (at least originally) 2 people and still feels fluid is incredible. I've also never heard anyone except developers complain that it's not app-y enough. True things like games shouldn't be written in phonegap or react-native but I think to say that we should completely stop trying to build any native-feeling apps is wrong. I actively push anyone interested in turning their app idea into a working prototype use Meteor or React-Native or just straight up phonegap and if they get traction they can do what facebook did and rewrite. I know this flies in the face of this article but I just don't think we should stop trying.
- capkutay 11y agoThis may not be in line with the author's point, but I agree with you. You probably don't need to build a super optimized native app to launch your MVP. Developers should use whatever tools they're comfortable with for getting the job done. In the event that they get thousands of users and need to scale, then they can think about hiring people to build native apps.
- Thiz 11y agoHtml is for documents served thru http. What we need is a language to build apps natively and serve them securely thru the browser using an app:// protocol. Working with html to build apps is a nightmare. Google, Apple and Mozilla should take the best from android and ios and work on something universal and truly disrupting. No, not java. No, not obj-c. Perhaps swift, go, nim or something better. Remember the hobbyists.
- Thiz 11y agoForgot to add. Let's use an emulator and not a browser to serve apps thru the app:// protocol. We don't need bookmarks, toolbars, addressbars, etc. We just need a surface and an app store, just like a mobil os, with the triangle, square and circle as universal UI for navigation. Best of both worlds.
- pinaceae 11y agoas long as you use a markup language plus styling language built for documents to somehow emulate app interfaces, the web will lose against native. the www is great for content. wikipedia. awesome. but it just sucks for apps. the constraints it puts on UI and hacks needed to get around them is staggering. let's center something on the screen with CSS.
- bovermyer 11y agoThe author's argument that news sites should go native and (presumably) forgo web ignores the fact that the web is highly shareable and discoverable, where native apps are a walled garden. Ignoring news sites' need to generate revenue for a moment, the ability for a user to search the web for a particular topic and discover a relevant news article from three years ago is a built-in advantage for the web. You can't duplicate that on a native app.
- snarfy 11y agoWe already did concede and invented asm.js. Unreal Engine runs in my browser.
- Zigurd 11y agoThere are several reasons mobile devices made it so that Web runtime and Web apps would not be the consensus app environment: 1. Web apps are no longer the answer to breaking Windows hegemony on apps. 2. The Web is a mouse UI. That may seem like a small thing, but look at how much trouble Microsoft has had turning Windows into a touch UI. 3. Mobile devices are the end of overlapping windows. The Desktop Metaphor was a cute idea, but it isn't appropriate for mobile devices. Web browsers, and Web apps, are made for resizeable windows. 4. The Web is great for simple interaction with dynamic content. Which is another way of saying "The Web is a big, linked, interactive e-reader." You can stretch it to try to do other things, but it isn't a comfortable stretch. 5. Pure canvas apps... ... ... </crickets> 6. No one development idiom. Nor two. Nor five. This is much less of a consensus than Android or iOS. IOS developers will be making a language transition, and Android developers will probably incorporate a dependency injection idiom into nearly all Android coding once a consensus is reached on what flavor of DI is favored, but that's about it for uncertainty.
- Zigurd 11y agoWhat's appalling is how many of the mobile app-related comments here are about cross-platform panaceas. Just don't. Well, OK, if you have an enterprise shop full of C# coders, you should evaluate Xamarin. But unless that's your story, and unless your tool is as well-engineered and well-funded as Xamarin, trying to get one code-base across Android and iOS is going end in pain and regret. It's back to a future of multiple platforms with fairly strong differentiation in features and toolchains. Deal with it.
- codingdave 11y agoI really don't know anyone who carries on these kinds of arguments. In my experience, we code for the web most of the time, because it works and is usable by most people on most devices. Once in a while, we have a compelling reason not to. When that reason comes along, we code a native app. It isn't a competition between platforms, and we aren't trying to make them look and act the same. We just pick the platform that offers the simplest workable solution. Maybe I just run in a more pragmatic crowd...
- jebblue 11y agojQuery Mobile and done.
- dfield 11y agoThe author is correct that modern web applications are hacks all the way down. It's crazy sometimes. But it fucking works! Look at Google Docs, which has changed the way teams collaborate. Look at Unity which recently ported their player to the browser. Look at the Autodesk 123D suite. With technologies like WebGL and Emscripten, we're just starting to see the tip of the "native on web" iceberg. And it's awesome. I know because I'm helping build the iceberg... :)
- fixermark 11y agoAnd yet Google Docs on a mobile device is a native app, not a web site. The website experience is significantly degraded relative to the native app. It's awesome on the desktop though, but there are frustratingly good reasons that Google couldn't just run that code on phones and expect it to be anything other than a terrible experience.
- MBCook 11y ago> It's awesome on the desktop though Not in my experience. I've never had a native app pop-up a dialog saying "You can't use that menu to copy and paste, you have to use the keyboard shortcut." Instead, the standard platform way of doing things just works. Then there is the fact that Google Docs just refuses to open certain docs in certain browsers. It's impressive for something on the web, but it's not better than a native app by any stretch.
- Too 11y agoGoogle docs is a great example of an app that should not be web. Especially google sheets, brrhgh, i get shivers just thinking about it. Even Excel 98 on a pentium 3 has better performance.
- kwijibob 11y agoMaybe for an advanced use you hit speed issues. But Google Sheets has transformed many collaborative functions in our organisation. Especially at budget time. The ability to have concurrent access to the latest version always is wonderful. No more emailing attachments. Also being able to sit down at a budget meeting and everyone can view the work-in-progress as it develops on their own device. In all of this we've never had any one bothered by lack of speed.
- mangeletti 11y agoI think web technologies are a bit like AI and VR; they catch up in waves to meet demand, but they don't simply replace everything all at once. IMHO, the reason for this is that once web technologies almost catch up to native, native buckles down and fights back (e.g., Swift), and as long as native can keep developers happy, it will always own serious market share, since it is closer to the metal (faster) and necessarily has access to more APIs (more features).
- leecarraher 11y agohonestly, thought the defeat was in the other direction. Although I am a stubborn native app loving, cloud denier, i fear more and more that I am in the minority. Everything new is web based. It is true that there are perfomance hits, communication bottlenecks and added complexity to developing on the web, but i think the biggest hurdles are slightly less obvious. Security requirements are exacerbated and the size of development units have grown intractably large. More speed requires better direct HW access but every time the web wants more, the blackhats push back and take even more. What happens when facebook advances react, google remodels appengine, AWS changes... Nothing is static, making the "who broke what" question sometimes impossible to answer.
- markbnj 11y agoThe web only works for mobile when you have a responsive design, which again for most people means adopting one or more frameworks. News stories should be perfect content for a mobile web presentation, and yet we have Flipboard. The app is free and the service is sub-based, so business model wasn't the issue.
- hypertexthero 11y agoTwo problems with native applications: 1. The systems and hardware they run on are usually controlled by one company. 2. There's no View Source. Where would we be if 'native' applications had won over the '[worst is better][1]' approach of the web? What do you do if you lose your phone and the only way to get information is through an app that was installed on it? [Alex Martelli][2]: > Our culture’s default assumption is that everybody should always be striving for perfection – settling for anything less is seen as a regrettable compromise. This is wrong in most software development situations: focus instead on keeping the software simple, just “good enough”, launch it early, and iteratively improve, enhance, and re-factor it. This is how software success is achieved! Good enough is usually good enough and can be accomplished through a web browser if your design is good and not compromised by 'business decisions'. There is a place for native apps. Communication applications such as Whatsapp are a good example. Applications whose main purpose is to provide information, such as websites, are not. Finally, lets not forget that the internet is not a thing, [it's an agreement][3]. [1]: http://hypertexthero.com/logbook/2013/07/europython-2013-notes/#goodenough http://hypertexthero.com/logbook/2013/07/europython-2013-not... [2]: https://ep2013.europython.eu/conference/talks/good-enough-is-good-enough https://ep2013.europython.eu/conference/talks/good-enough-is... [3]: http://worldofends.com/ http://worldofends.com/
- tim333 11y ago>I think we web developers should advise news sites, and in fact any site that prioritises features over performance, to go native As a user, I have no interest in that. On the front screen of my phone I have the 'google news' icon which is a link to a web page with links other news web pages. Why would I want 20 siloed apps for the various news sites I visit occasionally?
- dasil003 11y agoI get where he's coming from, but I don't think it gets to the root issue: Why does everything need to be a "rich experience"? Note I didn't say app-like, because many apps themselves are straightforward. If we don't make that distinction then we risk throwing out the baby with the bathwater. He cites having to build "2 native apps" instead of one website, but that's only true if you ignore everyone who is not on recent versions of Android or iOS (certainly you're not going to support iOS < 5, right?). The web is no-install, ubiquitous and accessible in a way no platform has ever been and which is impossible for any proprietary platform ever to be, and is backwards-compatible. Given these strengths, the effort to bring app-like functionality to the web is well-founded, because difficult though it may be, the converse of bringing the web's benefits to native apps is just plain impossible. If when marketing wants to create these experiences we tell them to go native we're just shoveling the problem off to the poor native developers. The first step should be determining the value of said "experience", is it there to serve the user or is it there win awards?
- Too 11y agoThe no-install, and especially the sandboxed environment, is an argument that hasn't been discussed much here yet but which i think is a big factor to take into account. Sure, device and browser compatibility is nice but really, being able to open a news site without having to share my contact list, my location and my photo album is what makes the web a no-brainer. If native installs could cater more for this scenario it could win some more ground, especially on windows, there you have no control at all over what goes into your machine. Android Play store and and Apples App store would also need to improve on this. They pretends to have support for it but the permission categories are all to vague, they are either all or nothing, and they have been abused to the point where nothing can be trusted any more.
- zokier 11y agoJava offered no-install sandboxed apps well before the webapp craze. But of course being based on Java made that attempt quite stillborn. http://en.wikipedia.org/wiki/Java_Web_Start http://en.wikipedia.org/wiki/Java_Web_Start
- netfire 11y agoHere's the problem: native apps require downloading, updating and take up more space on the device. I think the mobile web could be much more than it is now, but mobile web browsers seem be improving at a very slow pace. One of the main reasons for that is that the creators of the mobile operating systems benefit more from an app store environment, where they can take a cut of the sales of an app, so they have little motivation to improve the browsing experience. There's no reason that mobile browsers couldn't expose more internal access to components or controls (and UI options) that are currently only available to native apps. Do we really want to be restricted to a environment where if we want to release an update to our app, we have to submit the update to the third party for approval? (which can take days) and where users have to download your app to try it out? We shouldn't concede defeat. We should demand more from mobile web browsers.
- EGreg 11y agoI disagree. Especially for startups. I will take "the Web" to mean, essentially, that the rendering of the front-end is done in a browser. Then, what you have to do (I learned this the hard way) is stick to things that get rendered well in browsers on all the platforms you plan to target (which usually includes Android, and that means no fancy scrolling). Develop your aesthetic around that. You can put out native versions later, when you have the budget to support all the platforms. Now, this has several advantages. One is if you want your app to become popular. If your app involves inviting friends in any way, realize the friends don't have the app, and the invitation will probably happen via SMS/email/facebook/twitter. When that link is clicked it takes the friend TO THE WEB. Yes you can have the friend look at an app in the store, download it, launch it, confirm their phone number again, and then finally engage with whatever they were invited to. Or you could have a WEB EXPERIENCE ready for them and create an account with 0 steps! And since you've already coded this web experience, why not wrap it in Cordova/PhoneGap and add bridges to the underlying OS APIs via plugins. This keeps separation of concerns and actually maintains much better than many native apps. And you'll also instantly have access to many cool components like Mobiscroll, which have been tested and work everywhere. Compare that to downloading some native lib for iPhone, like Layer, only to realize it doesn't work on other platforms like the web, and you have to wait or roll your own. So now you've got a unified codebase which renders in a browser, you can wrap it in MacGap and other wrappers, hire specialized developers on each platform and pay them to maintain ONLY the part they need to. Not the interface, not the server interaction, backend etc. Which all has a standard and RICH environment to build in, including localStorage, local database, sockets, even WebRTC for peer to peer time topologies! All that stuff is good enough in 99% of the cases to build an app. Well I'm a little biased. My company has been building a plaform for years to make apps successful and viral, a sort of open source alternative to Facebook (which just shut off app access to people's list of friends, something that was the biggest selling point at the very beginnng and the reason you shouldn't build your business on someone else's centralized platform). The idea was to have a standard platform finally to power SOCIAL apps, that was all well designed and take care of everything for you, so you can focus on write your app, not getting cordova, user signup invites and privacy / security right etc. It also supports a growing ecosystem of reusable widgets and plugins that you can just plop onto a page, like chatrooms or docs. I thought it would be done in a year but it's been four years and we're still at 0.8! http://platform.qbix.com http://platform.qbix.com
- gwu78 11y agoAs most nerds know, internet messages or "email", and later Usenet, preceded the www by a number of years. The original issue was opening different types of files from within an internet message. No matter how browser-centric today's computer users have become (thanks to marketing) the past remains close to the surface. As far as I can tell, the original solutions were never replaced; they were just ignored in favor of "middleware" (for lack of a better term) like Adobe Flash. In fact today's developers probably are very familar with the original solutions, though they might not recognize them as as such. I wonder how many could tell you where the Content-Type: header actually originated from? What is MIME? What is mime.types? What is a mailcap file? My browser of choice still has full MIME support so I can configure it to launch other, dedicated programs ("native apps"), based on file extension. For example, in theory I could launch ffmpeg or a video player (mplayer, vlc, etc.) from "within the browser" to transcode or play video. I imagine all the major video-serving websites used or still use ffmpeg in some capacity, not to mention any "web browser" that attempts to play video itself. Now the truth is I actually do not use MIME anymore. I prefer to download video first to my own localhost server using my own scripts and a program like tnftp, and then play it on any device I choose, with whatever "app" I choose. I like my video as a file that I can manipulate, not a pseudo "broadcast". As a side benefit perhaps, I have observed that with few exceptions I never see any preroll advertising when I download video to a file and then play it. The only time I have seen such advertising is when using a "modern" browser to "stream" (or progressively download) video. Not sure what this means if it means anything but I am happy to avoid the advertising.
- xtx23 11y agoso...concede defeat 'cause some engineers using the tech wrong or like the tech to be used certain? Who's to say how the web is supposed to be? Some comments above mention about loading 100+ assets for a Phonegap app...why would you load 100+ remotely when you are building a phonegap app? Every tech comes with a trade-off, at the level of dev-cost vs. runtime-efficiency. It also depends on what kind of existing code base and developer team you have to build the new product. A typical right or wrong clickbait like this one is not even worth the lengthy discussion bordering on random mass venting.
- jorangreef 11y agoBrowsers need to move from tens of fine-grained permission requests, to "Do you want to install this web app and give it full access to your computer?". And from there, they need to stop building abstractions such as WebSocket, IndexedDB etc. and start exposing raw OS capabilities such as UDP, TCP and FS. Basically packaging node.js would do the trick here. Also see Tim Berners-Lee in "Installing web apps": "If I can't give power to apps, then the web app platform cannot compete with native apps." https://lists.w3.org/Archives/Public/public-webapps/2012JanMar/0464.html https://lists.w3.org/Archives/Public/public-webapps/2012JanM... He asks for something fine-grained but the main point is to expose the real power of a machine to an app if the user so wishes. That's not possible today with browsers. Browsers need to stop controlling innovation, and push core services out to the edges and let innovation happen there. Stop drafting specs for WebRTC etc. and give UDP etc. back to the people. If the browsers don't do it, maybe someone will make a new kind of browser that will with something like Electron.
- methehack 11y agoDoesn't it depend a lot on what you're trying to do? Seems impossible to generalize to me. The win is huge if you have a web app that works well as a mobile app and does so on the same code base. And that does happen, it just doesn't happen every time -- but I don't think anyone ever said it was supposed to.
- hypertexthero 11y ago[Lets not](https://en.wikipedia.org/wiki/Wabi-sabi https://en.wikipedia.org/wiki/Wabi-sabi).
- kogus 11y agoI think this is a false choice. To a large extent, the web is native, and the browser is the operating system.
- perseusprime 11y agoIsn't a browser a one size fits all container that was never destined to last?
- netfire 11y agoCouldn't you say the same about the operating system in general? That is, that apps are limited to the the container provided by the APIs given by the operating system? The browser is only limited by the restrictions given to it by the operating system. There's no reason a browser can't run frame-less graphics-intensive apps with something like WebGL or why it couldn't have access to more of the operating system and hardware or function more like an app. The main workflow of a web browser (urls and browser history), don't seem any more restricting than having to use an app store to download an app or using the home screen to launch an app. (I would argue that they are, in fact, less restricting). The restrictions are a factor of the operating system, not a inherit restriction of a web-based workflow.
- perseusprime 11y agoWouldn't they then become emulators of app runtimes?
- netfire 11y agoIn a way, but there are advantages to running an application in a web-based platform: - Distribution is easier and doesn't require approval from a third party (such as app stores), providing less of a barrier to entry and more freedom to create what you want to. - It can be easier to develop and you may not have to provide support for multiple versions of your app (because you can easily ensure that everyone is running the latest codebase). - Common components/libraries that would be used across different applications can be part of the browser itself or cached for one app and re-used for other applications, instead of having to be packaged into a downloadable runtime application. - Depending on your application, creating a cross-platform product may be easier and require less development and have built-in functionality like copy-paste and other functionality provided by the browser (history, bookmarking, etc). - The user and other people have more power to use your application in more flexible ways (due to browser plugins), and the content can be crawled and made searchable/discoverable by a variety of companies. - The hyper-linking built into the web offers easy, cross-site (and application) integration that may be difficult to accomplish in a native app. In fact, you may be limited to a certain application provided by the operating system (like a particular map provider on mobile)
- oldpond 11y agoRemember when the browser was actually a thin client? Now it's got everything but the kitchen sink in it. It's funny how we just assume anything new will be in the browser. Innovation is happening at the native client now. Do one thing, and do it well. Talk to my server.
- thegoodlab 11y agoIt's all relative. At any moment in time there are going to be "native," fat client, devices that are running more powerful applications than that can be run on a thin client, "web," model. There is always going to be a transition to move those applications to a thin model, for scaleability reasons, and when that happens there is going to be some new software that requires the fat client model again. Either way, it's about figuring out which model works for the problem you're trying to solve. If you advocate for one, you're missing opportunities in the other.
- dgreensp 11y agoTelling mobile devs not to use web tech is an example of where this argument (request? demand?) falls down. There are a lot of advantages to using web tech (cross-platform, easier to develop, easier to push updates), and the result is not necessarily worse; it may be better. CSS is a very expressive and powerful layout language, for all its rough edges. Apple uses WebViews in their own apps (like the mobile App Store app). The rough edges of web tech will only get smoother over time.
- gedrap 11y ago>>> There are a lot of advantages to using web tech (cross-platform, easier to develop, easier to push updates) Yeah. This shouldn't be overlooked. In some cases (solo founder, bootstrapped, etc) a hybrid app might be a great choice for v1.0. Sure, it won't be fast on the phone... But it probably will be pretty fast to develop and release on multiple platforms. Especially if you don't have experience in iOS/Android dev but plenty of background in Web. Sure, if it gets traction you should rewrite it in mobile at some point. But as a prototype, I am totally cool with hybrids. I'd guess these are not the cases that the OP meant, but still I wouldn't say that the web lost. It has fair share of use cases.
- dgreensp 11y agoI think quite a wide variety of app devs might find it more important to be able to, say, develop an app faster than have a snappy start-up time, especially if there's no fundamental limitation at play but just limitations of the current tools! I trust app devs to choose the best tools for the job.
- piyush_soni 11y agoThe article, when talking about the simple 'Technical' reason why web can't beat the 'native' falls flat. When most people talk about 'Web', they just mean 'Browser' - and a browser is native. The extra layer, yes, the rendering engine/JS framework - how are they very different from JVM and XAML in theory? Both add an extra layer. Are Android apps not native according to the author because it adds a layer too?
- bcg1 11y agoI think that a title of 'Browser vs. Native' instead of 'Web vs. Native' might be more accurate, since the web is so much more than just a UI... most 'apps' I think are just web apps with a 'native' GUI, not disconnected applications in and of themselves. PPK ignores the original reason that the browser won so long ago though... it is the universal GUI for the Internet. We've had access to 'native' extensions for web apps for a long time (Java applets, Flash, Silverlight, ActiveX, etc, etc), yet these things are all but dead. Eventually the Apple vs. Android war will probably go nuclear, and users will demand that things 'just work' too. If its time to revisit this conversation, then fine... but there is nothing new under the sun.
- kenjackson 11y agoThe web already won. When was the last time someone told you about an app idea and you had any vision of a Win32 (or WPF) app? The web is the defacto app platform for desktop today. Sure, there are some legacy and niche apps that are native on desktop, but mostly everything new is web based on desktop. Mobile is a different story, but I think its due mostly to the fact that mobile HW is still changing rapidly. New sensors and other capabilities to tap into, where a standards web has to catch up. But I think this gap too will go away as mobile HW matures.
- ghostly_s 11y agoIf someone tells me about an "app idea" without specifying platform I assume a mobile-native app. I do not think I am in the minority in this.
- vezzy-fnord 11y agoWhen was the last time someone told you about an app idea and you had any vision of a Win32 (or WPF) app? What does a specific widget toolkit have to do with native applications in general? Sure, there are some legacy and niche apps that are native on desktop People still use office suites, raster and vector graphics editors, CAD, 3D modelers, meshing and texturing tools, media players, P2P clients, text editors, various format readers and writers, etc. that are largely dominated by native. Even most non-HTTP network clients (like IRC and FTP) are still predominantly used as native apps because of the far superior experience they bring.
- kenjackson 11y agoYou went on to list a bunch of niche app classes except for two: 1) Office suites --> Largely a legacy app suite... and web use is increasing for both Office and due to Google Apps. 2) Media Players -- Playing local media is done by native apps, but that is becoming an increasing niche scenario. I posit that Netflix, Pandora, and YouTube dominate other media players. The most important native app that most people use is their web browser. :-)
- 11y ago
- anthony_romeo 11y agoI've always felt that (with respect to desktop work specifically), the reason we're in the web-app world we are today was in part due to the stagnation of window management in the 90's - 00's Windows era. Tabbed browsing for web browsers was created pretty much to overcome having a gluttonous and unusable taskbar after opening just a few applications. From then on, I feel users were primed to want an alternative, which web apps ultimately gave. Further, to talk about the speed hit associated with web apps, laptops and desktops are still primarily sold with old-school hard disk drives instead of solid state (unless you pay a big premium, which most people don't want to do). Frankly, opening a native application on an older machine (and for most users, one which hasn't been defragmented even once) was an awful waste of time. But Google Docs? Heck, you didn't even need to be on the same computer to access those files. Not to mention trying to update these apps being a massive chore on these slow machines. Heck, most people didn't ever update their browsers until Mozilla and Google imposed it upon users. While I'd like a world where the browser is not an abstraction layer for an OS, I do recognize that these web apps exist for a reason.
- fixermark 11y agoI wonder how many of the issues listed in this article could be addressed if my HTML-based web technologies allowed for a real concept of an "app?" Right now, we're very much faking it. Every time you go to a website, you're downloading a document that unpacks into code and data in your client. Caching aside (because the cache timeouts are never long enough), you have to do that dance every time you visit. The performance hurdle ppk describes is less about having to go through the DOM to get to your elbow (though that's true) than it is about never being able to precompute and locally cache view logic. That's a performance hurdle that can't be crossed. In contrast, native apps divide everything quite clearly into "code" and "data." The code is the app you download from the store; it can auto-update, but it isn't trying to pull in a new copy of itself from scratch every time you use it. It can also be pre-optimized for your architecture. The data is only what is needed for the app to present it appropriately to you. I'd be willing to wager that if a real standard for downloadable client-side cached web apps ever caught on, ironed out the optimization issues, and stuck, quite a few of the performance issues of web vs. native could actually be addressed. (Of course, we'd still have to deal with the DOM really being the wrong language for a lot of high-interactivity use cases, unfortunately).
- kuma4 11y agoWeb has some advantages and remains important. But HTML/CSS/JavaScript will decline in the future. They are limited compared to native app technologies. Web will be used mostly as a JSON API.
- kuma4 11y agoWeb has some advantages and remains important. But HTML/CSS/JavaScript will decline in the future. They are limited compared to native app technologies. Web will be used mostly as a JSON API.
- javiramos 11y agoI consider Onshape to be a big win for web: https://www.onshape.com/ https://www.onshape.com/ It is a full-fledged CAD design software (think Solidworks) that runs on the web.
- feld 11y agoFastmail's mobile webapp is a webapp done right. It feels like a native app that is a natural extension of the full-featured website -- but it isn't native. And it's blazing fast.
- orthecreedence 11y agoI agree that if you have the resources to build something on native, go right ahead. But as someone who has very limited time to work on a passion project (https://turtl.it https://turtl.it), I will be damned if I split up my efforts writing different apps for iOS, Android, and desktop. Thanks to things like Cordova, node-webkit, and various other tools, I can write my interface/logic once and run it everywhere (with a bit of wrapping). No, it will not be perfect. Yes a native app would perform better. But I don't think my users (who continually ask me for a mobile app) are really going to care that much if the page transitions jump a tiny bit. And once Safari pulls its head out of its ass and supports AES-GCM in its WebCrypto implementation, I can do fully-native crypto in my app, and the largest performance hit goes away. HTML5 is the poor person's write once, run everywhere. Although not perfect, it continually strives to be better. It enables someone like me, who can only work nights and weekends, to build something useful and usable for a lot of people in a minimal amount of time. If people don't like Turtl mobile because it doesn't have seamless transitions, they can use Evernote instead and have their plaintext data leaked all over the web.
- BadassFractal 11y agoI use the web in big part because I want to avoid the week (or sometime two-three week) long gauntlet of going through the Apple submission process.
- pbreit 11y agoI mostly agree but the one category that does NOT need to be native is pure (text-heavy) content (i.e., news sites). In fact, it's beneficial to be web in order to support easy linking, distribution and discovery. Also, while I agree with the "home page" question, the author makes no mention of the other important criteria: are notifications important?
- downandout 11y agoIf we "concede defeat" of the mobile web, then we wind up with only content and functionality that has been approved and can be killed on a whim by Apple and Google. Wars have been fought over freedom of expression, and now we have voluntarily given two companies and the governments they answer to the power to crush anything they don't like, based upon cryptic and arbitrary guidelines. That is a truly scary world. Like many things, native is perfect in theory but in practice comes with massive drawbacks. A happy medium may indeed exist, but "conceding defeat" of the mobile web isn't going to get us any closer to it.
- mathattack 11y agoHow much of this is due to the # of native technologies that need to be developed for? In Mobile there are 2. Web wins over 20 competitors, but just 2?
- carsongross 11y agoI broadly agree that trying to completely mimic native apps with the web is unlikely to yield great results, and the author points out some of the obvious strengths of web applications (in an update he mentions URLs, which he missed initially. https://www.youtube.com/watch?v=i28UEoLXVFQ https://www.youtube.com/watch?v=i28UEoLXVFQ). I think he underestimates the web, though. In particular, there is one aspect that he fails to mention and that I view as one of the best aspects of web development: the stateless request-response-replace-UI model of remote application development. This model is incredibly simple for developers to understand and work with, and eliminates a whole slew of complex problems found in traditional client-server development. (I would note that these problems are being reintroduced with client-side web technologies like Angular and de-reintroduced by technologies like intercooler.js.)
- krick 11y agoI pretty much agree with the post, even more — I like to agree with it. But there's a problem. The problem is there's no stuff like "web/native app": there's a bunch of functionality that solves user's problem to greater or lesser extent. Although there are needs which are technically complex, usually what the user can ask for is pretty simple. Clean UI, button here, button there, some lights and pop-up notifications, a couple of graphics, which actually represent something pretty trivial internally. Sometimes the data needs to be pushed to or retrieved from some centralized storage or it needs to talk to some other "service". "Browser" is absolutely awful and pointless invention, but unfortunately we have accepted it somehow and learned to satisfy the needs of typical customer using it. HTML might be stupid and CSS might be inconsistent, but it's actually pretty easy to make the UI user wants with that and a handful of js code. As a bonus, it will work on Linux, on Windows, on Mac, your Android phone and even some not-a-real-OS which happens to have a capable browser. And what we have on the desktop? I don't know a single good UI framework for Linux ("good" in a sense it would make it easy to achieve what I want). There's somewhat usable WPF for Windows, which still doesn't really allow to implement anything user could wish for as easily, as I'd do with HTML. Somewhat usable, but pretty verbose is Android UI framework. I don't know what happens on Mac. And, yeah, none would be compatible between different platforma (except Qt, maybe, which is overly complicated by itself). So there isn't much choice between "web and native". The "native" is currently pretty much non-existent.
- zokier 11y agoHave you checked QML? It is kinda mix of "native" and "web".
- krick 11y agoVery superficially. I haven't completely understood how much better it is than the standard layout, but I doubt that it can be comparable with what we have today in a browser. I mean all the libraries, no need to compile anything, writing components in a general-purpose high-level language… Even making Qt friends with Python isn't as effortless as it could have been. And it isn't the real thing anyway: in order to be a good platform for building GUI apps all that functionality should be directly provided by the DE in the first place (like it is in Android, for example). You cannot play with representation the same way in (say) xfce, as you can in the browser console, moving containers and changing colors as you like.
- z3t4 11y agoSo the customer/users want a native app!? Fine, I'll just bundle the web app with Webkit (io.js) and they'll be happy, and I don't have to worry about the app breaking in new browser versions. Browser based applications are much easier to distribute and adds a security layer, so that you do not have to worry as much about trojan/viruses, compared to installing executables from cr4cksnh4cks.com The browser is the perfect container. And it also helps a bit that the API's are super easy to work with. Making even the novice programmer productive and able to create powerful applications. One reason why native is bigger on the smart-phone market is that Google, Apple and Microsoft want to completely control their ecosystem and lock people in. What I can not understand though, is why Google does it, as the majority of their income still comes from the word wide web. They have however made major advances on that front lately, to there is still hope. As the virtualization technology advances, the gap between running on bare metal and in a container continuously decreases. So there will be little incentive running native/bare metal in the future. We have a long way to go though. If you want to build something that make full use of the hardware, you still should go native. But I hope that changes in the future.
- wvenable 11y ago> And it also helps a bit that the API's are super easy to work with. Not in the least, you have to be comfortable in a half-dozen technologies at least to make a half-decent web application. You want performance, animations, etc then you better be an expert. And even then, your perfect container is a limited container. By comparison, building a native app is much simpler. The complication comes in with cross-platform applications.
- drew-y 11y agoI disagree. I'm still a fairly new developer and I've tried developing both native and web apps. I found that there are thousands of great well documented frameworks that make developing web high performance web apps easy. Meteorjs for example took me about a weekend to learn, has a huge community with lots of prebuilt packages, and countless tutorials that make implementing smooth animation incredibly easy. Before Meteorjs I used Angularjs and found mostly the same thing. I'm sure it's similar with React or even just standalone Javascript Html and CSS3.
- sixQuarks 11y agoIf native won, then why didn't "native" desktop software win? Everything on desktop has moved to SAAS and on the web. We're being way too short-sighted here. In 5-10 years, I don't think smartphones will be the primary way we get information anymore. Yes, we will have mobile, but it might be mobile VR, with an infinite canvas. I think the reason native works well on mobile today is simply because there isn't much room to work with. Once the viewports start getting larger again, the web will take over.
- JoeAltmaier 11y agoDisagree. Currently I'm running Audacity, Sococo (network conferencing tool; soon to be a web app), NetMon, Outlook, WeatherEye and a few services. Also have 3 browser tabs open. So from here it looks like native apps are winning. Especially since I despise two of the three web apps (TeamCity, Jira) for being laggy, buggy and very slow.
- sixQuarks 11y agoOK, but are you a mainstream consumer? No, you're a developer and have specific needs that are solved better with native desktop software.
- JoeAltmaier 11y agoTrue, but ALL the web apps I use are laggy, slow and annoying. Its unlikely that developers have the only bad web apps.
- mike_hearn 11y agoIf native won, then why didn't "native" desktop software win? Everything on desktop has moved to SAAS and on the web. That's a common belief but isn't correct, of course. Otherwise ChromeOS would have crushed the competition and it's hardly made a dent. Many, many people use and heavily rely on native desktop apps every day. Usually for getting real work done once they tire of procrastinating on hacker news/facebook/etc :-) Just consider the importance of MS Office to the global economy! This is despite the fact that the shepards of the native desktop app platforms all screwed up big time and dropped the ball in major ways allowing the totally unsuitable yet somehow still good HTML platform to take over: • MS/Apple/etc STILL have never shipped non-terrible software update mechanisms. Only Google managed to get this right and their online update engine was basically just for them. • No discovery or sandboxing mechanisms worth a damn. • Heavy reliance on languages like C++ that many developers can't handle (whereas web was fully built around scripting languages) • The only people who cared about doing cross-platform desktop apps were Sun and Trolltech, and they sunk into the swamp of pretending to be the real OS-specific widget toolkits. Web apps collectively said "fuck that shit" and we saw a glorious explosion of styles with every web app branding its own widgets differently. Users, it turned out, wanted attractive apps rather than native-looking apps and so the web won on this count even though the HTML widget toolkit is a joke. • Speaking of widget toolkits, the web is so primitive that it forced developers to abandon UI design paradigms that were actually rather user unfriendly, but commonly used because the platforms made it so easy. Prime offender: tree widgets. Hardly ever used these days as they have terrible usability. HTML has no such widget. Also: context menus. Users never find things in them. HTML simply doesn't let you customise the context menu at all. So the combination of uber-simple programming languages, a cross platform UI story that ignored the massive must-look-native timesink, clean/invisible online updates, etc ....... almost by accident it ended up being compelling. Now compare to mobile, where native apps are doing a lot better. Android is Java based so no problem with C++ being unapproachable. Objective-C kind of sucks but is still easier than C++ in many ways, plus iOS is big enough that developers just have to suck it up. There's built in sandboxing and app updates have been a part of the platform right from day one. The widget toolkits are way more modern and result in simpler/more intuitive UIs than classical desktop toolkits, so dysfunctional design patterns like nested trees with tiny icons don't crop up. The toolkits almost force simple, intuitive designs. And because mobile is (a) very new and (b) very restricted, writing the app twice with two different user interfaces isn't necessarily a killer problem, whereas for gigantic mature desktop apps with a 20 year feature set like Photoshop/Office it's just really expensive to do that. So the not-quite-native issue is less prominent.
- superuser2 11y agoIf we do this, then we need to pick a platform. There is enormous replication of effort in creating "native" experiences for all of OSX, Windows, Linux, iOS, and Android, probably worse than what we spend on replicating native tooling in the browser.
- piluyas 11y agoI read this post just in time!... now I'm am sure, the future is create a new PornHub for android, and other in xCode for Iphone, and for nokia, another one, and what about Blackberry, sounds very portable! is a remarkable idea!
- lewisl9029 11y agoRehashing part of my comment from a different article: I don't see why any reasonable developer would choose building the same app on multiple platforms over building it once on a standardized platform and having it just work everywhere. That is, of course, if the standardized platform had all the capabilities and great UX of the proprietary platforms. This is where the web often falls short, and the companies behind proprietary platforms have some very sweet incentives to keep it that way, i.e. developer lock-in and maintaining complete control over the platform's direction. Despite all the impediments, I still firmly believe an open platform will triumph eventually, because developers love efficiency, and the current landscape of wasting work on rebuilding for proprietary platforms is anything but efficient. In fact, the web is already capable of competing with native in terms of general UX, not yet quite enough to match or surpass it, but definitely within striking distance. Just take a look at some of the apps made with Ionic: http://showcase.ionicframework.com/ http://showcase.ionicframework.com/ Additional tip for for Android Cordova development: bundling an external WebView like Crosswalk makes quite a bit of difference for <5.0, both in terms of performance and in terms of not having to worry about legacy WebView quirks for older phones.
- s73v3r 11y ago"I don't see why any reasonable developer would choose building the same app on multiple platforms over building it once on a standardized platform and having it just work everywhere." Because it provides a better experience for the user, and because it allows for access to hardware native features.
- lewisl9029 11y agoYes, but that's a side effect of native already having better UX and more capabilities. "That is, of course, if the standardized platform had all the capabilities and great UX of the proprietary platforms. This is where the web often falls short, and the companies behind proprietary platforms have some very sweet incentives to keep it that way, i.e. developer lock-in and maintaining complete control over the platform's direction." Given the choice of a standardized platform that offers the same UX and features, would anyone still choose to develop for proprietary platforms? That ideal standardized platform doesn't exist right now, but the web is getting closer and closer to being that platform. It doesn't make sense to give up all the progress we've made so far and resign to native when the web is already almost there.
- bcheung 11y agoI strongly disagree. Why wouldn't we want to continue making progress on the web? Where would we be today if cavemen decided spears are good enough so let's just stop there. I think the web should continue innovating and that trying to compete with native apps is a worthy goal. The web is the most compatible and universal format. There is a lot of benefit to having a universal development platform. With Node.js the web has been brought server-side as well. There is a huge convenience factor to being able to use 1 platform for browser, server, and mobile. I don't like being locked into a vendor or being uncertain if this platform will be around 5-10 years later. Having the web around as an app platform creates certainty that encourages commitment from companies.
- meesterdude 11y agoI'm a web developer with a side project. I don't have the time/energy/knowledge/bandwidth to implement native apps for IOS and android. But I CAN make it work well enough in a mobile browser. And lets not forget: developing for android is hard. You have a variety of models, with varying versions of android, running on different carriers; i've read some horror stories, and no thanks. I'm quite quite happy developing for the web and keeping it on the web. I don't have to answer to any gate keeper or ask for permission for a new release. I don't have to share 30% of my profits, or be careful about implementing native functionality. They just need a browser that can render HTML/JS/CSS. That browser will only get more optimized over the years, the network will only get faster, and the technologies to optimize the web for mobile will only increase. I say more web, not less.
- ajkjk 11y agoThings native apps do well: - Run offline - Access large files stored locally - Start from the command line or from clicking buttons on your desktop - Use graphics cards and other peripherals - Run quickly - Tune performance to your platform - Come pre-installed on your system - Exist for all time - .. and don't require servers and money to host them - Security is pretty easy to solve - Coded in pretty much any language - Interop with other native software - Run in windows or fullscreen - Easily convince people that they ought to be purchased - Obfuscate their functionality from users [important for, say, video games] - Handle large amounts of data without network bandwidth constraints - Standardized UX, so it's less easy to screw up. - .. and when you do get an app running, you get a lot more UX for free [well, depending on what you're doing] Things webapps do well (and sometimes get for free) that native apps do really, really badly: - Save data across multiple systems - Allow you to login anywhere - Get off the ground with a UI + basic functionality in O(1) lines of code. - Receive updates from their maintainer - Run multiple instances at once with a single click - Are completely sandboxed from other apps (app freezes? just refresh the page) - And are sandboxed enough that you can pretty much trust them (barring tricking you into downloading viruses / exploits, etc, you don't have much to fear from a website,, at least compared to running untrusted executables) - Live in easily manageable tabs - Are easily augmented using scripts (Greasemonkey, etc) or simple scripting (JS in the URL bar) or plugins (browser extensions, etc) - Are globally namespaced and addressable and bookmarkable. Every webapp runs in the same container so other programs can manage them in a uniform way: via bookmarks, hyperlinks, etc. - Support an intuitive format for entry points into the internals of an the application, via URLs. Compare to: save files for a video game. - Must forcibly reveal their source code (except for Flash & similar plugins) to the client, so it's hard to deceive [technically savvy] users. - .. and are largely open source by default (except for Flash & similar plugins) - .. which means anyone, even someone who's never coded in their life, can learn to peek under the covers and see how things work. - Use global identities (logging in with Google/FB/Twitter/etc) and sometimes global configurations (Google cloud sync, etc) - Elegantly support collaboration (send someone a Google Doc link and you're done) - Standardized ways of handling authentication and security - Default implementations for hotkeys, linking, text editing, scrollbars, image insertion, etc. - Default implementations for caching of resources, shared libraries, and asynchronous messaging [sort-of] - I'm sure there's lots more. It seems to me that web browsers accidentally solved a very general problem with a very specific solution. -- The problem: how do I run software that's dynamically loaded over the Internet with minimal barriers to my users? (Well, initially, it was 'how do I display text from the Internet to users', but we've moved passed that.) The solution: I'll build an application that interprets a single programming language (Javascript) with a single standard UI language (HTTP+CSS) and supports a global namespace of all other apps (DNS,etc) and allows data to be encoded in resource identifiers (URLS) and maintains state across sites and across sessions (cookies), and I'll call it a Web Browser. This relatively arbitrary solution got a lot of things for free, but it left us with a lot of implicit constraints that divided the line between web and native applications. Ever since then everyone has been trying to break back out: adding tabs, isolating tab processes, supporting extensions, embedding Java, embedding Flash, supporting SVG and Canvas and WebGL, etc. We're still largely trapped in the interpreted-Javascript world.. but Dart gets points for trying, and alternative languages/paradigms (stronger-typed variants like TypeScript, from different programming paradigms like React, and from entirely different languages (asm.js, GWT..)), are gaining traction. Meanwhile, AWS S3/Lambda/etc, Dropbox-based storage, Google cloud sync'd settings, or global credentials via Twitter/FB/Google/Etc.. these are all reducing the amount of work you have to do to build and maintain a simple app in a website. I wish I could use Dropbox or Google Docs as my 'filesystem', if that made sense, or use the local filesystem as my filesystem, if that made sense instead; I wish I could use local keyfiles or global accounts at will, or build low-level compiled libraries and mix them with deployed, interpreted website code at will. I also wish my window manager and my tab manager were the same application, and my Omni-bar searched my desktop and the entire internet, and my stickynotes were synchronized on any computer I log in to. And I also wish I could write a native (but sandboxed, except for the permissions I grant it) application with a UI that interops with my OS and other programs by simply writing 10 lines of code and then saving the file with a .html extension. Basically, I don't particularly want to write native apps; I just want to casually slide back and forth over the line between native and web.
- eighthave 11y agoGreat article, I'd just like to point out that URLs by no means only work on the web. Android has deep, native support for URLs in apps, and are frequently used for inter-app communications that have nothing to do with the web. Another OS could easily add such support for URLs.
- nickpsecurity 11y agoAbout time they see it! I've been pushing native, Internet apps over web since the beginning. The reason: native apps can be faster, leaner, more flexible, more portable, and more secure. Most things I can do to boost one of these metrics in a native app I can't do in a web app. Let's put this in perspective: my 1998, 200Mhz P2, 64MB RAM machine's native apps did more than modern web apps and with fewer problems. And could connect to people and use presentation layers... just like the web! One writer argued the reason the Web took off was a standard, usable presentation format (HTML) combined with easy distribution (hyperlinks) and instant updates (connect for freshest version). This spread across both users and platforms. Its continual improvements gave it new features. To this day, the standardization and easy distribution makes it easier to push content to most users via a web browser. Yet, even big Web companies such as Facebook are switching to Internet-style apps for mobile and Internal use. That's saying something. I guess they don't want around 500 bytes of HTTP overhead to send 1 byte of data. The best solution is a portable platform that solves the presentation, linking, and updating issues that web did. There's a number of older vendors and startups doing this. Hopefully we'll see more take-up and network effects on that side of things. Note: Anyone still thinking the Web was the best route to go, but wanting native advantages, should Google Tannenbaum's Globe distributed toolkit that supplemented Amoeba distributed OS. They even ported the Web to run on top of it and side-by-side with its applications. Compare work like that to the web of the same time period and you'll see why Web is inherently inferior due to bad (or misappropriated) design. It always will be to a comparable, native product.
- sclaxton 11y agoThis post is right, but only because web technology is currently a second-class citizen on almost every native environment. If you were to expose system API's directly to the JS layer à la FirefoxOS, then I'm pretty sure you could recreate most "native" apps on web based platform with minimal / narrowing performance hits. Booting to browser technology can only get better with projects like ChromeOS and FirefoxOS pushing the envelope, so why not push for a unified application runtime in the web across devices? Sure, html wasn't made for rich applications, but html5 was...
- RRRA 11y agoGiving up on the Web, is: 1) blind to all the progress we've made closing the gap with all the other platforms 2) Irresponsible to the future by basically saying "let's drop the only open and common platform we have and let's embrace the corporate fragmented world, let's relive the 80s all over again just because of some short term performance issue!"
- click170 11y agoAs a user, my policy is simple. Unless there is functionality in the app that I want to use that requires direct OS access, the app doesn't get installed, period. I care about my privacy and I won't install your app just to get a slightly snappier experience or because you didn't want to make a web version. There's way to many "apps" that don't deserve to be anything more than a website and I take a hard line when it cones to my privacy. Ask yourself if this can be done in a website. If the answer is yes, then why aren't you?
- slvrspoon 11y agothis is crazy and obviously wrong.
- shiggerino 11y agoThe biggest problem with this HTML app craze is that the tooling is abysmal. People still debug JavaScript by looking at trace output, are you fucking kidding me‽
- minusSeven 11y agoOnce upon a time I would choose native over web anytime. But my experience in native has got worse over time. Now I don't even feel like it. I don't want to install 2 dozen apps when I can simply go to that website directly. Native if anything has been a failure.
- hlfcoding 11y agoThis coming from a guy whose website hadn't been made responsive.
- AnonJ 11y agoThis argument is really rough. Many programmers surely will only want the content, but websites are interacting with the whole world, and they have business models which cater to the general public. If they only exist to serve simple, static contents then the companies running them might add well just shut up shop. Why even bother spending expenses maintaining such stuffs if they don't appeal to the general public and easily call them to actions anyways.
- MidoAssran 11y agoInteresting points you bring up. A point to consider is that many companies, even surprisingly many of the Fortune 500s, have very dedicated IT teams to support a web infrastructure, and so choose to make their applications using web frameworks to emulate native. I give you that it's not ideal, but when looking at the bigger picture, it just makes sense to use web rather than to force them to spend months getting ready for native support; so there might be a place for each use case.