Slashdot Mirror
Web Apps: the Future of the Internet, Or Forever a Second-Class Citizen?
An anonymous reader writes "This article takes a look at whether web apps will ever match desktop and mobile apps in terms of performance and usability. Jo Rabin, who's leading the push by web standards body W3C to get web app performance up to scratch, is optimistic web apps will eventually be the default choice for building the majority of commercial and business apps, while the article weighs up just how much web technologies need to be improved before this could happen. Quoting: 'Native apps are generally first to gain access to new platform-specific hardware features — be it navigating using a phone's GPS and accelerometer or taking pictures with a phone's camera. But if a particular hardware feature becomes popular, standards to implement that feature in the browser will always follow, Rabin said. Work is taking place within W3C to standardise APIs for web technologies to access many of the features found on modern smartphones. Ongoing work this year includes setting out a system-level API to allow a web app to manage a device's contacts book, a messaging API for sending and receiving SMS and MMS, new mechanisms for capturing photos and recordings, new event triggers that could handle mouse, pen and touch inputs, a new push API to allow web apps to receive messages in the background, new media queries for responsive web design, an API for exchanging information using NFC and precise control over resource loading times in a web document.'"
iOS/android app: installed base on day one: 0
web app: installed base on day one: a hundred million +
Do you mean a website?
Do you mean a website that has been optimized for mobile devices with a small screen and minimal javascript capabilities?
The reason the internet became popular is because of standards. Any web browser can connect to any web site, issue client http requests, and get an http response.
The web browser/web server is the better way to go. I don't need to download & install a program on to my laptop to bank online. I don't need to download & install a program on to my laptop to read the news.
I just just my web browser. It's the better model.
GRRR, this is STUPID!
This drives me mad. Webapps have pushed usability a good 20 years back.
Look at all the pb in desktop apps: we still don't have it "right", and now, there is an increasing tendency towards webapps that have even worse choice of widgets, and don't respect many usability issues that desktop had barely started solving... And don't get me started on the browsers compatibility issues!
This is CRAZY! People are building GUI inside a medium that has never been designed for that, using always more complex layers for a worse user experience (and don't get me started about the performances loss).
The primary problem is not JavaScript, but that so few web apps are licensed under AGPLv3+ and thus are not suitable for any kind of serious use.
From the article: "...today web browsers are capable of supporting sites that are getting close to the look and feel of apps we run directly on our phones and computers."
Great. So with a lot of work we're almost BACK TO WHERE WE WERE! How about some innovation and making better, more usable UI's rather than just trying to catch up with what we already have?
This is the same problem I have with Linux - constantly reimplementing things we already had. At least Linux is replacing closed software with open software. What's your excuse with webapps? Is MS Word in a browser inherently better or just different?
Not as long as we have to write them in javascript. javascript might be an ok language, but it is not designed to help with development of large projects. The type system, and the prototype based objects are not good at doing that.
It seems like some of the problem here is that Javascript is the Lingua Franca, and also that it has to use the web pages DOM. If the system were being designed from scratch, it seems unlikely this would be the choice.
Javascript is a very sloppy language for doing anything in. People, like me, continue to write in it because there's little alternative.
Writing in javascript drives people insane.
sent from my padded cell
A feeling of having made the same mistake before: Deja Foobar
There is a big disconnect here. Using HTML 5 to implement an app interface on a smart phone is not the same thing as implementing a general browser web interface using the same technologies.
Heaven forbid a web interface should ever have data to manipulate anything more than the cookies it needs in my browser. That would be a security hole you could drive a whole fleet of trucks through.
However, I don't believe that web interfaces will ever equal custom client code or custom apps for the simple reason that you get hesitations and delays during page and AJAX refreshes. One of the worst culprits for this is trying to implement drop down choice boxes that adapt their contents to previously selected data, such as country-state interactions. The only way I know of to do that with a web interface is to refresh the whole page, which is obscenely slow compared to the repopulation of the choice box data itself done by a custom interface.
There is also no way to perform performance tuning and UI tricks like dynamically making widgets visible/invisible with a web app, something that is very common to high performance custom interfaces. In part, this is because web apps don't have the necessary layout management interfaces that a custom application does, which allows them to position those hidden widgets appropriately so that they overlay each other to the pixel.
I do not fail; I succeed at finding out what does not work.
Apropos of the question in the title of this post. I had a meeting today with a guy in his early 20s. I mentioned that one of my current projects is a Java Client/Server application. He found this totally bizarre, because "Web apps are the future".
Now, I'm a geezer by IT standards. I cut my teeth on an IBM 360 (yes, that old). One of my standard charts that I use in general IT presentations is a spiral. I'll do it here in text, a bit oversimplified:
The pendulum swings back and forth, but you only start to recognize the pattern after you've lived through a couple of cycles. In fact, it seems that by the time one trend has established itself as inevitable, the next (opposite) trend is already well underway. Right now, Web-apps and Cloud computing are the buzzwords, but mobile computing is already well underway for dominance by 2020.
So, if I may answer the question posed in the title: Web Apps: the Future of the Internet?
No.
Enjoy life! This is not a dress rehearsal.
Seriously, to add some content to my snark. Until internet speeds reach parity with access to local resources, web apps will always be second class compared to a similar application running locally.
file:
iOS/android app: installed base on day one: 0
web app: installed base on day one: a hundred million +
What are you getting at here?
Let's say you launch a new website/web app. How is your install base not ALSO zero?
It doesn't matter how many devices CAN run what you have built. What matters is, WILL they...
With either an Android or iOS app, your POTENTIAL install base is in the hundreds of millions. Furthermore, there's a huge increase in the chances of you getting paid for some of that work, which you can use to develop it further. With a web app chances are good you launch it, zero people notice, and you fade into obscurity.
Lastly, compare how many applications there are in the iOS/Android app stores, vs. how many web sites exist. Which has better odds of someone using what you have created?
"There is more worth loving than we have strength to love." - Brian Jay Stanley
It seems like some of the problem here is that Javascript is the Lingua Franca, and also that it has to use the web pages DOM. If the system were being designed from scratch, it seems unlikely this would be the choice.
I think there is a Peter Principle for technology: Every good, proven technology gets extended, extended, and extended until it breaks. Back in the day, the web was a fast, simple, stateless request/response document retrieval system. Then some markup was added to make the documents look pretty. Then CGI and server-side processing were added to make the content more dynamic. Then tricks were applied (eg- cookies) to break the statelessness, albeit in a limited way. Then the prettiness of documents were improved with CSS and liquid design. Things got slow. Then the scalability of the server side was improved with various frameworks like JSP, Struts, Rails, etc. Applets had been tried, but they didn't take off because security and version control turned out to be much bigger problems than anyone realized at the time. Then JavaScript was extended and made to do all sorts of unsecure things. AJAX came along and broke the request/response document retrieval model for good. Things got slow again, especially on newer mobile platforms. Now the w3c wants to figure out how to run fat applications on this platform, ostensibly because it's shared in common across OSes.
Rinse, Repeat. Everything old is new again.
The answer will forever be NO, until the bandwidth between me and the server == the bandwidth between my CPU and its main memory. AND latency, AND availability.
The fastest web apps TODAY are still slower than the native apps that shipped on the first iPhone in 2007.
Besides that, both kinds have their own unique advantages. The first two: web apps allow you to get at the data from any device with an Internet connection and browser, but native apps work when you don't have network connectivity at all. If neither of those is a make-it-or-break-it aspect of your app, then you look at all the other advantages each offers, which have been discussed ad nauseum elsewhere.
Dear Slashdot: next time you want to mess with the site, add a rich-text editor for comments.
Most of the criticism I see around javascript comes from people who know almost nothing about javascript, much less writing quality javascript. Read "Javascript, the good parts", and js becomes an extremely powerful and flexible language. Trying to cram OOP into a language centered around functional paradigms seems to be the source of most people's issues with the language. As for the DOM... the DOM is a pretty sane system, until you need to support 3+ browsers of differing quality. The language is really not that big of deal anyway. Look at iOS, which uses a language many would say is no where up to snuff with other modern languages.
I've wondered if there could be a scheme where you have an app that in installs say, the sproutcore framework, server, and a minimal web server on a device. Further, use a rendering engine in such a way that you would not have to run it in a full browser. It would look just like an app and you could use it offline. Of course there is a matter of how to make something so crazy secure. Then I suppose the question is why do it at all. Fact of the matter is: I have no idea what I'm talking about. Any input from someone who does know what their talking about?
Brought to you by Carl's Junior.
Whether you want to call them web apps, thin clients, or terminal based apps - the premise is the same...the name just changes over time.
Web apps will succeed where users can tolerate the inherent limitations of the basic network-based thin-client app paradigm and are content accepting a limited or common native hardware feature set.
Native apps will succeed where people want the highest possible performance and instant access to the latest native platform features and are willing to sacrifice cross platform capability.
Yes, there are hybrid tools that try to bridge the gap...some are better than others. Still, you lose something in the translation when you try to be a jack of all trades and master of none.
To get the best of both worlds would imply that innovation has to stop until manufacturers all implement feature X into the common hardware feature set and the hooks in the "browser" to access them. And, you will need a persistent storage model that will allow web apps to have the same sort of performance that native apps possess. Only where the native app and web app are storing data on a server will that metric ever be even remotely close to a native, persistent store (unless we have FTL wireless networking).
I recommend advocating web-based apps and/or native apps when they are appropriate to the requirements from both a business and user perspective. Unless you can do that, you are little more than a fanboy for whatever paradigm you have locked yourself into.
Best part of it all? You can make money regardless of which way you to choose to go...Web, Native or Hybrid. That's what it's all about, right?
I develop web apps for the enterprise, and have for well over a decade. In the past few years tools like AJAX and HTML5 have made the user experience much richer, but in general even these weren't needed for web apps to take over the enterprise. Bulky, finicky fat-client apps were always a pain, which was why technologies like 5250 and VT220 were popular so long past their expected life spans. Many many businesses still run back end systems with these interfaces because they are simple to deploy and simple to maintain. Web apps are the best of both, and performance is more than good enough these days. For games, maybe not, but for the enterprise we're already there.