The future of web-development (RIA vs. traditional HTML) [closed]

Question

How do you see the future of the web development? will HTML, CSS and Ajax continue to lead the web-development or do you see a shift towards Rich Internet Applications (flex, silverlight & JavaFX)?

I am not looking for a clear cut answer, and I know you are programmers and not prophets, but a smart analysis of how do you see the current trends in web-development would be appreciated. Links to such debates on the web are also most welcome.

I am asking this question since we are now evaluating technologies for a complete rewrite of our GUI. Since it's a relatively big (actually huge) product, we tend to do things slow. We need to consider where do we see the web is going to.

I am interested in the near future (3-5 years from now).

Anlage

Javascript + AJAX counts as a RIA environment in most places, just more browser dependent and sometimes a bit slower. – Housebreak

Answer 1

The trend in web development is to move more towards the client. This has several consequences:

• As the codebase grows on the client, a framework is going to become indispensable. You're going to see a heavy increase in framework usage across the board.
• The server will become less and less involved with page rendering, and gradually become more of a services provider. The new stuff I build today only talks with the server across JSON-RPC, and in my personal opinion this develops faster and scales much better.

Which frameworks will "rule the roost" is hard to guess, but if I had to ...

• JavaFX can be safely ignored, it's going nowhere.
• Web apps are going to be predominantly javascript-based. For some types of apps flex and silverlight will be the chosen solution (e.g. video, games), but for most business apps the edge is just not there.
• GWT is going to become the go-to platform for enterprise java web app development.
• The number of projects that "roll their own" code is going to dwindle to near zero. It just isn't worth the effort anymore given the quality of the frameworks out there.

Within the javascript framework subset (discounting GWT), I'm assuming the following:

• JQuery is going to remain the primary solution, but it's going to be augmented with extensive component collections (JQuery UI), so as to match what the other guys can do.
• Dojo, YUI and Ext are going to fish in the same pond and will all do well marketshare-wise.
• Prototype is probably going to remain a major player for website augmentation, but is not going to be relevant for web apps.
• You're going to see a few frameworks grow in marketshare at the expense of all the rest. The market is going to standardize on a few major players. This will happen primarily through the job market, with employers preferring "standard" platforms over "whatever the new guy likes".

Slosh

Answer 2

HTML5 really rocks, and the newest implementations of Javascript have vastly increased performance, so I'd expect the RIA frameworks to have a hard time beating the new generation of HTML + JS + CSS apps. (While it will be a while before HTML5's widespread, some of its functionality's already available e.g. through plugins like Gears, and frameworks such as Dojo do a great job of transparently using the best implementation at hand -- plugin, native or w/ever -- for any given functionality).

I do expect Silverlight (and its Moonlight free software clone) to gain some developer mindshare -- mostly thanks to letting developers use their favorite languages for web apps. But I also predict this will remain a niche of sorts. (Just remember that "it's always difficult to make predictions, especially about the future";-).

Ripen

Answer 3

In the near future HTML+CSS+Javascript will rule with Flash, Silverlight et.al used for things that HTML just cant do (or if you are Microsoft for things where you want to annoy non-Windows users)

Sudor

Answer 4

This really depends on what you want your GUI to do. If it's going to be highly complex, with a lot of interactive elements, fancy transitions and rich media, then I think Flash will still be the way to go for the next couple of years (large install base, present in many corporate environments.) However, if you're primarily concerned with dynamic data and fewer 'eye-candy' animations, HTML/CSS/Javascript will definately get the job done, without reliance on third-party plugins. Another bonus for the HTML side of things is a larger pool of developers to pull from (in my personal experience, it's hard to find good Flash people. The good ones all already have jobs ;)).

As for Mastermind's concerns about the ability to index content, Adobe has been working with Google to make Flash indexable via a 'headless' player, see this post from the Google blog:

http://googleblog.blogspot.com/2008/06/google-learns-to-crawl-flash.html.

Eventually, HTML/CSS/Javascript will be capable of everything the Flash/Silverlight is. We've seen huge leaps in Javascript engine performance already from Chrome, Safari and Firefox. And HTML5 adds a lot of the rich content ability (ie. video) that Flash/Silverlight are used for. However, IE will always be an issue for rich HTML based content. And if you are targeting the enterprise, IE6 will still likely need to be supported for the next few years.

Unsex

Answer 5

I believe HTML/CSS/JavaScript will lead the dance for years to come. Even with their shortcomings (how many people are stuck with tables for design for not having a workable alternative) there is just nothing else to use if you care about things like SEO.

Flash/Silverlight work well for projects where there is another path for new people to discover these projects other than search results. For example, eBay and Amazon do not really need their content to be fully indexed on web. People already know them or come from affiliate links, so they could have probably rewritten their interfaces to become very rich.

If you care about SEO visibility, you should better spend your resources on optimizing this aspect. If not, just pick up whatever you like. You need however to be sure that your present and future users will be willing to install Flash or Silverlight plugins. For example, I have Flash plugin disabled in IE because it hangs the system and crashes IE on a daily basis. I suppose only YouTube was safe for me. As for Silverlight I haven't installed it yet and see no reason to do it.

Woodhouse

Answer 6

Some recent thoughts on Silverlight in particular, but also Javascript, Flash and JavaFX: Battle for the Rich Client.

Choe

Answer 7

If your target is a public website then the future will be HTML, although not necessarily "traditional" HTML. Frameworks such as JQuery, improved common features of browsers (such as cross domain messaging) and increasing average speed of PCs may lead to richer experiences for the user without resorting to things like Flash or Silverlight.

Latest advances standards such as HTML5 and CSS 3 sound promising but I can't see them having the impact they promise in the 3-5 year time scale. Not unless the industry finds a way to convince the majority to move away from IE.

Hence for the public current HTML/CSS/Javascript will continue to be the main stay of web development.

Intranet will be a whole different ball of wax, Silverlight and its community hasn't yet matured yet but I predict that the SL will become very popular in this arena.

Ammonium

Answer 8

Both technologies have advantages and disadvantages. Some things just can't be done, like nice audio/video streming in DHTML, but many things are not needed to be created in Flash technology. Google still dislikes Flash pages, as pages with no real content and availability problem.

I think nothing will change in big way. Maybe Silverlight will gain another 5% of market share, flash will lose 2%, but nothing really big.

Danita

Answer 9

I believe there will always be a place for web apps that are simple/fast to code, and don't need a fancy UI.

For apps that do need all the beels and whistles in the UI, Flash/Silverlight/RIA will continue to evolve to allow more and more rich graphics on the web. We also see a similar evolution on the windows side with WPF, etc.

In both cases, I believe the majority of apps will continue to be developed using whichever framework is the fastest to code in, as most apps are internal business LOB, and just need to be functional.

I would suggest you consider a hybrid approach--mostly use current technologies (HTML/CSS/Ajax/etc.), but have some sections of the site where the UI can most benefit from it really shine with Silverlight, or other RIA technologies. This way you get the best of both worlds, get the bling where you need it, but don't increase project risks.

Deason