The State of Web 2.0, The Future of Web Software

SphereOfInfluence writes "Despite some disdain for the term Web 2.0, the underlying ideas seem to be genuinely taking off from the seed of successful techniques of the first generation of the Web. Here's an in-depth review of the future of Web 2.0 and online software from Web 2.0 proponent, Dion Hinchcliffe. Like or hate the term, the actual ideas in Web 2.0 are turning out to not only usable but a growing cadre of companies are actively being successful with them. This includes the Ajax phenomenon being actively pursued by Microsoft and Google, widespread social software, and massive online communities like MySpace. These trends are all leading to predictions on the ultimate fallout of these changes, something increasingly called social computing. "

Just to clear things up...

[TechnoGuyRob]

This includes the Ajax phenomenon being actively purused by Microsoft and Google.

They mean pursued (I'm assuming), not perused.

This is a pretty long article, so I'll sum it up for you guys by taking the important passages:

Key Aspects of Web 2.0

- The Web and all its connected devices as one global platform of reusable services and data
- Data consumption and remixing from all sources, particularly user generated data
- Continuous and seamless update of software and data, often very rapidly
- Rich and interactive user interfaces
- Architecture of participation that encourages user contribution

...
In a way similar to how open source software (OSS) democratized and decentralized control of software creation, commoditizing it relentlessly along the way, Web 2.0 sites is doing same thing with the control structures of society and business. Web 2.0 represents the unyielding shift towards putting the power to publish, communicate, socialize, and engage, using an almost-dizzying array of methods, in online two-way discourse and interchange. The Web is the medium, but it's powered by people.
...
We are seeing surprisingly active interest in the conference circuit, with a large number of sessions about SOA, Ajax, and Web 2.0 in the enterprise in the next few months.

Also, this image [hinchcliffe.org] is a particularly interesting comparison of the growth of various Web 2.0 sites. The author finishes with some predictions:

Remaining predictions: 1-The hype is going to ramp down quite a bit this year. 2- People will focus much more on using the ideas and ignoring the Web 2.0 hypesters more often. And 3- A lot of folks will still hate the term Web 2.0.

Re:Why is it called web "2.0"

[MrRogers2]

Not really. Mostly it uses a javascript object to make calls to the server without refreshing the page. Imagine validating entered part numbers in a table on the exiting event of the field. You can them make the field backgrounds turn red on the invalid part numbers (without the full round trip for the whold document).

There's lots and lots of hype, but underneath there's some really powerful tools.

StumbleUpon

[gihan_ripper]

One of my favourite innovations in recent years has been StumbleUpon [stumbleupon.com]. It's a very simple idea — you install a StumbleUpon Firefox toolbar and click the "Thumbs Up" button when you come across sites you like, or the "Thumbs Down" button for sites you don't like. This way, StumbleUpon builds up a profile of the sorts of web surfer you are, and will then offer up a suggested website when you hit the "Stumble" button.

Using StumbleUpon, I've been presented with many really cool websites I woudn't have been able to find using Google, because I wouldn't have known to search for them. It seems my own interests are interactive flash websites, mathematics news, food, and philosophy. You mileage will vary, but will be catered for none the less.

Not anymore...

[Frosty Piss]

That's what really drives the web technology!

Old school thinking. That was only really true years ago when "legit" business was still new to the Internet. In my opinion, Porn really hasn't moved that much since the 2000 timeframe. Sure, there are better video codecs, but they are nolonger the product of porn production.

The irony of X

[amightywind]

It's rather ironic that we're trying to get browsers to do what other application platforms have been able to do since the late 1970s. I sometimes wonder if the web browser, like the gopher client before it, should be dropped for something, well, a little less kludgy and arcane.

It is also ironic that these days the distributed capability of X Windows (-display host:server:screen) is very portable, efficient, universal, and ignored for a less universal solution, HTTP.

Re:The irony of X

[jdeluise]

I'm sorry, but X is neither efficient for the client or the server. The premise behind X is that the application is running on the server (not the X server) and merely displaying on the workstation (the X server). Each instance of said application is going to consume massive resources (on the server..again not the X server), and is ABSOLUTELY NOT SCALABLE! Network-wise this is not ideal either as their is a tremendous amount of inefficient bi-directional communication just to click buttons and type in fields. This equals poor performance for the client (and what about printing, or other types of interactions that are now made far more difficult because the application is not actually running on your machine?) Have you ever tried to run X applications over a modem connection?? Well let me tell you the performance is miserable, even when you use X compression protocol modules.

I think the requirement of any scalable solution is for the application to in fact run on the client and merely communicate with the server. This cuts down on excess bandwidth usage, memory usage and CPU usage on the server while providing a much faster and better experience for the end-user. I'm not saying AJAX et. all is the solution but X certainly is not!

Re:HTTP deficiencies

[jdeluise]

I think you're trolling here, and your signature and the several troll or flamebait downmods in your past seems to support it. I wonder if you have thought through your original proposal or your response carefully? First off, spawning HTTP processes to handle a new connection is not nearly as costly as spawning a process to handle an X application. Why? Because generally speaking an HTTP connection will handle a very discrete and limited amount of data. Most of the time, the request will be processed quickly and will not use a significant amount of memory. However, spawning an X application is different because you will be spawning an instance of an entire application (the stack and the heap) for every user that connects. Lets say 1000 users wanted to simultaneously use OpenOffice for example. What kind of machine could handle this load? Remember, the client contributes very little processor time and memory to running these applications remotely.

The biggest problem with X is that the application is not running on the client machine, it is only being viewed there. Because of this, much of the functionality that would expect is not present. For example, suppose you were using an X-based mail application. What would happen if you wanted to print an email? Where would you be printing to? It would not be the X server (ie. your client for those who are not familiar with the way X works), it would in fact be the server you are connected to! Can you think of any non-hack that would allow you to print to your printer on your machine? Do you think any of these hacks would be secure enough to use in practice?

Now say you were using an X-based word processor. Let's say you want to save your document so that you can close the program. You click the save button. Where are you going to save the document? I'm sorry to tell you but it would be the server you are connected to. Sure, this could be coupled with some kind of remote storage facility so that you could get access to your documents, but it's very much still a hack, and not everyone is interested in saving their work to remote storage. The only way around these problems are hacks that are not only insecure but also far more costly in terms of bandwidth and loss of functionality for the client and processor time, memory, and bandwidth for the server.

Besides the above problems, you are still paying a lot more in terms of bandwidth costs for every single interaction with the GUI. As more people use said application, the slower it will become for everyone. What if someone discovers a bug in an X application that causes it to hang with 100% CPU usage. This would hose every other user on the server instantly. Just imagine the DOS attacks that would result from your proposal???

I don't know if you are a troll, an X fanboy, or you just plain didn't think it through, but I would suggest in the future that you argue on matters that you actually know something about and have thought through a little more. Thanks for your time.