Slashdot Mirror
Where is the Real Ajax/Flex Revolution Happening?
andzik writes "Even with all of the buzz around Rich Internet Applications these days, using toolsets like Ajax and Flex, most sites that utilize these technologies seem to be incremental improvements, not revolutionary interface changes. Where does the Slashdot community feel the best opportunities are to substantially create different/better user experiences using RIA tools? What will be the killer app? Are we just not seeing them because the best improvements are being made to web based applications and not in the public space?"
On a related note, Vertigo asks: "Not so long ago everybody believed that it was a good thing to have the freedom to modify your software to suit your needs or to mangle your data in any way. But now that users are flocking to non-modifiable, one-size-fits-all web 2.0 apps like Gmail or Flickr, are we moving away from our open source ideals? Those services do provide many important benefits, but in the process of their enthusiastic adoption did we not loose sight of the most important issues?"
One key question in this Ask is
Just my opinion, but I think the killer app may be out there already but in stealth form. It's mostly a question of discovery and trust, and I think both lurk right around the corner.
Just my anecdotal internet experience, but I'm migrating virtually all of my work into cyberspace and allowing internet services to manage my data and backup. I'm not completely there yet, but I've been a heavy gmail user for over a year now, and have almost forgotten how to use local pop clients (though I still do for peace of mind pop/download the e-mails for local storage -- I haven't gotten that far with my trust). And the sheer convenience of being able to "do e-mail" from any browser has been more beneficial than I'd predicted. I now have complete threads at my disposal whereas I used to find myself re-constructing threads dispersed across multiple machines (typically laptops "on the road").
Lately I've tried some of the on-line word processors and calendars, and yes even some of the spreadsheets (some of the on-line spreadsheets are very responsive and offer functionality 99% of excel users typically tap). They're not all there and ready for prime time yet, but they're getting close.
The word processors for my general use are already good enough that I'm willing to do my word processing on line and let "them" do the management. I wouldn't even consider (not that I did anyway -- I'm an OpenOffice user) any of the pricey Microsoft Word Processing/Spreadsheet options. Again, the side benefit, almost unexpected, is the universal access to my work with NO effort, just a reasonably current browser.
So, from my perspective, that's the "killer app"...: the security; the ease-of-use; the convenience; the cost; the true benefits reaped from a net where your data is created and managed in the internet "cloud" (sorry about all of the "quotes").
(As for the one-size-fits-all, I think the eventual internet app winners will be those who provide the functionality with the flexibility. And if you shop around you'll find these on-line versions seem to providing reasonable (maybe not complete, but reasonable) flexibility)
instead of complaining about the lack of killer apps, maybe you should be out building one.
Do you even lift?
These aren't the 'roids you're looking for.
Since when? Web applications are built on a foundation of HTML, which is simply a way of marking up information. How user-agents interpret that information is up to them. UserJS/Greasemonkey can modify web applications on the fly too. It's a bit of a stretch to claim that web applications aren't modifiable.
Bogtha Bogtha Bogtha
most sites that utilize these technologies seem to be incremental improvements, not revolutionary interface changes.
I like the idea of AJAX being used to enhance applications, not completely rebuild them.
If I wanted to do something like change the menus/site navigation I could already hose up the browser's controls with a flash based site.
If i want to do a quick validation in a form against a remote database, I'll use AJAX
If I want to add a quick way to change a record(ex. disable a user) in a table, I'll add a link that makes an AJAX call.
If I want a text box to do a spellcheck without posting a complete form, I'll use AJAX
"If they have both, tell them we use Linux. And if they have that, tell them the computers are down." -Dave Chapelle
There's only so much you can reasonably do with a web app before it becomes more feasible to make it a proper application. Wait for the day when we have a real programming language (Read: C/C++) as script, instead of Javascript, and then you'll see how powerful the internet really is.
I'll subscribe to Slashdot when I see a month without a dupe, a typo, or an article the "editors" didn't read.
BTW: Flex is a popular lexer based on Yacc and not some web2 buzzword.
What users are you talking about? Those who use OSS or your typical Internet user?
Bare in mind that the internet, aside from the technical sites, has become a huge business, ecommerce, entertainment, and anything else that non-IT people want to use it for. The latter folks have no idea what OSS is. They just want thier music, porn, buy books, etc... And they'll use whatever canned software that is offered - they don't want to mess with code.
I guess what I'm saying (and what others have said) is that the internet and computers are just a home appliance now. Anyting that makes computers more of an appliance will sell BIG!
Saturday is April 1. Slashdot will be shut down. Sorry for the inconvenience.
Those services do provide many important benefits, but in the process of their enthusiastic adoption did we not loose sight of the most important issues?
;)
The killer app is one that automatically fixes "lose" misspellings?
See my journal for slashdot ID's by year. Mine created in 2005. http://slashdot.org/journal/289875/slashdot-ids-by-year
No. "We" have the vision clearly in mind in all of the cases you mentioned. Gmail, for example, is available to you from almost any contemporary communication device, including your handheld, laptop, public terminal, cell phone, etc. The same functionality is also provided by nearly all other hosted email services, including those one might build for exclusive use. This is the "freedom" users value. Precisely which "ideals" and "important issues" do you suspect have been so "enthusiastically" abandoned?
Lurking at the bottom of the gravity well, getting old
I see some really nice web designs out there, but when it takes a minute (or more!) to load a page with a DSL line, then I get a little testy. And many times, I absolutely agree with you. I just want the information, the graphics/Flash/whatever do not add anything. And many times, it makes site navigation difficult because the page becomes so cluttered, it's hard to make out what you're looking for.
Saturday is April 1. Slashdot will be shut down. Sorry for the inconvenience.
Iv'e found some apps that allready kill their software counterparts. http://meebo.com/ is an IM client that I use at home and at the office because of its portability and ability to function over trillian or other IM clients. A killer app is something that should do it for YOU not the market as a whole. I swear by Meebo but other AJAX apps fall short. Microsofts "Live Office" is grand for the avarage user, but he tech community will find it lacking... it is all about useage for YOU.
Go ahhead and run lad... I'm a good shot
The latest project on my plate is a console app even. So refreshing to deal with an application that actually responds in a reasonable amount of time to button clicks. Console apps aside though...
.net or some other new fangled thing more times then I can remember but we always come to the same conclusions. It works and works well and has no limitations another technology could overcome any easier, switching to something else would not benefit us any beyond being able to brag that I'm coding in [insert fad here]. Downsides include spending ALOT of time re-engineering a pretty complex system and almost certainly introducing a bunch of bugs that would almost certainly cause the product to loose customers.
The main web app I am working on these days still happilly uses VB6 controls. We (the engineering crew) have discussed switching to
In short, we aren't AJAX revolutionaries because we would realize no benefit from it. Google used it in one app and that spurred a frenzy because it was cool and it was Google. But that was an implimentation of somthing special that couldn't really be done well any other way. To say we need to automatically apply that to all the old functions just because Google uses it (for something mostly unrelated to what we are doing) is dumb. When we have a problem that analysis shows AJAX is the best solution for, only then will we dive into it.
Do I think AJAX will go away? No! It certainly has it's uses. Do I think AJAX has been over-hyped and a bunch of managers will make their engineering teams use it for no good reason because of the hype? That is a resounding yes. Oh well, life goes on.
I've just discovered OpenLaszlo earlier this week. It's a (now) open sourced web RAD system. It compiles into Flash files so almost anyone can run the apps, and it feels a lot less hacky than Javascript ever did... blasted browser wars, "standards", and all. Pretty interesting technology -- especially if you can't wrap your mind around building an application in the Flash "everything is a movie" model. The IDE is an Eclipse plugin.
I think the original point to my post was that AJAX is nice but I don't think that the standards are there yet.
The SMF forums use AJAX for post previewing. I was a bit confused at first, but it was somewhat gratifying to see your post previewed without having to reload the whole page.
VBulletin also uses AJAX in its latest version, specially when you edit a typo in one of your posts.
But I think the real revolution will happen in intranet apps, where there are TONS and TONS and TONS of forms!
Brag up my own site. I'm only using a modest amount of ajaxy stuff, but I'm making heavy use of javascript/svg, for things that people wouldn't think possible without java.
It is sort of a picture trading site, I suppose you could say. Free users get credit for uploading pictures, and for using a webapp to enter metadata about those already uploaded. All pictures thusly cataloged are searchable (and no, it isn't just tags). Wish I could show some of the applets, but they're not really SFW.
AJAX is a transition point in the web's evolution beyond the browser. The real killer app is what happens when these applications' communication protocols standardise. It's not so long ago that when you wanted to run a blog, you either hand-coded HTML or had to have a server running slashcode. Now you can choose between dozens of free and non-free web apps or hosted services, and it doesn't really matter which you choose because your audience can aggregate from any of these via RSS, and view your content in whatever client they choose. I'm sure Blogger.com is still a viable business, but it's increasingly irrelevant.
Similarly, it's rapidly becoming possible to share calendaring information with others via CalDAV without caring which client and server options you and your collaborators prefer.
Whenever I do a Drupal site for an organisation I like to encourage them to set up an LDAP directory, so that they can use the same authoritative data source for authenticating to Drupal and other systems, internal address books (usable from a multitude of clients), and finely-grained control over sharing personnel data with affilliated organisations. The ability to do all this is very cool, and not at all dependant on my choice of OpenLDAP (which is, frankly, a bastard to get working), as the critical element is the LDAP open standards.
These are pretty simple examples, but I don't think it's too much to expect that open standards for interacting with applications like Flikr and Del.icio.us will emerge, along with increased choice over back-ends and interfaces and effective commoditisation of the services. Value moves up the chain, innovators move on to the next big thing, and it all starts over again.
At least that's how it should work.
Matthew.
You want to use C for web coding? The two just don't mix!
JavaScript is flexible, simple, and already has the features needed for web development (like DOM).
C is NOT the answer to everything.
We are a small web development team ( 4 people ) doing web app work to interface with our mountains of textual data stored away in our DBs ( oracle and mysql ). We used to use the traditional click/reload web app design but have recently made the switch to focusing EXCLUSIVELY on Ajaxy clients doing a lot more with a lot less development work. I wrote the framework that we use to generate Ajax logic on the fly with our scripted templating engine ( think RoR, only with a lot more power and a lot more flexible and maintainable ) and it has turned our standard application development time from about 2 weeks down to about 2 days. We have some pretty advanced stuff already.
I've gotten our applications to basically mimic a natively compiled application. Our default 'skin' is a Win2k look and feel, but I've already been tinkering with whipping up CSS designs to give us OS X looks and also some KDE looks. The templating is what gives us the most power. Automatic code generation is our best friend.
Anyone seriously working with Ajax/JS on the large scale really does need to use some kind of code generation mechanism for all of your logic definitions. The technology to power most of the cool functionality you'll need changes too rapidly to allow yourself to have a codebase that can't adapt to the changes. With 6 tools currently in production using our new framework ( that we released two weeks ago ), if we wanted to change our primary JS library ( prototype ) for whatever reason, I would change the JS logic in our 'logic generation' template and all of our tools would be up-to-date automatically.
It's really changed the way we do things.
-E
One of the things I see holding back all these web-based applications is printing...how do you format envelopes and labels, insert page breaks and format pages etc. when you are at the mercy of the printing capabilities of the user's browser? Online is great but sometimes you just want a hard copy.
Heh, you better not rely on javascript to validate your forms. What happens when the malicious user disables/modifies the javascript? You still need to send that data to a server-side process for validation.
Using asynchronous javascript to send the data to the server and get the response is a way of saving time by drawing less of the page. But you still need to server-validate.
you network believers are going to be fucked when one of the systems you rely on goes down. a little too failure-prone / can't-fix-it-myself for me. it's hard enough to get people to do backups now... what happens when you assume your stuff is safe because it's in some digital bank vault somewhere? what happens when you can't get in?
so to all these web-apps, i say thanks but no thanks. if you want to give me a stand-alone version i can run in my browser *locally*, that's great. if you expect me to trust you, then i'll join the rest of the people that keep functioning in disasters and say "no".
so here's the point: "universal access to my work with NO effort, just a reasonably current browser"
not trying to rain on anybody's parade here, just considering reality here.
Perhaps there needs to be a distinction between "we"s here. We, as in the open source believers/advocates/etc, may indeed be partially giving away our open source freedoms in exchange for convenience in using new web-based apps like these. We, as in the mainstream, are doing exactly what could have been predicted we'd do: going for convenience as we always have. That's never changed, and is why open source is still only used by a fraction of end users. Where open source users move towards web-based applications is at the fringe of open source where it meets the mainstream: Those who aren't super dedicated to open source as a concept, and only really used it because of the value proposition it sold them on initially.
If you look at things in other areas of IT, you'll find confirmation of this there too. For example, this is partly why the iPod is successful with less options than its competitors, why most people don't bother with replacing MSIE even after all the security nightmares over the years, and so on. In the case of the iPod, ease of use is more convenient than learning how to use extra features. In MSIE's case, it _appears_ to most people to be more convenient to do nothing and fret than to find an alternative.
Web-based applications succeed based on their ability to sell convenience over desktop equivalents. In cases like Gmail and Flickr, they've succeeded against their desktop counterparts. The sharing of photos is way more convenient with Flickr, and Gmail is not only easier to use (ease of use == convenience), but it's also online, which is more convenient in itself for many people.
putfwd.com - 1GB Free file storage with a twist
the killer AJAX app will be a XUL app, not a web page.
Why do people feel a need for technologies to have a killer app? I mean, yes, killer app = good, but if a technology doesn't have one, maybe it's just not destined too.
Okay, so I'm a little fed up about hearing about AJAX. It's not a cure for cancer people, it's a way of updating web pages without a page refresh! For the people that find it useful, great. I really like GMail, for example.
I see comments like "most sites that utilize these technologies seem to be incremental improvements, not revolutionary interface changes." and think - well, that's great. I don't want incredible shiny web applications that sing, dance, and have unspeakable numbers of points of failure, I want web sites that let me do what I'm there to do, as quickly as possible.
Let me give you an example; at work, expenses are submitted through an online Java applet. It's pretty, it's shiny, it auto-verifies your data on entry, provides useful tips... and you know what? On half the systems I have access to, it refuses to work. Sure, it's mostly an issue with having been poorly tested, but if they'd given me a web page with some nice standard form/input fields, I'd have been done in a fraction of the time...
Even if the web is a very democratic place only the biggest player get a top place. Just look how Yahoo declared itself as satisfied with second place or like Microsoft won't overtake Google anytime in searching. To overtake any leader of a certain kind of a web site you need huge amounts of resources, resources which only the leader can afford. So whenever there's this killer idea any big player simply takes it over and integrate it into their own side. Only when big players sleep like Microsoft did with IE, outsiders (Firefox) get a chance.
l ) needs a few millions to get started. But if I get this money it will completely change the desktop environment, finally fixing Ubuntu Bug #1.
On the other side it's quite easy to destroy someone's business if you use the right leverage and are prepared to drop enough money out of the window to push such a leverage (see OpenOffice). Of course that means you neither can make a business yourself except after you are the undisputed leader. Yet another sample is the Xbox360, even if the PS3 will be a success, Microsoft will simply launch another XboxHDTV and will crush the PS just a year or 2 later since they have these resources.
What amazes me is that nobody understands these simple mechanics. The web is the ultimate amplifying factor which soon will influence even normal business like shopping or banking etc.
Besides just look how Microsoft has an OpenSource lab to know any emerging thread from the OpenSource community many times even before the OpenSource community knows it itself. All the big players, who can afford it, have such labs and also the ordinary business has soon to use this technique.
To summarize, it's quite unlikely to see any killer idea emerging from small sources since these don't have the necessary sources. Another sample my own idea (http://wyoguide.sf.net/papers/Cross-platform.htm
O. Wyss
See http://wyoguide.sf.net/papers/Cross-platform.html
The killer app is functionality! The use of ActiveX and now java script are great for adding functionality but, with functionality comes major problems. When you make it so that an unknown web designer is able to run unknown code, unknown to you, on your machine you are just begging for a dick in the ass! That's what has happened with Internet Explorer and ActiveX. Developers started to exploit the functionality to their own benefit, screwing the users, and now you can barely use Internet Explorer at all. If you do you are guaranteed spyware, viruses, cross sight scripting, total pwnage!
Then everyone discovers java script. I don't know why it took so long to discover it but, they finally did. Using java script you can manipulate the client in almost any way. Given a little bit of time, the developers will figure out how to use this functionality for evil the same way that they did with ActiveX. Now many will argue that it can't happen because 'java script is more secure' or 'java script doesn't have as much power' or 'java script is sandboxed' but, it's only a matter of time before someone gets creative and ruins java script for everyone. Everyone will be forced to turn off java script support and Web 2.0 will be gone in a couple of clicks. It's already beginning. Note the recent discovery of a GMail java script exploit by a teenager. Sure it's minor. Sure it's already fixed. But, the exploits have begun.
Besides, when you get right down to it, the functionality that is being offered isn't all that wonderful. Type ahead autocompletion is nice and it's way cool because it is so new in web pages but, after a while the novelty wears off and it isn't all that special. But, who really wants character by character relay all over the web? Who wants web sites turned BACK into telnet applications? If you put lipstick on a pig, it's still just a pig! But, now the pig is giving you a disease!
You're right about forms.
A lot of things will change when anyone can start making and using business quality electronic forms the same way they use word processing and e-mail today. All you'll need for taxes is this year's ODF/XForm. 80% of "business app" software will go away. Standards committees like HL7 and X12 will become un-necessary.
We're getting closer. (see: freebxml.org, PDF, OpenOffice, XForms, Mozilla) It just isn't quite there yet. Who knows, some revolutionary forms client may use asynchronous Java script, but it doesn't really seem maintainable to me.
With implementable standards, the software should follow and small business will finally be able to replace a lot of their paper/pens and hand-crafted form applications.
While there is a cool factor to all this, its essentially a hackup of a document presentation interface. The code ends up an unmaintainable mishmash of 2 or more languages within a file. Sure, you get a charge out of making it all work, but it really is crappy software engineering.
For instances where you really can't run an executable on the client, and all you have is the browser, it's obviously the only solution. I think these situations are the minority, however, and users can be much better served by a client side app running ssl over the internet to a backend database. Response is snappier, browser compatibility is a non-issue, and it uses a lot fewer packets.
For clean printing, webapps can output PDF files.