Stop using Ajax!

In this controversial article, James "Brothercake" Edwards argues that we should stop using Ajax (unless absolutely necessary) until we have ironed out the issues surrounding its lack of compatibility with accessible technologies such as screenreaders, and suggests how you can stick to accessible web standards-based techniques, instead of resorting to Ajax.

( Read the article )

Let me just say - "I hear you, brother!"

This echoes what I've been thinking for quite a while now, although I wasn't thinking so much about accessibility issues, but more about how so many sites seem to over-complicate things to such an extent that reliability suffers. Not to mention how much time, money and effort must have been expended creating those sites.

Originally posted by Albert Einstein:

Make everything as simple as possible, but not simpler.

Google Maps need Ajax to work, and so I have no real criticism of them, and accept that the pure accessibility issues are (as far as I can think) unsolvable for now.

This is not true at all.
Can you remember how online maps looked 3-4 years ago? Look at current MSN Maps for example. There is no fancy "grab n' drag" interface -- only simple arrows on the sides of map. And before Google Maps there were only this kind of interface on most online maps. I expected that they just add additional interactivity to that approach, but they decided to destroy accessibility and build all from scratch. Some users just can't press some button for long enough time to scroll the map, but they can move cursor and click.
So, what can be done to make online maps more accessible?

• Re-add side arrows to interface, put on arrows simple HTML links, so even users w/o JavaScript can navigate a map
  
• Load static HTML page with map initially, but plug JavaScript with Ajax for users who will try to drag a map. If user clicks on side arrows, continue to reload whole page, if user drags a map, load map parts with Ajax.
  
• This is it! Simple!

BTW, great article! Thanks!

Post edited Friday, 25. April 2008, 20:00:57

First off, I couldn't disagree with this article more or this comment more.

To start, accessibility is not a concern for me, and it only ever will be if I'm designing towards an audience that consists of a disproportionate number of concerned parties.

Let's face it, when talking about accessibility, we're mostly talking about blind and hard of vision users. There are about 1.3 million legally blind individuals in the United States (http://www.afb.org/Section.asp?SectionID=15). Now, that's "legally blind", not flat-out blind. I have three people in my family who are "legally blind", and I am nearing that status myself. We're all non-severe, however. We can see with contacts, glasses, and are candidates for laser eye surgery.

Of those 1.3 million legally blind individuals, a smaller percentage of THAT NUMBER is classified, "severe". So, let's just say a safe estimate is there are 300,000 severely blind people in the United States.... approximately 1/3rd of the legally blind. Of those, a significant number are elderly, which, for now at least, takes them out of my core audience. So, I'd estimate about 200,000 blind internet users *in the US* (fair enough, number increases internationally, and the proportion changes, but let's focus on my main market of concern).

The July 2007 approximation of the US population is 301,139,947. So, 0.06% of the population is severely visually impaired. Even if you consider ALL LEGALLY BLIND PEOPLE, 0.4% of the population is visually impaired.

But because it's not particularly handi-capable (Dojo's getting there, though), you're trying to tell me that we should abstain from or use moderately one of the biggest boons to intuitive and cost effective interface designs since the web standards were created? Please....

After reading that article, I feel you are immature.

Originally posted by Andrew Gregory:

This echoes what I've been thinking for quite a while now, although I wasn't thinking so much about accessibility issues, but more about how so many sites seem to over-complicate things to such an extent that reliability suffers. Not to mention how much time, money and effort must have been expended creating those sites.

And I think this user highlights your true motives. Accessibility is your straw man. You do not like *bad AJAX design*. That's fine, and I feel you. There's plenty of it. However, to blame AJAX is like blaming lumber for shoddy construction, or bricks for weak foundations. It seems as though you've really got a grudge against bad developers and designers, not AJAX.

But all that's speculation.

Good AJAX is cost effective, time effective, and labor effective. Trust me, as someone whose been doing AJAX since 2005 who has also been web developing for 10 years, wise use of AJAX has done nothing but increase the productivity of myself and my user base.

In conclusion, accessibility is a statistical non-issue, is not cost effective, and is a bitch b/c of varied implementation of web standards. I'd rather just tell blind people that I'm sorry and ignore their problems than I would spend an extra week's development time to fix them (and yes that is a lenient estimate considering the variables involved in being accessible). It's the cost effective and the mature choice. Sucks, but it's true.

Oh, and by the way, this site's main page has 52 top priority web accessibility errors.
http://www.tawdis.net/taw3/cms/en

Rafajafar, sorry, but you just don't understand accessibility problem at all. You just want to close your eyes on the problem.
Remember, disabled users are not only blind users...

I'd rather just tell blind people that I'm sorry and ignore their problems than I would spend an extra week's development time to fix them

I'm pretty shure that soon you can't just say sorry for disabled users, just like stores that don't provide acceess inside via special ramp for wheelchair.

Post edited Friday, 25. April 2008, 20:10:35

I agree with you (the author), but don't you think aggressive AJAX usage by developers would, over time, help to establish better design patterns for the trilingual problem of web design, as well as address the various other problems with AJAX itself? So, to some extent, I would be concerned that your position is too conservative. Point in case, I had to get a forum account to post this message, and to check the availability of my handle, your site used a pop up window. AJAX would have been perfect there! Don't be afraid!

i registered with dev.opera just so i could comment on this article since it irritates me so much.

my overall impression of this article is that your views are in direct conflict with the progress of technology, and are therefore extremely short-sighted.

first, your argument to stop using AJAX until it is more mature will prevent it from becoming more mature. let's use a simple analogy: the internal combustion engine. the reason it is so mature and robust is that it has gone through many revisions, starting from an immature technology to a mature one. granted, accessibility issues are moot with this, but the point is valid: technologic progression occurs when there is demand for something better, and telling developers to stop using AJAX will destroy its demand, and therefore its progression.

i want to use another analogy: the printing press. let's say gutenburg decided not to print anything until his printing press supported braille. i don't have to spell this out for you, but that would have been absurd and while you're not suggesting something so extreme, i think the point is still valid.

if everyone takes your advice to stop using AJAX, you'll kill the technology completely. your arguments, if adopted by a great many, would greatly reduce the progression of web technology at great costs to the many with benefits for the very few.

applying your arguments to any mature technology would cause it never to have been developed in the first place.

you ask us to develop web apps for the entire world, but let me redefine progress for you:

"The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore, all progress depends on the unreasonable man."
George Bernard Shaw (1856 - 1950)

So how do you feel about you tube? It cuts off a whole group by being inaccessible Oops I forgot....nobody counts the deaf as real people. Nice post, I agree about Ajax. I accessibility is an important consideration.

@dredfive & @Rafajafar

You don't see the point that BrotherCake is making. I bet you didn't even read the article fully...

AJAX requires a javascript engine in the browser to work, if that browser doesn't have a javascript engine and the site relies on AJAX only for it to work then the user is screwed over. Accessiblity issue #1.
Plus many people are accessing the internet by their phones these days, not everyone has an iphone or a powerful blackberry that runs a browser that can handle AJAX(currently opera mini is server based, but doesn't really handle AJAX very well on java for mobile). Accessibility issue #2
Not everyone uses their mouse to browse the web, and at times people may not even have a mouse handy to browse the web, if the AJAX app only relies on the user needing a mouse to work, then that seriously cripples the user. Accessibility issue #3

i did read the article fully, but i doubt you read my comment fully.

let's face it, if you have a browser that doesn't support javaScript these days, you're not experiencing the web and no reasonable person would demand that all or even most web applications should work both with and without javascript.

mobile handsets are a problem for browsing the web, true. but it should in no way restrict the majority of web applications that are developed. first, the number of users on mobile devices is way too small, and their technology is too far behind. it's up to the handsets themselves to create a browser that can use the real web. see the iphone for an example.

your argument is that web design should accommodate users who don't have javascript and who may use handsets, and my point is that web design shouldn't be restricted by those two cases. i think the users should get a real browser whether that be on their handset or on their desktop. saying we should slow the development of the web for those two is just...whew... outrageous in my opinion.

let me just reiterate my last point. by having web developers cripple their applications in order to accommodate those who do not use common technology, we will give those users or handset manufacturers even less incentive to upgrade their browsers and we'll be caught in a brand new problem: we'll stifle innovation.

@Fatal
Oh, I know the accessibility "problem" well... and it pretty much boils down to hundredths of a percent of the population complaining. I understand it, I just do not care. It's not a "problem" for me.

I am not going to pander to the extreme minority... ever. It is not the same as not having handicap ramps. See, private businesses do not need these ramps. There's no law that says a business even needs handicap parking, at least in my state. That is a business's personal choice, and a minor choice to make, so most do.

To adjust your website to handle accessibility... it's no small choice. And it contributes to the development and testing phases significantly.

It is not financially responsible to engineer your entire system just to accommodate a few cripples.

To put it more clear, accessibility isn't a problem, being blind, deaf, or armless is. If you're one of those, you deal with it however you can or stop being blind. :-P Either way, no tears from me b/c you wanna use the visual medium too. I have no responsibility to those people, and to suggest that I do is an offense to me.

@shoust

Counter Issue #1:
If you're suggesting I should care whether every user can use my site, again, I do not. Embrace change or gtfo, I say. If I gave a damn about every single user, I would make all my sites Lynx compatible and fire the graphics designer. This is archaic, absolutist, and immature thinking... and I personally do not subscribe to it. If you do not want to run a browser with javascript, ok, but you deal with the consequences. You are such a minority, and really you're not the kind of minority I care about. Learn to deal or get off the web.

Do.Not.Care.

Counter Issue #2:
Build better phone browsers. Not my problem. Do not care.

I'm actually working for a Wireless Media company now. When I design for them, of course, it's gotta work in a reasonable wireless device. Unfortunately, there's so many of these devices, with so many different browsers and features and bugs... this problem is not standardized and therefore, I can't really deal with it in a comprehensive manner.

Again, it's about knowing your audience. This is still such a small audience at this point that I really cannot bring myself to care much. It is not cost effective to expect all web developers to give up good tools and techniques to satisfy some jerk who wants to read your site in an airport.

Do.Not.Care.

Counter Issue #3:
Seriously, mouse? Are you serious? ... no mouse, touchpad, or fingerballthingy... no touch screen or any means of moving a pointer?

That's not an accessibility issue, that's a matter of personal problems.

I use tabbing whenever possible to try and navigate my computer and the web, but let's face it, most of the time I just use my mouse. Even Stephen Hawking has a way of using a mouse, and that guy's immobile as hell. OK, so yeah, no mouse?

Do.Not.Care.

Originally posted by Rafajafar:

Either way, no tears from me b/c you wanna use the visual medium too

Wasn't that one of the points in the article? That the web wasn't supposed to be strictly a visual medium, but rather platform independent?

Originally posted by Rafajafar:

I am not going to pander to the extreme minority... ever. It is not the same as not having handicap ramps. See, private businesses do not need these ramps. There's no law that says a business even needs handicap parking, at least in my state. That is a business's personal choice, and a minor choice to make, so most do.

That's not the case in the UK. And sooner or later a country is going to pass laws ordering that websites either hosted or created in that country must be 'accessible'. It can't apply globally because of the nature of the web, but it's far from unforeseeable that a company that exists entirely in one country would be required by law to accomodate the disabled. It would not surprise me if mine or the US was the first to pass laws on this.
As for Javascript being disabled or unavailable, I see more and more comments here and in other places by FF users saying they use the NoScript extension due to the increasing number of JS security issues as well as increasing abuse of its capabilities.

Of course economics will sometimes dictate that small subsets of users will be locked out. And for a developer, doing a lot more work on something that might never be used is very frustrating. But economics and 'do not care' are not the only driving forces here.

Originally posted by Rafajafar:

Do.Not.Care

Yeah, I think that sums things up quite nicely.

Originally posted by Rafajafar:

this site's main page has 52 top priority web accessibility errors

No, it does not. That tool simply flags potential issues that need testing by a person. Accessibility testing cannot be automated. That tools flags every script element three times no matter what, and every image and style sheet at least once each, often more. Not to mention six other points that look like they'd be included for even a totally blank page. That's just the priority 1 messages!

It's really nice to know that someone claims to use "AJAX" well. My problem is with the teeming masses of developers who do not.

There is a huge and rapidly growing market of people (often in the third-world) using the web from very simple mobile phones that have slow data rates, slow CPUs, small memory, no touch screen, etc. That is what accessibility is all about, IMHO. Accessibility is far, far more than catering to disabled people.

What a conundrum. I'm afraid that I see both sides of the issue, other than the obviously inflammatory comments like:

Originally posted by Rafajafar:

It is not financially responsible to engineer your entire system just to accommodate a few cripples.

and

Originally posted by Rafajafar:

To put it more clear, accessibility isn't a problem, being blind, deaf, or armless is.

.

While it would end up being quite a bit more work, is it possible to use navigator.plugins to determine of the client has a screenreader installed and then serving the appropriate page?

This would ensure that Ajax technology is served to those who can use it, while a more accessible page is served to those who require accessibility. Sure, this would create a lot more work, but the company that appears to embrace its differently-abled customers has a HUGE PR advantage.

You're right, it's inflammatory.

I felt this article was quite inflammatory as well, as was the insinuation that I had not even read it.

Sorry.

Whilst rather long winded, I agree with this article. The title is rather blunt, but how can you argue with, "if you don't need to use AJAX, for the sake of accessibility, don't use it." There's nothing wrong with that. It seems like common sense to me. I'm surprised by the number of people who disagree and are even proud to passively discriminate.

Also accessibility is more than just about disabled people. Some AJAX sites I dislike for the very same reason that I dislike flash sites. They often kill the URL making it impossible to link to specific content. It's annoying when you have to include instructions with a URL.

Originally posted by drworm:

They often kill the URL making it impossible to link to specific content. It's annoying when you have to include instructions with a URL.

I have to agree with that observation.

But on the other hand, many well written AJAX sites have now special URI's to allow that direct linking to the contents in question, having the unchallengeable advantage of only rewriting the changed, new contents into the page, W/O refreshing everything, which is really important for me...

I believe, W/O AJAX this couldn't be done in such an elegant way, but I am no web designer...

My view is that we should continue pushing new technologies like Ajax despite the issues to the minority of users.

Economics will take care of it.

If this inconveniences 1% of users they will not be left in the dark. That's a big niche market, and somebody (hopefully me) will be hard at work responding to the demand this problem brings.

Originally posted by browland601:

Economics will take care of it.

...accessibility is not profitable, yet we (or those who write the laws, anyway) feel that those with disabilities have the right to concessions from the rest of societies.

(from comment on Reddit)

Read the article with great interest and have to agree with it.

For those that say "I don't care", I just hope you never experience discrimination of the sort. Having a disability myself, I know what it's like to have the door slammed in my face.

Originally posted by dredfive:

let's face it, if you have a browser that doesn't support javaScript these days, you're not experiencing the web and no reasonable person would demand that all or even most web applications should work both with and without javascript.

Yeah, but we shouldn't forget that more and more large organisation block javascript at firewall level. The user might have javascript enabled at the browser level, but getting out in the world through their corporate firewall, they never get to see the javascript.

Accessibility is NOT just about people with disabilities guys!

I am not surprised that the usual comments came up here, but am shocked by the 'do not care' attitude displayed in some of them.

Yes it's a small percentage, but it's a percentage that any one of us could find themselves in. Then see how you feel about being shut out?

Something that perhaps some of the people commenting here don't know, the author is not a grumpy old man having a go at these new fancy techniques that he doesn't like or doesn't get - James Edwards is an absolute sharp-as-anything world-class JavaScript genius and has been responsible for numerous techniques that he's shared with the community, either on blogs or in the books that he's written.

The fact that he's making these comments suggests to me that they are comments worth paying attention to - he is not doing this because he is *incapable* of producing an absolute diamond of an AJAX powered interface; he could do it standing on his head and still make you look foolish. He has, however, researched the issues that it throws up significantly more than 99.9% of the JavaScript web dev community and knows what he's talking about.

So let's not dismiss this article out of hand as 'the rantings of a crazy man', but rather the alarmed tone of a man who feels like he's the only one to have noticed this to be problem and who is, in all probability, completely fed up with trying the softly-softly approach and getting nowhere.

Originally posted by vavroom:

Yeah, but we shouldn't forget that more and more large organisation block javascript at firewall level. The user might have javascript enabled at the browser level, but getting out in the world through their corporate firewall, they never get to see the javascript.

This is *absolutely* correct. For a loooong time, every site that redeveloped with a JS library (pick one, any one) would immediately become useless at my place of work, because the content in the library would trigger off the potential malicious code alarms. So, what would happen is that *some* JavaScript would be blocked - a whole .js file - while others would be let through.

In this scenario, placing alternative content in a <noscript></noscript> would not work, because the browser would not act on that - the browser is JS-capable, it's just not getting the full picture.

This is a rel problem, and if for no other reason you should

1) Make the site work without JS
2) Add the bells and whistles on top, unobtrusively, perhaps with a library, perhaps hand-coded.

If the critical JavaScript is blocked, yo still get something.

It's a mindset that more people really need to get. As the quoted commenter said, it's not just about disabled people. In the building I work, there are 3000 *non-disabled* people who could not use many of the web apps that use certain JS libraries until just a few months ago when the firewall software was changed.

This just popped up in one of my newsfeeds (Dean Edwards' bookmarks?)

http://domscripting.com/blog/display/41

Somewhat topical...

Ok, I could have ranted here, but instead I blogged my response (also to the post above):

http://www.wait-till-i.com/2008/04/29/oh-look-using-ajax-in-a-stupid-way-is-not-a-good-idea/

In essence: James is right, but how is this a daring controversial article? The big problem about Ajax and accessibility is not the technology, it is the environment we apply it in. There is not much support for clever, accessible and usable development as we are asked to rush things. And we are asked to rush things because we've been preaching to the wrong people about best practices for too long assuming that "we'll never reach" beyond our own community.

@Chris:

I'd say the right people to access are the Paciellogroup people, especially Steve Faulkner.

Marco Zehe from Mozilla blogged a response to this article:
http://www.marcozehe.de/2008/04/29/are-ajax-and-accessibility-mutually-exclusive/

Marco talks about promoting WAI-ARIA. BTW, I'd like to thank Opera for being one of the driving forces behind that, so we can work toward a solution.

Good points in this thread, I also like the response article. I'd like to point out that accessability software also has a responsibility to keep up with advances in technology.

Ajax is immature and far from desktop application experience. However, while desktop applications are far better than Ajax apps, desktop applications can satisfy all accessibility requirements. So, the problem is with tools, browsers etc. I would stop asking people to develop accessible sites, rather we need to adopt existing tools to be compatible with current inaccessible Internet.

Meanwhile, from the developer point of view it is not hard to make applications compatible with pure HTML browsers if you are doing so called served-based AJAX assuming your have enough money to pay for extra servers for your application. Otherwise, AJAX can reduce server load and make servers cheaper.

I can't really see the web without ajax, comet or any real time application. I think the aricule should be about how to create readers more smarter in order to be avail to read any type of data added into a HTTPRequest.

The time and the industry is demanding high performers in order to make your site a successfully one.

And by the other hand we can always try to get around that issue... Like using a DOMScript in order to display ur Ajax or JS effects,.. Or add you rich text as a hidded one and just display it in any case that the js is off...

We can't tell to a developer to just code in regular HTML and CSS.

{ Hope you can get my point...}

Great article and illuminating comments. I look forward to seeing more accessible sites - including those with AJAX - in future.

Just commenting to say that I quite like the Flickr Edit Photo page drafted by Brothercake and linked in this article.

As a Flickr user who appreciates the design, feel and responsiveness of the AJAX on the site, I have sometimes wondered whether there's a way to make it more accessible without losing usability. Putting all edit functions on a single page sounds good to me. What I'd lose in waiting for the edit page to load, I'd gain in being able to make multiple changes in one spot. (Some things can be changed in-place on Flickr, but some take you to a separate page that only lets you change a couple of things.)

A good example of stupid way to use Ajax: invoicemachine.com
I can't even get final PDF from this "web 2.0 application".

Just for contrast here is a better service for creating invoices online that doesn't use Ajax and uses JavaScript in appropriate unobtrusive way: invoiceplace.com

Post edited Wednesday, 14. May 2008, 18:20:14

Originally posted by uklloydi:

This is a rel problem, and if for no other reason you should

1) Make the site work without JS
2) Add the bells and whistles on top, unobtrusively, perhaps with a library, perhaps hand-coded.

Exactly. _This_ is what is called "progressive enhancement" (the article implies that simply using AJAX is prog. enhancement). You can have all the bells and whistles of Javascript, as long as you don't depend on it.

Part of the registration page for this site is a perfect example of what not to do. The link to check name availability is a javascript:function() url. If I have JS turned off, the link would do nothing. In the progressively enhanced version of the page, this link would be your regular old http: link, but there would also be an onclick event handler that fires, does all the fun Javascript stuff and then returns false (to prevent the browser from following the link). Also, the onclick itself should be added via JS (instead of hardcoded into the HTML) so that (in the case of a firewall blocking JS), a browser with JS enabled wouldn't be looking for a non-existent function. Yes, debugging this would be a pain, but that only affects the developer, not the user.

I don't know how many times I've gone to a site and middle clicked a bunch of links (open in tab in background), only to find 10 tabs of javascript: links.