Kevin Cheng  

Mo’ Money Mo’ Problems

November 7th, 2003 by Kevin Cheng :: see related comic

“Universal design is the design of products and environments to be usable by all people, to the greatest extent possible, without the need for adaptation or specialized design.”
- Ron Mace

Accessibility, or Universal Usability, seems like an obvious property which should be inherent for all designs. However, many systems and their designers seem to ignore accessibility as a requirement. Many web sites, for example, use small, fixed sized text which cannot be scaled to the user’s preferences thus hindering visually impaired viewers. Hence, many websites have been dedicated to not only providing guidelines in designing an accessible site (often dubbed universal design) but also in why it’s important to do so.

I myself, only recently discovered many of these guidelines and surprisingly, they’re not that hard to find and for the most basic guidelines, not that difficult to adhere to either. Before everyone jumps on the wagon and tells us that OK/Cancel isn’t universally usable either, I’ll just say it now. We know that, we’ll do our best to move towards that. Just because we’re not doing that doesn’t mean we can’t talk about what we should be doing, though! Before your next design, I encourage you to read some of the sites I mention or in the least, read my summary of them:Why Universal Usability
The INCLUDE for Telematics Project, a site dedicated to telematics for elderly and disabled, gives numerous considerations for why accessibility is important. Their reasons included marketing, ethical, legislative, cost and industry standards considerations and benefits. Telecommunication Industry Association (TIA) Access’s Resource Guide for Accessible Design for Consumer Electronics espouses similar points. However, they delve deeper by debunking several myths related to universal design.

Most unexpected were the two practicality issues one contends with when designing for additional audiences: cost and benefit. The general assumptions are that universal usability requires additional development time and the benefits gained are not worthwhile because the audience is inconsequential. TIA Access’s guidelines explain that the increased cost of development is marginal and will regain the cost spent either through the additional market gained or by redistributing the cost among a much larger customer base.

What Does Universal Usability Encompass
Not only is it necessary to define why universal usability is important, one must also define what universal usability encompasses. Most sites offer similar classifications of disabilities in this area: visual, hearing, cognitive and physical. Each of these classifications can be broken down into more detail (severity, color blindness vs. legal blindness, paralysis, etc.). Interestingly, each of the prominent sites also discussed accessibility in a less traditional context: outside of disabled and elderly. The tendency is often to associate accessibility with access for disability. Sites such as Trace Research describe how accessibility includes people who in everyday settings: strollers, rolling luggage, baby carriage, etc. They also include more limited control settings such as driving a car, where the driver must pay full attention to the task of driving with minimal distractions.

In contrast to their common definitions of accessibility, each set of guidelines seem to take a different approach to how they solve the problem. Some sites such as the World Wide Web Consortium (W3C) tackle the issue by specifically addressing accessibility from the Web. Their approach is quite thorough, however, and provides not only a set of guidelines but also criteria for success and varying levels to permit some flexibility for sites using their guidelines. They also provide examples of usage and reasoning for each of their guidelines (e.g., providing subtitles, audio equivalents, or typed scripts for animations or videos if they are time sensitive).

Similarly, the Royal National Institute for the Blind (RNIB) have a very specific audience and thus, discuss mediums including but not limited to the web but focus primarily on visual accessibility through audio alternatives. The Center for Universal Design at North Carolina State University (NCSU) takes a slightly different approach. Their guidelines and principles are wide reaching. They cover factors beyond computer systems, considering differences such as physical attributes (e.g., left and right handed-ness). Equally, the principles are very brief and do not provide as much guidance as the W3C or Trace Research. Nevertheless, an overview such as what NCSU provides is valuable for those looking for a place to learn the fundamentals.

Government Regulations
Aiding in the enforcement of universal design, the U.S. and U.K. government have taken steps to ensure accessibility for all federal agency material. It comes as no surprise that laws such as Section 508 have been instigated to ensure access. Previously, many manufacturers and their designers considered this law and those similar to it irrelevant for private systems. While technically true, the less obvious point discovered was the indirect responsibility to create accessible systems. Because most employers are required by law to provide tools that do not discriminate, only tools which provide such functionality would ever be purchased. As a result, systems designers should always consider their universal usability as a necessity for themselves as well.

The Corporate Bandwagon
Recognizing this need, corporations are also creating internal guidelines and checklists. IBM has a particularly extensive site on various forms of accessibility. Those related to the web generally refer to the W3C but additional guidelines are provided. Particularly noteworthy are the IBM accessibility checklists used during development to indicate whether a feature has been implemented, plans to be implemented or will not be implemented. This approach is useful and provides more structure than a loose set of guidelines, such as what Microsoft uses.

Rather than publicly provide their guidelines, Adobe elects to advertise their accessibility features in some of their key products. Adobe illustrates how accessibility is not detrimental to a product’s market and may increase its marketability.

IBM and other companies also offer software specifically designed to increase accessibility. IBM HomeReader is a web browser which reads the web pages to the user whilst JBliss offers software designed for visually impaired users. Their features include intelligent magnifiers which vary colour and contrast to suit the user.

Tools
In addition to tools to directly aid accessibility, several tools are available to help designers create universally usable sites. The ‘Bobby‘ evaluation site uses the W3C guidelines as a basis for evaluating sites submitted to their web site. A quick test of their site with OK/Cancel revealed many accessibility issues but more importantly, suggestions were made on how to alleviate the problems listed. Many of these solutions suggested were quite trivial to implement, furthering the point that the additional cost to develop an accessible system is marginal. A slightly more detailed evaluation tool called WebXACT also uses the W3C guidelines and provides free evaluation and suggestions for improvement.

Summary
Clearly, there are a great deal of guidelines and ideas on how systems, even just computer systems, can be made more accessible. The key is first for the designer to understand what accessibility is and then determine the needs of the user base. In general, all of these guidelines, however varied, have illustrated that creating an accessible tool is not difficult once the designer takes it into consideration. While the level of detail, categorization, and emphasis of each of these sites are different, their overall principles and solutions are consistent and ultimately, invaluable.

20 Responses to “Mo’ Money Mo’ Problems”
Bob Salmon wrote:

If you’re interested in this area, particularly if you live in the UK, have a look at http://www.abilitymagazine.org.uk/, which is the magazine of the British Computer Society’s Disability Specialist Group (quite a mouthful!) They do a dead tree version too.

Bob Salmon wrote:

On the usable HTML front (not that I’m pointing any fingers), but I notice that OK/Cancel uses tables for layout. I’m assuming this is an artifact of Movable Type - certainly it’s a very common technique. What do people think of the pure CSS approach of say http://www.csszengarden.com/?

Jon wrote:

As an educator, I face the same issue Bob has brought up, although for a slightly different reason; universal design aside, convergent design trends point to the integration with form and content, allowing my toaster to display some kind of web content. Ignoring the ridiculous cultural commentary of needing to get my email on my appliances, what do I teach my students with regard to HTML coding? Tables are easier, faster, and for all practical purposes, much simpler to learn; the visual-oriented find them to be very straight forward.

But, of course, they’re a big hack. Learning and properly applying xhtml techniques and allowing content and presentation to split is a big, big headache - especially for a novice.

Any thoughts on these questions:

A: is it ok to teach a philosophically wrong method that pragmatically works?
B: does anyone have experience teaching css for complicated layouts?

Joshua Kaufman wrote:

Jon, in response to your questions:

A: Yes, I think it is wrong. The same mentality is why a lot of software is poorly written. Yes, it works, but it’s very messy and is difficult to maintain. Using tables for layout is no different.

B: I don’t have much experience teaching CSS for complicated layouts. But it’s important to remember that a lot of complicated layouts exist because tables were used to do something they were never created to do. You can sometimes take a table-based layout and duplicate it with pure CSS, but it’s not always possible and is usually a bad idea to start with. CSS was created to style HTML, not to replace tables.

Meri wrote:

I think that the complexity of correct CSS and properly formed (X)HTML is overrated. I realise that those who began with the “table and font-tags” approach (as my more standards zealous friends are wont to his) might find it quite a leap. However, if you begin with this approach and are taught it well, then it becomes much much easier. After all, isn’t separating content & structure from specific presentation eminently logical?

Simon (http://simon.incutio.com/) has noted a number of times how easily his girlfriend learnt the standards-compliant approach from scratch — because she hadn’t already learnt all the hacks. Admittedly there is more initial outlay to learn the approach, but it is not significant if the right methods are used to educate … just as all you need to do to convince a computer science student of the importance of commenting is to make them write a program, leave it alone for a week or so and then go back to modify it.

Is the hack-and-exception based approach really more pragmatic, or are those who use it just more loathe to give up their hard-earned skills and watchouts?

I am not a standards junkie or zealot, so please don’t flame me for that. I just mean from an educational standpoint those learning for the first time can handle all sorts of levels of complexity we would not expect from those with even basic skills already. The equivalent is the ease with which very young children learn multiple languages … I have a number of friends who speak more than 5 (African and European) languages, simply because they learnt them as young kids because their extended families were multilingual.

Can you imagine the challenge of 4 extra languages for an adult? I think we’re really talking about learning curves … and that’s probably a whole other comic ;-)

Meri wrote:

Apologies, KC and Tom, for hijacking your comments section a bit here. We should probably continue the discussion elsewhere as it’s not particularly to do with this particular article … what’s the ETA of the forums or will the bandwidth not cope? ;-)

Joshua Kaufman wrote:

Meri, I think you’re correct to some degree. However, learning new and improved techniques is part of working in IT. Yes, it can be difficult. No, it’s not necessarily intuitive. But once you understand it, it makes a lot more sense than hundreds of tr and td elements. I originally learned web design using tables too, but then I saw the light and it was clear: valid CSS and HTML is the future the web.

Jon wrote:

Joshua,

I think you are absolutely correct with regard to valid CSS; however, your statement about “working in IT” is a bit shortsighted. Considering the trend towards ubiquitous display of information and the convergent design patterns between product and interface, Industrial Designers are now in an exceptional position to create hypertext-driven content.

These designers are not information technologists - they are humanists, first and foremost. And while they must clearly have the potential and capability to see their creations through to fruition, it’s fairly safe to say that learning the nonsensical nuances of xhtml is not a great use of their time. Designers are problem solvers, first and foremost, and as Mies said, “god is in the details”; the details are currently much too difficult to achieve using css.

Steve Donie wrote:

Getting back on topic…

I spent the first 6 years of my programming career in accessible software - it was a great thing. I spent 4 years writing screen readers for people with visual impairments, then 2 at Microsoft (http://microsoft.com/enable) working in the accessibility group. The hardest thing was convincing people that it was worthwhile doing the extra bit needed.

I was more focused on actual software, but the web and information access is another really important area that doesn’t get enough attention. I think most programmers are still in the “I had no idea that blind people used computers!” camp.

Anyways - thanks for the plug, althought the comic is a bit annoying - not exactly the kind of message that helps ’sell’ accessibility as a feature.

For the programmers out there - if you don’t see accessibility as a feature, think about your testers - if your program is accessible to people with disabilities, it is likely that testing tools can test it also…

Elly wrote:

Steve, while I agree that the cartoon is perhaps a little cynical about accessability I think it illustrates quite well the need for integration of accessability - rather that it just being an added extra.

If you’re designing a public building you have to make sure that (for example) someone in a wheelchair can get to all the areas any other user could get to. It’s not enough to say “well if they do come in they can ask nicely at the desk and the kind and friendly receptionist will guide them round 3 miles of corridor and take them upstairs in the service lift.”

So in the case of the cartoon - there shouldn’t be an extra option for the wheelchair-user/short or stupid - these people should be catered for within the design.

And when it comes to computer/web access it’s not enough to say “oh there’s a version of (whatever) for the blind/deaf/old/young/cynic” For a start most people would object to being labelled in such a way. Secondly, surely this makes everyone’s life more complicated - if you (as a designer) have to create half a dozen (and this number’s not going to shrink any time soon) different versions of your design that’s not making your life easier - and if the user at the other end has to go through some equivalent of the “if you are in a wheelchair press 1, if you are blind press 2, if….etc” they’re not going to be a happy camper either.

Meri wrote:

And equally, making assumptions about the various competencies or abilities of the user can be rather disastrous, as I think the comic is trying to illustrate.

Kevin Cheng wrote:

Bob, Jon and Joshua: MT doesn’t restrict us to using tables. I would like to use CSS where applicable (as Joshua mentioned, tables do have their place) but we wanted to get something out the door first and slowly massage the site to our image as it were. On our long list of todo’s is to convert the site’s implementation.

Jon: I think separation of style and content is the philosophy you should be teaching. I also agree with others that it’s not necessarily pragmatically more difficult. I think dealing with tables and the different browser interpretations of them is in fact more troublesome, especially for those that don’t need to learn the nuances as you pointed out.

Steve: Meri and Elly have both said it much better than I can. Many of our strips are in a similar vein - we look at an issue and how it can go terribly wrong if done incorrectly. A lot of people’s views of accessibility revolve around separating “those other users” from “normal” ones which I’m sure your experience has told you isn’t always the best idea.

We try to illustrate these ideas in a more light hearted way. I doubt (or rather, HOPE) that nobody looks at the comic and decides that is the correct way to implement accessibility.

Equally, we believe sensible ergonomics and reasonable interpretations of Don Norman’s ideas are the right approach.

Bob Salmon wrote:

As a pusher of small children in double pushchairs, I appreciate the fact that accessibility features (e.g. wheelchair ramps, kerb cuts on pavements) can have more general use in the wider population. I hate software that relies on the mouse, with no keyboard shortcuts - it’s quicker to use the keyboard, if you can bear the cognitive load of a less wizzy interface.

Jon wrote:

Kevin -

Not to be a pest (ok, to be a pest):

“I would like to use CSS where applicable … but we wanted to get something out the door first”

“I also agree with others that [css is] not necessarily pragmatically more difficult.”

I understand that MT comes with table-based templates, but if css is so simple, wouldn’t a conversion have been fairly straightforward?

I guess I’m playing the advocate of the devil here, because I *use* css quite frequently and enjoy the flexibility it gives me; however, writing valid xhtml is a fierce pain in the ass. And quite frankly, I don’t have the first idea how to TEACH valid code at all! Tables make sense to visual learners because illustrator & photoshop can simulate a grid via guides; css’s absolute/relative/fixed/static positioning is a nightmare for my students. They simply just don’t get it, and I can’t think of a good metaphor or systematic way to teach it.

Nope wrote:

There is no such thing as a “U.S. and U.K. government” unless you know something that the rest of us don’t. The UK doesn’t have any ‘federal agencies’ either.

Kevin Cheng wrote:

Jon:

My statements sound contradictory but I don’t believe they are. I have developed HTML since the Table tag was new in Mosaic and thus, have invested a lot of time into learning the nuances of tables. Conversely, developing using CSS and valid xhtml is something I’ve only recently started learning. I think a site like OK/C could probably be moved in about a full day of work. Unfortunately, that full day is usually committed to making a comic and writing an article already so when I say I hadn’t time to do it, I meant it on a fairly small scale.

I feel separation of content and layout is inherently logical. The implementation details may not be quite so. As I mentioned, I haven’t done a lot of non-table layouts so I can’t offer suggestions on how to teach it but I do believe it should be taught over table layouts.

Kevin Cheng wrote:

“Nope”:

The brevity of your message leaves me with some confusion as to exactly what point you are protesting but allow me to try to address it.

If you are pointing out semantics issues, you are correct. “US and UK government” would imply a government that oversees BOTH countries. I, of course, meant to say “US and UK governmentS”. Equally, I made a mistake in mentioning “federal agencies” when I was specifically referring to Section 508 in the United States and not the U.K.

In the U.K., the laws are not as explicit but the Royal National Institute for the Blind (RNIB) have gone so far as to interpret how the Disability Discrimination Act may apply to websites ( here ).

My original point was to state that one of the reasons companies may wish to become more aware of universal usability issues is to avoid liability issues.

I hope you at least found the remainder of the article helpful and did not discard the entire piece due to my misleading language. Thanks for pointing it out, I’ll correct it ASAP.

Tom Chi wrote:

To inject a little pop psychology in the mix, I’d like to suggest that CSS vs Table is also something like Type A vs Type B personalities (or Filers vs Pilers). All pilers feel this sense that if they were filers that everything would run so much better, but still they live with their piles. In fact they can live quite well.

Filers are great at telling the world how wonderful their organization systems are. Although I like AListApart, the way that they push standards is definitely pretentious. It reminds me of how one of my old piano teachers would tell me that only a particular technique was acceptible. When I got into the real world of playing gigs, it was more about doing things that worked, regardless of technique. Still the influence of filers puts a nagging voice in the head of most musicians: ‘oh, if only I had better technique’.

In reality, both influences are important. The “get it done” urge is as important as the “do it right” urge. There will always be tension as long as there a way to get it done without completely doing it right. As it stands with the web today, the “right” way is not even that “right” yet. Case in point with our font size, which is still tiny in Mozilla (KC glares… I was supposed to fix that). Well it turns out that the size is different in every browser implementation. In fact the only way I’ve found so far to make it look right in most major browsers and still have it be resizeable was with the ‘font size’ tag. I don’t know of a way to do it in pure CSS, without invoking javascript for browser detection. In this case the “right” way of doing things is painful and still wrong (we can’t assume everyone has javascript turned on). In fact it’s so bad that some of the major major sites solve it in the “bad” way. CNN.com doesn’t have resizable fonts. Salon.com uses the ‘font size’ tag! Wired just gives you that little font control thing, which I’m not fond of.

We certainly can’t expect people to move to CSS in droves until it is certain to give us a better result. Ultimately we need to please the users, not the standards people, so if that requires the ‘font size’ tag, then so be it.

Bleizig wrote:

Regarding your bandwith problems … do you know you can compress the html sent to the browser? I personnaly implemented it and the pages size was divided by 3.

There are multiple ways to do it, you can use the apache mod_gzip or you can implement it directly in php like here: http://leknor.com/code/php/view/class.gzip_encode.php.txt

If you happen to read french, we had a nice discussion about it not a while ago:
->http://cybercodeur.net/weblog/commentaires/detailsCarnet.php?idmessage=685

Kevin Cheng wrote:

Bleizig:

That’s a good idea. Although our files are generated in movable type, we could precompress our html in the templates. However, as thegoodtomchi pointed out, our major b/w comes from the comic, not the code (and in the code, it’s our articles that take the most space). We’ll definitely add it to our toolkit though.

Unfortunately, though I took French for years, it’s quite rusty now.


Leave a Reply


OK/Cancel is a comic strip collaboration co-written and co-illustrated by Kevin Cheng and Tom Chi. Our subject matter focuses on interfaces, good and bad and the people behind the industry of building interfaces - usability specialists, interaction designers, human-computer interaction (HCI) experts, industrial designers, etc. (Who Links Here) ?