Putting the User at the Heart of the W3C Process

Back to Accessibility SIG Meeting 24th July 2007: Putting the User at the Heart of the W3C Process Notes.

This page contains the transcript and MP3 files of the presentation Jonathan Chetwynd, Peepo.com, gave at the 16th Accessibility SIG Meeting on 24th July 2007 at the University of Wolverhampton. Due to file size constraints, the presentation has been split into two parts.

Back to Top of Page

Part 1: Transcript
[Jonathan] Then I started working with people with learning disabilities and the first thing I created was a Java tool for people to author, to get them creating in any way, some sort of communication - but probably a written, rather than temporal... something a bit more permanent than just communicating for the moment - so something that's going to hang around for a little longer rather than just in the present.

So by 1998, the web was coming along and I was getting involved in the web and considering about how people with learning disabilities might access the web and I got rather drawn away from authoring and towards how people with learning disabilities might find things – so searching, basically, looking for things - and that led to Peepo.

I started Peepo in 1997 or something, but the Peepo project (as such) was launched in 2001/2 and lasted two years. It had very substantial funding. It was with Lambeth College and Wandsworth Borough Council and we had about £150,000 cash, which was supposedly match funded... so a £300,000 budget in quote marks. Most of the money went on assessment – looking at the issues with learning disabilities (we assessed about 200, over 200 students), the issues they might have, how to motivate them, why they might want to use the internet and creating resources that would be of interest to them.

That ended in about 2003/4 and I've returned to looking at authoring, as well as the accessibility. I'm really concerned that... I think the whole web is being derailed because the ownership of the means of production is being retained by charlies such as these over on my right [points to Pebble Learning and laughs], but... and the major corporates as well, not just companies. But what I'm concerned about is individuals, people like you or me, because I haven't been working with people with learning disabilities for about three years... how we go about creating stuff.

Even though the web's been around for ten years or more, it's remarkably difficult to create web resources... you know, there isn't a good HTML publishing tool. You can write stuff in HTML - you know, Word or whatever and the code's rubbish. But it's not that easy to publish it. But how do you get it up on the web for the average person, let alone somebody with learning difficulties? Scalable Vector Graphics, which is my preferred tool of choice... there just aren't any tools that are... unless you’re into graphic design and fairly obsessive, you've got a problem similar to Illustrator. How many people have ever used Illustrator here? Four people have used Illustrator. It's just not my type of tool.

Has anyone used iSketch? No? So we'll just have a quick look. This is a game similar to Pictionary. [Points to screen] Somebody's drawing something they know... they've been given a word and they get points for choosing what the word is. [Sound plays] When you hear that sound you know that somebody's guessed that word. One of the most exciting things we have... we went into a room on our own with people with learning disabilities... was... I would whisper to (we had 10 students all at computers, or 8 or whatever)... and I'd whisper a word to one student and they'd draw something like a flower or a house (something fairly iconic) and other students would all be looking at their computers and they'd just call out what they thought was being drawn. And this was an extremely motivating experience for the student, because the person who was drawing knew whether other people were understanding their drawing or not on a very instant basis.

What I want to show you is the interface and [points to screen] we have to go out of this to see the interface. That's a game, so you’ve got some idea of how quickly... and you do see some wonderful drawings. So we're going to go into "Animals Easy Spanish". Nobody's there. That's the drawing interface. We've got sort of four buttons there. There's nothing exceptional about it. It's just it's a relatively simple interface that people can use. And what my concern is that, whereas there was something called MS Paint and there probably are some Linux tools... Tux Paint. Tux Paint's just like this – a very simple interface for young children, but it makes sounds, so like that game, you can learn to associate using a tool with a sound. However, it's got a very cludgy (if you look at the graphics and stuff)... it doesn't have the wonderful smoothness that we've all learned to expect – eye candy – in our tools, when we're using them as people. Everything looks clean and functional.

When we were at Euston, the coffee pots are designed to look like dogs so you get three coffee pots sliding together and it looks like three dogs are meeting to sniff each other. And that's our expectation. But the tools for the ordinary user don't exist. They're not being refined, because of the... what I call... it's sort of the cult of the guild. The producers want to tie everyone in to their designs. So Illustrator gets more and more functionality, but people who want simple tools to use don't get them, including people with learning disabilities. What I'm coming towards is the pencil and paper. If the pencil and paper had come along... had come along as a computer cludge, it would never had got off the ground. It just wouldn't have been accepted by the guilds. It might have gone by the guild route, but it would have been incredibly slow. And what I'm nervous about and concerned about is: where are the very simple and easy to use tools?

Now one of them is twitter. Has anyone... who here has used twitter? Only one person's used twitter.

[Repeats after delegate] Jaiku.com.

[Delegate] You can join your twitter feed and your blog feed.

[Jonathan] So you're not allowed more than 140 characters and it just produces a feed, which you supposedly tell your friends about. It's literally that simple. There's nothing to be done. The problem you've got with it: is how do you find things? If you do a blog, it's quite well set up, so if you want to look at the history, you can go to a particular date or you've got a topic – some of them let you search on a topic – you can do a search for a word. This has got none of that. But it is a very simple... anyone who's got reading and writing skills can use it, but we haven't got a symbol-based version. People like illustrations (I like illustrations with my text), lots of illustrations, photographs, pictures... not for you. You know, this is not going to work for you. And it's pretty limited. You can configure it. Have you configured yours at all? Did you change the background?

[Delegate] Yes.

[Jonathan] Can we look at it?

[Delegate] Twitter?

[Jonathan] Yes.

[Delegate] In Jaiku - go to Jaiku.com. If you look at all those, those are Jaiku. They've got twitters in as well.

[Jonathan] If your friends or acquaintances... OK, so these are things that are being added automatically... any messages from these people on the right are being bunged in there.

[Delegate 2] I wanted to use twitter with our... a group of learning disabled adults, I work with in Lewes, and I just wondered... I think twitter to me had such a simple layout and it was such an obvious one. I just wondered if you...

[Delegate] The whole thing I like about that is it brings in lots of things and you can add comments. With twitter, you don't link... you can comment on another twitter, but you can't link it, with these it does. If it's yellow... I've commented on some of the other people's, but again quite short.

[Jonathan] Can you change the contrast? I mean, it's under cascading style sheets... are there? So you can probably change... exaggerate...

[Delegate] I think so. With twitter, you can make the text really big. Yes.

[Jonathan] Has anyone come across Amaya? No. So Amaya's a browser that W3C... (I'm not particularly indebted to W3C), but Amaya's their browser that they've developed. All the W3C technologies, they try to keep them... and this is Scalable Vector Graphics, which is my particular little... it's similar to Flash, but completely different. It's much more open and supposedly accessible. I hope this morning, you gathered that perhaps Flash isn't accessible as it might be. Is that unfair, I wasn't here?

[Delegate] Deeply unfair!

[Delegate 2] Deeply unfair! Terribly unfair! [Laughter]

[Jonathan] There's some question as to whether Flash is accessible, if you're interested in the web or how accessible it is and whether Adobe's heart is in the right place. Now, one of the concerns I've got is Scalable Vector Graphics... the whole W3C process. I don't know if you’ve gathered this, but this talk is about W3C and technologies, but also the process, and because the process is run by large corporates rather than small businesses (who'd of course do the right thing!)... Because it's run by large corporates, they’ve got their own goals and Adobe was running the SVG Group... it would be an exaggeration, but they're a major player. They've got a huge amount of money from Illustrator and PhotoShop and they sponsor like IBM, W3C. And so, it's not that what they say goes, except they can afford to have people that go to the face-to-face meetings and represent them. And it's the quorum that supposedly reaches the unanimous decision, but if you're not there (and not many individuals have the time to represent the ordinary bloke on the street at W3C), then your view is not going to be represented. And this is what I was going on about the ownership of the means of production... the separation of power... Who makes these specifications? Because if they're just put up there like HTML, are they designed for people or for corporates? And the fact is, you can't get around it... if the corporates are paying for it, then they're designed for corporates.

Getting back to Scalable Vector Graphics, which are supposed to be accessible... [Points to screen] This is a little bit of Scalable Vector Graphics with some animation there, and I can edit it, copy it... It seems to me the thing that one might obviously want to do is... I might want to copy and paste the whole thing. I can go around and select the individual bit (rather than the group of them) and paste them into a new document. This is the only browser that allows you to do that. So you link into Amaya, if you wanted to do that. The standard doesn't say anything about copying and pasting graphics. It's a complicated issue. Because with text... because we write...some people write linearly in a line, it's likely you want to drag over something, highlight it and select it. And it's a process that's relatively simple. With graphics, it's not quite so clear. Because the graphics might be on top of each other and what do you want to select? But there should be a methodology. You still should want a means to do it. However complicated or simple you can make it. And it should be in the specification but it isn't. Obviously people that use symbols to communicate, which is most of us, would like to be able to copy and paste symbols. It doesn't seem like an unlikely thing that you might want to do.

Now the majority of you know about e-learning, e-resources and you evidently know a lot more about it than I do. I would hope that most of that has got some RDF in there or something similar... some meta-tagging of some sort is involved in the e-learning resources, otherwise there would be serious problems.

Has anyone been to BigLou? What are BigLou's products? We haven't got time to go and have a look at BigLou. He's got a wonderful story in there. He's a very, very successful top-flight... He initially worked out something about captchas – it's a wonderful story, well-worth listening to and I strongly recommend all of you to listen to that. The captcha – is when you copy something into a box - and they couldn't understand why these captchas were being broken so easily. And it turns out they were showing naked ladies on these porn channels and if you wanted to see some more, you had to say what was written there straightaway. So they built up a huge library of captchas very, very quickly, because these people were desperate to get on with their business. And they so type in as quickly as they can to get on to the next picture. He tells the story much better than me! But then this got his mind thinking along these lines. This was the turning point that got him into these products. And if you have a look at his games, they're very, very clever. The ESP game... you get two people anywhere in the world and you show them a photograph and they have to say what that photograph is, and then you can choose words, but you're not allowed to use them. So you're not allowed to use that word so you have to find other words to describe it. It's amazing what people will agree... you get points... you both agree on the same word, you get points. And the point is that is meta-tagging. Instead of the boring tedious way of meta-tagging, which probably all of us have done in our time, this is an enjoyable way and people with learning disabilities would be extremely good at it... other than the text, you'd have to show them images. You'd have to invent something similar using images... assume that all words, sounds... probably images that they click on to say "this is a flower". And there are more developed ideas than that, but I don't have time...

Are you all using Cascading Style Sheets? Everyone uses that and you're all familiar with it? What we're going to have a quick look at is where we ended up with the Peepo project... it's actually slightly after the Peepo project. It could be that people have got bad eyesight and they need it to be this big, but it's the amount of information... We all get flooded with information, but do we necessarily want that information, all of it... and you won't necessarily want all of the information at the same time. Not everyone wants as much information as other people but the users... getting back to the theme of the specification and W3C... the user style sheets...has everyone heard of user style sheets? Who knows anybody that when they're using their computer they use a user style sheet? I don't mean people doing it for research or in particular circumstances. Does anyone know anybody that uses a user style sheet on a daily basis?

Back to Top of Page

Part 2: Transcript
[Jonathan] For the background?

[Delegate] Although you can have colour changes, text changes, type of fonts that they want...

[Jonathan] Type of font.

[Delegate] You can have a bar that it will do it automatically. They actually provide it. There are many bars now, like the assessment bar in Mozillla, I mean IE, that you can attach to the browser, and Opera... which makes it much easier.

[Jonathan] The project had a huge back end, but not having the people to maintain it, which I hadn't during the project, I couldn't maintain it myself and also it needed continuous feedback. There's no point in me maintaining it in limbo – it's not a sensible approach. I'm going to switch browsers and hope that we can... Safari. [Points to screen] This is an example of the BBC News page, which I’ve nagged the BBC about it. They’ve got something like 250 links on some pages and when I had a good relationship going with them for a short time, we got it knocked down. And it has happened. All this is is a style sheet, but unfortunately there's problems with the specification. This white space... they're not going to discuss with me, but it's pretty technical, the two people who invented CSS. It's been agreed... it was designed in such a way that the BBC can't actually do anything about this particular issue. They could do it in a different way – possibly. This has just removed everything so unless there's a picture associated with the text, we're not interested, it's not shown. It's obviously much nicer if it rearranged the information so that it was all on a small page just with pictures and text. That's just an example. The BBC were looking for how they might make their site more accessible and this was something that I created for them to show the way that things could be changed with CSS.

We're now going to look at the latest Peepo website. [Points to screen] This is only CSS here. It's only CSS and Scalable Vector Graphics. There is some RDF in there. Nobody's been able to show me a way that you can use RDF with client-side scripting or CSS to actually use the RDF, because I'd dearly like to be able to because years ago I wrote something... a graphical user interface schema so that you could choose what was going to be the tool bar, what was going to be the content, and label it as such. If you didn't want the toolbar, which would be the menu bar in common speak, you could get rid of the menu bar. A lot of things have far too many menu bars for most people's use. They're offering you access to all sorts of stuff you're just not generally interested in. Or the number of links... you can link... you only want ten content links per page. It just shows you, just like I showed you how you could do it with just one image, it could just show you ten at a time. You could click somewhere or scan down the page... a different way of displaying the page.

This is only CSS. There's no sound... as far as I know it's the only page that exists more or less that plays sound and it's not playing sound on JavaScript, which would be the obvious way to do it. It's playing it in a very controlled fashion.

[Points to screen] You notice the language is wrong. This is something to do with the content negotiation, and I'm not an expert on this - something to do with being here specifically. If I take this home, it speaks English, and it's something to do with they're adding headers as they send us the requests out, and I'm getting a little confused about it. There's actually about six languages in here with the sound and it'll do it. Unfortunately, it's doing it in Italian and English. I think it wouldn't be wise for me to say it's down to the specification again, but this content negotiation is slightly more complicated than it could be in a simple sense. But you can do it with Cascading Style Sheets. The unfortunate side of it is, and this is down to the specification in my view, you can't turn it off. If you've got somebody who's got a screen reader or some software like that – unfortunately, none of them work with Scalable Vector Graphics yet, as far as I know... there was a demo about five years ago that was produced... you can't turn it off. So they're going to have their screen reader reading music or whatever it is and they're going to get the sound coming over as well. The argument, I'm getting at the moment from them is that's down to the plug-in. But as you can see there's no plug-in displayed, which is very... it's the way they way they set it up like that. And so in my view, it's not just that you leave it to the plug-in. It's got to be it at every level that you have the controls. And that's not been arranged. This one also has keyboard navigation. Am I boasting? Yes?

[Voice from Peepo.com] Game. Music. Radio.

[Jonathan] You can see it's not working quite as we would like but, the keyboard accessibility is not part of the specification. Scalable Vector Graphics has no keyboard specification in it. The one that works – if you can call that working – SVG 1.1, which is the one that the current developers, Safari, Opera, Mozilla work to - and Mozilla for one are definitely not going to work to 1.2. It's not that they're never ever ever, like let's say in this next decade... but it's not part of their plan. And SVG 1.1. has got nothing about keyboard accessibility. It took me something over three years to find out what the problem was. Because initially, they thought it was something to do with the DOM because I was talking about keyboard, that meant it was to do with the content, and it's the Document Object Model. And actually, I was talking about the user interface - keyboard navigation, I have to be very careful now. Tabbing, as well. Now, we get tabbed windows, which we didn't two years ago. I managed to go behind the back of the specification and speak to the Opera, Mozilla, and web developers. And Opera and Mozilla, I believe, have already instituted keyboard navigation. They saw it as a natural thing that they ought to be offering even though it's not part of the specification. Generally, they don't blank specifications for very good reason – they've all agreed that things are going to work that way, that's why they called it a specification. But in this instance, they went behind W3C and they've instituted keyboard navigation.

Things that haven't been so successful is – for instance, Google have asked them why they don't provide a Flesch-Kincaid... It seems completely obvious to me, that if you're talking about people with reading disabilities, which most of us are, we'd want resources to be sorted by their reading difficulty. Something that's completely unreadable I'd be very grateful not to be pointed to, and there's plenty of that. For people with learning disabilities, Flesch-Kincaid isn't perfect, but there are plenty of others that are even more sophisticated. I couldn't trip them off the tip of my tongue, but there are ones that work on more clever... Does anyone know the name of any? There's some commercial companies that produce... Does everyone know what Flesch-Kincaid is? It's a scoring of readability. Basically it works on word length, the number of long words there in a sentence, how many sentences there are. That's the crux of it, I mean there's more to it than that. But the more sophisticated ones work on how frequently the word is used in the general English language, because some long words, such as the names of footballers, might be long and so like the Sun and the Mirror (and those type of newspapers), must pay quite a lot of money to have their editing checked by some machine that knows what their vocabulary is. It's not just word length because some long words are allowable because they're in common parlance or whatever. But Google... T V Ramen... I've spoken to the people there, or e-mailed them and been in negotiation and talked about it. But I don't know what the issue is, it's not a complicated issue. You can see it on the web, it's just a simple formula.

On a similar level, Ed Parsons was stolen from the OS and moved to Google – I'm not putting the oar in for Google – but just this particular issue. And I was asking how come there's only one icon? When you go to a Google map, there's just one icon, which is like a drawing pin. And so the equivalent in English, in printed map form, is that if every town said "town" and if you wanted to know what it was called, you got to look in the gazeteer in the back and find the reference for where it is and it will tell you "oh yeah, this is London". I was saying to them, couldn't they provide a few more icons? And this was his response: "We’re not talking about the printed map paradigm anymore, where you’re restricted to a massive abstraction of reality to allow communication using a single static symbol". And I said to him, "well, that’s what words are". They're massive abstractions of reality. He took the point to give him his due. He thinks there might be a use for having more symbols. Watch this space to see whether it happens.

Has anyone seen Dasher? About the time that Dasher came out from Cambridge... Basically it's letters that start on the right hand side of the screen, move towards the left, get bigger as you sort of start shooting them and you eventually start spelling a word and you can write with it without having a keyboard. It's a method of writing using a mouse, only a mouse. Do you know when it came out? About ten years ago. So about the time it came out, I'd been working on something similar, believe or not, with GIFs. It's not exactly the same thing, but if you imagine a similar sort of thing with GIFs... We're just going to have a quick look at that. [Points to screen] For my daughter, this is linked up to Flickr pets and it's not flashing through hundreds of them, but it gets a new one every day. Sometimes you get... I haven't seen any naked women, but you get one in a thousand... maybe one in a hundred is something that is a little bit odd. There was a whole load of them where... do you know what PET plastics are?

[Points to screen] This is the BBC Weather Feed, you click on it. So that's updated. That's what they're saying the weather is... it's actually for London. I didn't put it for here, because the weather wasn't quite so nice here and I thought you'd fancy a jolly one. This is Symbol World. These are all SVGs but they don't publish them as SVGs because they're worried about their copyright. But they produce them in Illustrator. This gives you an idea of... if you're going to produce an icon library... and we desperately need a public icon library... that's the public, Jonathan Chetwynd published weather symbols. I don't know if the BBC are going to attack me but you need at least that many and the BBC change the meaning of them very regularly, so it makes it a little bit tricky. YouTube has made a huge difference to this authoring issue, because people with learning difficulties can... it's still mediated quite heavily, but they can be involved in videos... There are some wonderful videos there. Some of the Peepo site links to it somewhere.

This is a 2.1D space, that you're supposed to be able to navigate. [Points to screen] So that's their Yahoo Newsfeed. The BBC have been very recalcitrant about putting up any pictures. They're claiming they've don't have copyright and they can't put pictures with their news. This came up because I was doing weekly newspaper with them at the time the planes went into the Twin Towers. It was only local and it was published on paper and it wasn't too much of an issue. I very strongly objected to the fact that the news is mediated... I've asked Yahoo and the BBC and I had a potential source of funding... and I think it's the disenfranchisement of the youth, I'm talking about teenagers. If you go to kids.yahoo.com/news, you'll see what I'm talking about. Please go and have a look and you'll see the type of news that Yahoo are producing, supposedly for children. It's not news... even to call it news is ridiculous. It's my belief that children and teenagers should be involved in mediating the news, i.e. using news photos and video footage and creating what their understanding, what they're trying to do with the news. And we need public access. Grotty GIFs of that quality have no copyright value whatsoever. It's completely absurd that the BBC are publishing in their feed graphics of very low quality. These photos, Yahoo are linking to them, Reuters, and the BBC are not.

Finally, I've produced... my grandmother's 94 and lives in Corsica. [Points to screen] This is something I've produced for her and I actually produced it for her in printed form. She lives in Corsica. This is a road. The river's here actually but you can't see it. You can imagine perhaps. That's a road. A little road. A windy road. It's call maquis – low bush. It's very low densely populated. There's two buildings in this area, that one and that one. That's the Google, going off their webpage. And that's to show you the sort of symbols. And she's actually got an olive mill. This is a horrendous, as you can see, hair-pin bend danger spot. It's not the best icon but I've got a busy life and it was the nearest I could get to something that was understandable! It needs a little car or something with it and then you'd understand that it was a danger spot.

Thank you very much for letting me come and speak.

End of Putting the User at the Heart of the W3C Process: Back to Top of Page