Left Brain Right Brain

A few people have expressed curiosity in it so I thought I’d run through how its done. At this stage its really very simple but I have plans to extend its operation to automate more of the running of it.

Macromedia first introduced the ability for the SWF format to parse XML data back in the days of Flash 5 (2000? 2001?). It was pretty basic was pretty shoddily implemented as well. It was never quite resolved whether it was an issue with the actual actionscript events or the Flash Player failing to parse these events (my money was on the latter) but suffice it to say, whilst it was great to be able to introduce another area of flexibility to SWF, it wasn’t until the release of Flash MX and the version 7 player that SWF/XML integration became a more reliable process.

It was around that time that Actionscript Guru, Sam Wan released his RSSDataFactory object which made coding the integration of RSS feeds very easy. He claims it only works with RSS 0.92 but it works fine with RSS 2.0 also.

So, I turned to this RSS object to handle the laborious work of processing the RSS. It was originally built to handle just one feed so I had a few tasks to make it operational for multiple feeds. I also had the thorny issue of the infamous Flash Sandbox to negotiate. More on that later.

Customising the RSSDataFactory is a doddle. First of all I simply inserted another keyframe and moved the DataFactory code to the second keyframe and applied a universal stop(); action to all keyframes. So, now I had two frames to play in I could put all my top level links in frame one and the parsed RSS feeds would display on frame two. Easy.

In frame one then I simply listed all the feeds I was going to be utilising as buttons and coded them appropriately.

Now you may notice at this stage that I’m not calling the feed directly. This is due to the pernicious influence of the Flash Sandbox. The Sandbox is a security measure. It doesn’t allow movies or scripts from servers on domains other than the one the SWF exists on to be accessed directly from the SWF. So what we have to do is build a sort of proxy – a script that _can_ access remote files and, as this proxy script exists on our domain, we can access anything it returns. Problem solved. Now you can see why calling remote XML files is pointless – nothing would be returned due to the Sandbox – so instead I call the proxy script. And because I have multiple feeds, I append a name/value pair (....php?feed=1) to each separate button. I can then detect these in my proxy script and hence return the desired script:

Now we turn our attention to this proxy script (which I called ‘antisandbox.php’). What we’re doing here is detecting the incoming value of the variable ‘feed’ fetching the appropriate RSS feed and opening it up for our SWF to process.

There are a few ways to do this, notably using readfile() but they are mostly (and again, notably readfile()) badly insecure. The way to securely get the feeds is to use (in PHP anyway) CURL. CURL opens a decent, protocoled connection to the remote RSS script. Its syntactically very like using a session in that you initialise it, use it then close it:

I should really have used switch to do the evaluation of the feed variable as its quicker. Probably will implement that soon.

So now we’ve got our data, we have to parse it in. Thankfully all this is done in Sam’s excellent Factory. All I had to do was pass it the value of the proxy script with the right name/value pair.

The last line of this script

rssObj.parseFeed(_root.theFeed, this, "receiveRSSObject");

calls the SWF variable I set in the previous frame that specified which value to send to the proxy script (and thus which RSS feed to fetch).

These are quite interesting:

In fame two I built a movieclip template. This template contains an empty dynamic text holder and an empty URL variable. When the RSS feed comes in, depending on how many posts each author specified their feeds help, each one of these posts would duplicate the movieclip and send the title of each post in the feed and the correct permanent link to these posts. It also spaces out each newly generated movieclip below the previously generated one.

My last problem was the ‘back’ button. Going back a frame in Flash doesn’t affect whats in the frame you’ve just left so I had to clear these newly generated movieclips, clear theFeed and reset the whole shebang everytime someone clicked the ‘back’ button.

Note I simply hide the ‘old’ buttons off stage – this speeds up regeneration if a user wants to revisit a certain feed in the same session.

And thats it for now. I will be extending this to make my life easier. For now, if I want to add a new feed I have to edit the FLA and ‘antisandbox.php’ which is a bit of a pain. What I hope to do is write a plug-in for WordPress that will create an XML file from a link category in the WordPress database architecture. I can then use this XML file to auto generate the feed list in frame 1. The second half of this plug in will drag these details from the database and execute them in ‘antisandbox.php’.

Read the recent ‘A Conceptual Framework for Accessibility Tools to Benefit Users with Cognitive Disabilities‘ report from WebAIM, its clear that there’s one overriding concern that has to be addressed before we can successfuly design websites for users with cognitive disabilities. We need to know our audience.

As web designers we (I hope!) perform extensive usability testing with differing Scenarios. I wonder how much of that testing concentrates on accessibility? When we develop our Personas do we cater for our markets and forget that some in these markets are disabled? Some even are cognitively disabled. I’m willing to bet that we don’t. In fact I’m willing to bet that lots of web designers don’t even build Scenarios let alone build Personas representing disabled members of the target markets to move through those Scenarios.

Its easier for the majority of web designers to not bother. And for some web designers who make the laudable decision to make their sites accessible, there’s a belief that simply adhering to Priority 1 (A) of the WCAG means they’ve done all they need to. I know I’m guilty of this on occassion.

However laudable and useful doing this is or isn’t, its still not addressing the key point – we, as web designers, don’t make enough effort to know this sub-group of our audience. We read articles telling us to validate to WCAG standards and we do it but that should not absolve us from getting to know our audience better. The use of Personas and Scenarios has given us a massively useful way of getting to know key users but they are based on aspects of ourselves or people that we know as well as aset of marketing data. The last time you built a set of Personas how many of them had Downs Syndrome? How many were wheelchair bound? I’ll be honest – I have the advantage of having an autistic daughter and thus have an insight into that condition and ways in which I can account for that group of users but 9 times out of 10 my Personas are based solely on marketing data – ages, gender, income, likes, political beliefs, religious beliefs.

Why is this? We’ve all taken on board the figures – the disabled community have a vast amount of spending power – but can we really say we feel comfortable at knowing how we meet these users needs. No, we can’t. Why? Because we don’t know them. Our liberal upbringings state that we need to see the people, not the disability and for all aspects of social interaction I totally agree. But when it comes to building something that works then we need to see the people and the disability. We need to be curious – how does someone with Downs Syndrome percieve an icon? How does someone with Dyslexia read content? What thought processes lead an autistic to make a choice from navigation options?

We won’t know until we ask. Its not rude to ask. Its not rude to be curious. Its rude to assume. I’m not suggesting that you patronisingly go out and befriend everyone you can find with some kind of cognitive disability but what I am suggesting is next time you go to the effort of building Personas for usability testing, try and include someone with autism, or someone with ADD, or a generic learning difficulty. And don’t make the disability an appendage, make it what it is – an integral part of who that person is. If you need (as I do) to learn more about cognitive disabilities then don’t research it on the internet, don’t chat to people on a forum – include these people in your focus groups, get to know them and your Personas will become more accurate as a result.

In a recent blog post, David Kirby repeated the oft-cited figure of 1 in 166 people now having autism. That figure, according to Kirby, used to be 1 in 5000 in the eighties.

He goes on to ask that if autism is solely genetic then where are all the old autistics now?

As a writer, Kirby is pretty good. As a researcher he’s suspect and prone to errors. As a critical thinker he’s woeful.

Something happened between now and the eighties alright. Something that can definitely account for the increase in prevelance. Simply put – the criteria changed. What constituted autism changed between the eighties and now. Changed drastically. Changed so much in fact that a lot of diagnosticians doubt that Hans Asperger would recognise what is now considered to define Aspergers Syndrome, changed so much that people who were once thought of as AS would be today diagnosed as autistic.

The diagnostic criteria in the DSM, which provide a differentiation between autism and Asperger’s syndrome, have been examined by several research studies over the last five years. There has been some criticism from clinicians and research that the criteria do not identify the disorder Hans Asperger originally described. The four cases he described in his original paper would be diagnosed, according to DSM criteria, as having autism not Asperger’s syndrome. (Miller and Ozonoff 1997).

Dr Tony Attwood

So here we are with this sudden reclassification of a lot of peoples diagnosis. Further, those who would one have been classed as having Aspergers are now routinely being diagnosed as being autistic. I don’t have any stats (and I doubt they exist) on these numbers but common sense would indicate that these people make a real and significant (and continuing) impact on prevelance rates.

Going back one stage for a moment – who exactly is Kirby referring to with this ‘1 in 166’ figure? Is he talking about the prevelance of ASD? Or is he talking about the prevelance of Kanners (classical) autism? You see, the words ‘autism’ and ‘ASD’ are, in modern medical history, interchangeable. I don’t know but my gut feeling is that when people quote this 1 in 166 number they are referring to ASD – the spectrum of autism and not Kanners autism in particular.

If I’m right then we have another huge amount of people to add to our prevelance rates – the recently diagnosed. As stated the diagnostic criteria has significantly widened and as it has, the number of medical and non-medical people trained to spot ASD’s has risen massively. It doesn’t take a genius to figure out what would happen if you vastly widened the qualifying criteria for something and vastly increased the opportunity to see if people did qualify.

There are people (I know some of them) who in their thirties suddenly discover that the problems they’ve managed to live with since childhood actually have a name and an overridding condition. These people get diagnosed too.

Why is it that I know about these people and the likes of Kirby and the various mercury zealots do not (or choose not to?). Because I talk to autistic people. I don’t talk to their parents, or their carers or their advocates. I talk to them. I hear their stories and it all comes out – some (most) were diagnosed in childhood but a significant (and growing) number are getting their diagnosis’ as adults. They can do this because more is known about autism and their are more places to go and ask. When you couple this with the widening of the diagnostic criteria then you suddenly include a whole range of middle aged people who weren’t able to be included in ‘official’ stats before.

So can these people account for all of the ‘1 in 166’. I have no idea. But then I also have no idea who Kirby et al are actually counting when they reach these figures. Kanners autistics? Those with Aspgers Syndrome? Those with a generalised Pervasive Developmental Disorder? Tourettes? In the eighties it was Kanners. Today, autism is all of the above.

Variation in the administrative prevalence of ASD is associated with education-related spending, which may be associated with better-trained educational staff who can recognize the problem, and more and better trained in-school specialists who can provide screening. It is also associated with the availability of health care resources. Increased access to pediatricians and school-based health centers may lead to improved recognition of ASD. Interstate variability in the identification of ASD should be taken into account when interpreting the results of prevalence studies based on administrative data and the associated system characteristics taken into account by policy makers working to improve the recognition of ASD.

David S. Mandell, ScD; Raymond Palmer, PhD

The incidence of research-identified autism increased in Olmsted County from 1976 to 1997, with the increase occurring among young children after the introduction of broader, more precise diagnostic criteria, increased availability of services, and increased awareness of autism. Although it is possible that unidentified environmental factors have contributed to an increase in autism, the timing of the increase suggests that it may be due to improved awareness, changes in diagnostic criteria, and availability of services, leading to identification of previously unrecognized young children with autism.

William J. Barbaresi, MD; Slavica K. Katusic, MD; Robert C. Colligan, PhD; Amy L. Weaver, MS; Steven J. Jacobsen, MD, PhD

We observed dramatic increases in the prevalence of autism spectrum disorder as a primary special educational disability starting in the 1991-1992 school year, and the trends show no sign of abatement. We found no corresponding decrease in any special educational disability category to suggest diagnostic substitution as an explanation for the autism trends in Minnesota. We could not assess changes in actual disease incidence with these data, but federal and state administrative changes in policy and law favoring better identification and reporting of autism are likely contributing factors to the prevalence increases and may imply that autism spectrum disorder has been underdiagnosed in the past.

James G. Gurney, PhD; Melissa S. Fritz, MPH; Kirsten K. Ness, MPH; Phillip Sievers, MA; Craig J. Newschaffer, PhD; Elsa G. Shapiro, PhD

I read a great post from Cody regarding SEO tips recently. I don’t agree with every point – in my experience it makes no difference how far up the code block your content is for example – and there’s a couple he missed – notably the importance of submitting your site to the major directories – but its still a great list.

Something thats rarely touched on outside of dedicated SEO communities is the best way to get a good Google PR structure for your site. Its not as simple as get a good PR on the home page. In fact sometimes that can be counter-productive.

What Is PR?

Google’s PageRank (PR) technology is a mark of importance that Google attach to a particular page. Allegedly. There is a significant school of thought that claim PR is useless as a measurement of the success of a site in Google and yet another school of thought think that it was once used seriously by the Google search algorithm but since the Autumn 2003 ‘Florida’ event, Google have been using it less and less and some even suspect its used as a double-blind technique to make SEO black-hatters (those who use dodgy SEO techniques) falsely optomise their pages for PR.

Personally, I think its all about site structure and that PR isn’t so much part of Google’s seach algorithm but more a mark of how well you are doing _in_ that algorithm.

The Trick

Not really a trick – nothing ever is in good SEO – but more a long hard slog to get the right balance. Its very difficult originally but once you get the idea behind it and put in the donkey work you start to see some real results.

Think of a web forum – any web forum, the subject is irrelevant – look at the structure of it. How a web forum builds its structure should be how you are aiming to get a good site balance that will give you a good, well disseminated PR thats passed through to every page of your site.

A forum is tiered. Every seperate tier ‘drills down’ to a more concentrated and specialised area of the forum. Lets take a web forum about cookery for example. Teir one (or level one if you prefer that terminology) is simply ‘cookery’.

Tier two breaks that down into slightly more targetted areas – say, ‘utensils’, ‘cooks’ and ‘recipies’ for tier two headings.

Tier three goes into even greater targetted areas. Tier three of the second tier about ‘recipies’ for example might be ‘french recipies’, ‘spanish recipies’, ‘italian recipies’ etc etc you get the idea. Tier four of each of these tier three areas are your ‘money’ pages.

Lets go back to tier one and demonstrate how we use keywords throughout these levels. On tier one, the keywords are very generic and concentrate _solely_ on the idea of cookery. You don’t make _any mention_ of the keywords you’ll be using on tiers two/three/four/etc (depending on how ‘deep’ your site structure is). Everything on tier one is fairly generic regarding cookery.

Tier two is again fairly generic about the seperate categories you have here and your keywords should make brief mention of tier one but should be weighted for keywords regarding tier two but again, make no mention of any tiers _below_ this level.

Tier three’s keywords make very brief reference to tier one, a decent reference to tier two but are very specific to tier three. Once more – no mention of any tiers below this one should be made.

OK, I’ve designated tier four as our ‘money’ pages – this is as far as my hypothetical cookery site content actually drills to. So this page can mention tier one in passing, make brief reference to tier two, brief reference to tier three but really go to town on this (its own) tier. This is where you start heavy (but not too heavy) mentions of your three and four keyword phrases. Make your pages as readable as possible for humans – _all_ humans – by which I mean make your site as accessible as you can. An accessible site always ranks better as it is readble by both humans and machines alike. If your pages are easily viewable and navigable with styles turned off (or in Lynx) then you’re definitely on the right tracks.

So the goal with your keywords/phrases is to make them more and more specific on each level. Tier one concentrates on a very broad subject and tier four deals with very specific areas within that subject. Tier one should concentrate on one, possibly two generic keywords and tier four should make use of four or five very targetted two/three word keyphrases.

Doing this properly should result (eventually) in a PR of 2, 3 or 4 on your home page and as you get more relevant anywhere between 6-8 on your heavily targetted tier four pages. These tier four pages are where you should aim to get all your backlinks pointing to (a backlink is exactly as it sounds – a link back to you. Please note: a backlink is _not_ the same as a link exchange which won’t get you as good rewards). Depending on your market (and some markets can take a very long time) if your tier four pages get a lot of high quality backlinks then your PR on these pages can get up to the 8-9 level.

Another thing to note – the PR level in the Google toolbar frequently differs from the real PR. For a more accurate picture of your PR, try and find your site in the Google directory. If its not there you need to submit your site to DMOZ as soon as possible as it can take a *long* time to get a site accepted in DMOZ. A site I submitted in Jan 2004 still isn’t listed today for example. Obviously this entirely depends on how busy your category is but don’t leave this standing – sort it out as soon as you can.

I read a post at Matthew’s blog today describing his current feelings towards his own motivation in which he said:

I think a lot of ‘creative’ types would agree that it is very easy to self-diagnose with borderline manic depression, OCD, Aspergers, and a variety of other personality traits that go with the territory of trying to create something that other people will appreciate…

Now obviously with my daughter being classicaly autistic and Aspergers being a companion to Kanners autism (they are both autistic spectrum disorders – ASD’s) my interest was immediately piqued.

I commented that I had noted that a _lot_ of my fellow designers seemed, from reading their comments or blog postings, to have definite ASD characteristics – most notably of Aspergers Syndrome – but I then chickened out and refused to name names.

It seemed only fair though to offer my fellow designers a chance to find out for themselves whether they display a disposition towards autism and to that end I’ll point you to the online ‘AQ’ (autistic quotient) test.

Unlike a lot of online tests, this one won’t tell you what sort of Star Wars character you are or what your Hobbit name is. This one measures your responses to key questions that determine whether or not you have a disposition towards autism. Its developed by a team at Cambridge University led by Professor Simon Baron-Cohen (if you recognise that name then yes, they are related – its his cousin).

If you score 32 or higher then you have an increased chance of having a disposition towards autism. Its nothing to worry unduly about – if you’ve coped fine up till now then there’s no reason why you shouldn’t continue to do so but if you do struggle and you get this score or more then maybe you should see your Doctor.

I’m also happy to tell you more about autism and point you to some great resources if you need it. Add a comment or email me privately if you’d rather.

NB: The CGI script thats processing the results seems to be broken – manual scoring is at the bottom of the test.