I've spent about four years as a medical tech working with epilepsy in a neurology lab. Sorry, but this just gets up my nose.

Please stop using 'flashing images' as a way to co-opt epileptics into defense of the issue-of-the-day (that's the only thing I can think of as to why you included it). Yes, some epileptics do respond to flashing images (~15% have the kind required, less actually respond), but the parameters are pretty tight for the bulk of them. The kind of flashing that can trigger a seizure is also monumentally irritating to regular users, and generally hasn't been used since the 90s.

I spent four years in a full-time job where I sat in the dark flashing lights at epileptic and suspected epileptic patients, 4-6 times a day while recording EEG, so I've got a reasonable subjective feel for the kind of flashing that fires it off. I've spent a lot of time on the web, too, and the only thing that even comes close in general use are the flashing-style 'you are the 999999th visitor' ads, and even then they usually subtend too small an arc of the visual field.

For the record, the usual range for photosensitive epileptics to respond to flashing images or lights is 12-20Hz, with a peak different for each patient and with some rolloff at each end. Above 25Hz epileptic responses are extremely rare - I've not personally seen one, but they do exist.

Yes, there are some epileptics that can have a seizure from a single camera flash, and I've seen them, but they're reasonably rare amongst flash-responsive epileptics (themselves an uncommon group). I've also heard of an epileptic whose trigger was orange circles - not orange other shapes, or other coloured circles (anything can be a trigger for a seizure). At what point do you set the bar for rarity?

Sorry, but it gets up my nose - almost every time I've seen someone 'defend epileptics' online, they don't actually know much about epilepsy, and seem more to be using common mythology about epilepsy in order to co-opt the illness to strengthen a point being made.

Of course, two other common forms of trigger, depending on the kind of epilepsy, are stress and hyperventilation (strictly speaking, these are less triggers and more threshold-to-event reducers). In this case, one could argue that frustrating site design works against epileptics...

I included it to make the following point: Accessibility is more than catering for the deaf or blind. It also includes color blindness, mental handicaps or going by the standard:

http://www.epilepsy.org.uk/info/photosensitive-epilepsy/web-...

http://www.evengrounds.com/blog/assistive-technologies-for-p...

Which have the following guidelines (or myths?):

  - allow users to control flickering, avoid causing the screen to flicker

  - allow users to control blinking, avoid causing content to blink

  - allow users to freeze moving content, avoid movement in pages

  While people with epilepsy are browsing web sites, they 
  may encounter pages that have blinking texts and 
  animation that may trigger seizures. Sudden loud sounds 
  and repetitive audio in some web pages can also cause 
  epileptic seizures.

I do not know much about epilepsy, so I will believe your experience, but believe me when I say I am not using it as a way to further a point. I believe in above guidelines, regardless if they help for epilepsy, because in this case they help other users with a mental handicap or ADHD and so remain good guidelines. Or do you rather have we don't mention epilepsy all together when making sites accessible and just call them common sense? It is easy to ignore even a common sense guideline when there are so many different ones.

It seems you have an uphill battle to fight if mentioning epilepsy and accessibility in one sentence gets up your nose. Like my quotes show, I am not the fringe 'defend epileptics' spreader of myths here, this is part of the accessibility topic.

It's fair to say that I overreacted a little, but I did admit it (via 'sorry, gets up my nose'...)

What I was really reacting against was if you search hard enough any site will breach some accessibility guideline for a ... epileptic user. The other three, sure. But I just do not come across any kind of website offering a service that violates these principles. You could throw a brick and find any website that violates guidelines for people with blindness or mental illness. You'd have to search for specific audio-based sites to find ones that discriminate against deaf folks (I can't see how HN does so, for example, but I may be missing something). But you'd have to go through a lot of service-offering sites to find a serious one that caused problems for epileptics.

The only kind of violation of these principles I could think of is a bug anyway - the issue where a mouseover changes the shape of a menu, moving it out from under the pointer, reverting the menu back to original shape, repeat ad infinitum.

Apart from that, I just don't see websites that cause problems. Served ads on some low class websites, perhaps (999999!), but not the content that the website itself is providing.

Solutions for epileptics to use poorly designed sites (if they were a real problem like they are for the blind) are here anyway - [Esc] in pretty much every browser cancels image animation. Flashblock kills flash dead, allowing you to throttle it to your needs. If you're really, really susceptible to simple movement as an epileptic (I've never heard of it, but it could happen, I guess - see the orange circle guy), you can install noscript, which breaks a lot of websites, but kills non-image animation dead. But if you're that susceptible, you're extremely rare and the web probably shouldn't be designed around your use case.

It seems you have an uphill battle to fight if mentioning epilepsy and accessibility in one sentence gets up your nose.

Please read my comment again. It's not 'mentioning them in the same sentence', that's a strawman.

This is true, but Dreamwidth is a blogging platform (it's a fork of Livejournal), so those errors weren't caused by the post's author. Presumably it's accessible enough for the user in question to be able to navigate and post at, though, since they say they're using a screenreader; I assume the "disabled and disgruntled" tag means this wasn't just an academic exercise.