Blind for a week! A web developer's experience

[Stomme poes]

So webdev David Ball wanted to really experience needing to use a screen reader for all his regular web surfing needs. Some things he discovered:

1. His screen reader wasn't of course limited to the browser.

2. It was hard.

3. Browsers are different (I dunno which screen reader he used, but screen readers tend to work better in certain browsers as well. For example JAWS works better in IE while NVDA is best (so far) in Firefox. We won't mention VO cause that's unfair)

4. Raise the speed to chipmunk if you want to actually get through a page before you turn 60 (writing for mobile, in the Jakob Nielsen sense, would be good here too)

5. Some populare websites... (ZOMG you guessed it) really suck and are poorly written. Versions written for mobile are... almost usable!

6. Link titles are a waste of bytes. yes older versions of turdpress and current versions of joomla and *EVERY* SEO "specialist" under those rocks, we are talking about YOU

7. Autofocus sucks. (Why it should have been written so: let users choose to add autofocus, not authors... hint hint browser vendors, cause we can't just turn JS off anymore thank you)

8. Surprise (not): you can be technically valid in HTML, doesn't mean you're semantic or correct. A steaming pile of divs and <span role="button"> is just as smelly by any other name

9. Heading-navigation rules!

10. We should be thankful every day that a rampaging army of blinks (Gonz' word, but I like it) don't hunt us webdevs down and eat our livers. We should be able to tell a blind web user by their lack of hair, since they pulled it out all long ago trying to navigate our steaming piles of HTML5/DHTML/AJAX/ZOMGWEARESOAWESOME code.

http://blog.silktide.com/2013/01/thi...nd-for-a-week/

[Sega]

5. Some populare websites... (ZOMG you guessed it) really suck and are poorly written. Versions written for mobile are... almost usable!

Could you elaborate on this point? Does responsive negatively affect usability for the blind?

Edit: I now read the article, this is really really shocking! We really have to give more focus to this. Sounds ludicrous that the EU should pass a cookie law, but do nothing to encourage the blind to browse their web pages.

[Stevie D]

Could you elaborate on this point? Does responsive negatively affect usability for the blind?

I think that what poes means is that a website that has been designed to be used on a full screen, and only on a full screen, is more than likely going to be an absolute nightmare for anyone using a screen reader. A website that has been designed to work on a mobile (especially if it is written just for mobile, I don't know where responsive designs would feature) is much more likely to work better for screen readers.

[Sega]

I think that what poes means is that a website that has been designed to be used on a full screen, and only on a full screen, is more than likely going to be an absolute nightmare for anyone using a screen reader. A website that has been designed to work on a mobile (especially if it is written just for mobile, I don't know where responsive designs would feature) is much more likely to work better for screen readers.

I have no idea what would normally happens. I initially thought that screen readers stripped out the styles and showed their text in hierarchical structure of their headings. Images would be replaced with their text equivalents. I don't know how a responsive website would negatively affect screen readers.

After reading the post I can see that it might not be the website developer's fault, but more to do with too much content on a page. Having 1000 links would certainly complicate things. Having said this, they could place those links on the very bottom. I might have to try the experiment myself and see the overall experience.

On that note, do blind people use traditional computers, or computers with monitors for that. In terms of paper documentation, any government documentation has a coded message on the bottom with a contact number to order it's braille equivalent, so as you can imagine there is a lot of support here.

[ServerStorm]

I think that what poes means is that a website that has been designed to be used on a full screen, and only on a full screen, is more than likely going to be an absolute nightmare for anyone using a screen reader. A website that has been designed to work on a mobile (especially if it is written just for mobile, I don't know where responsive designs would feature) is much more likely to work better for screen readers.

I have a friend that uses JAWS and she tested a couple of recent responsive sites I did. I was really interested in how much hell the forms would be. Overall the responsive functionality held up very well. As it serves just one site based on the browser sizing it is render the way it is, so as long as the document is structured semantically and it's not jammed full of JavaScript for primary rather than enhanced functionality then, JAWS works quite well.

[Stevie D]

I have no idea what would normally happens. I initially thought that screen readers stripped out the styles and showed their text in hierarchical structure of their headings. Images would be replaced with their text equivalents. I don't know how a responsive website would negatively affect screen readers.

I don't think that responsive designs are haveing a negative impact; there's noting that says they are worse than regular sites, and the implication (re sites designed for mobiles) is that are quite likely better.

Part of that will be because – whether you're looking at a separate mobile page or a responsive design – there's a good chance that the author has put more time and effort into thinking about the structure and semantics of the page, and that makes it easier for screen readers to sort out the hierarchy and the order.

[Stomme poes]

It's much more likely that mobile versions of spacewaste, amazon, news sites etc are
- written more recently and therefore was maybe an opportunity for fresh devs to write new, better, modern code (I noticed this at funda.nl, where the mobile version was using the hip new HTML5-style doctype and way way way less code... while the main site was still filled with old ASP form runserver blah blah and thousands of lines of inline javascript!)

So basically, StevieD's last point holds well.


- As a briefly alluded to, it's encouraged to write even terser content for mobile than for general web. You don't get this with responsive sites really, since usually responsive means the same content but simply styles that are flexible enough for many screens and maybe input methods (larger buttons with more space around them for touch, etc). The Big Sites who have a mobile version, tend to have mobile-focussed content as well.

Take a look at mobile twitter and mobile facebook and see the content difference.
(btw, related: forever-scroll and facebook http://webaim.org/discussion/mail_thread?thread=5671)

- It doesn't matter that a screen reader strips out styles if the code underneath is smelly and gross and bloated. Also keep in mind devs tend to do a lot less loading of content via JS than on main/desktop versions.

[Stomme poes]

I missed this earlier:

On that note, do blind people use traditional computers, or computers with monitors for that.

Usually, most people have regular computers, with monitors, and keyboards. Windows is still the most popular OS simply because... it's the most popular OS. Macs are increasing though and VoiceOver is built-in with all OSX systems, though by default it's English even if you bought the machine in a predominantly non-English-speaking country and the OS language is otherwise set in that other language... you have to go find and buy/get your own voices for your language for VO. Shame, really...

Some of the Linux users on the Orca mailing list have been figuring out a safe way to turn their screens off because, with laptops, batteries matter. But usually there's the regular monitor and of course this makes sense if anyone is sighted in the house and they also want to use the computer.

Some people have a Refreshable Braille Display hooked up, even though the percentage of the blind population who are Braille literate is low, like 20% or something. They're hideously expensive. They come in different forms: some have 16 characters, some have 32. Some have one row, some two or more. They're like a mouse, you just plug them in and you have software on your computer that the Braille display talks to. It can usually also work together with the screen reader and screen magnification/zooming software.

Here's I think a nice demo thingie movie of a RBD, here hooked up to an iPad: https://www.youtube.com/watch?v=tVuLGrab9JA

These thing just sit as regular software on top of the same software most other computer users have. The screen reader software gets information from the applications running on the Desktop for example (so this is the OS or at least the OS's windowing system getting involved, not just the applications... you can see now why it's more difficult for, say, Firefox or OpenOffice developers than for Mac software on Macs or IE/Outlook/Office/etc on Windows). So if you want to write a paper on Windows, you'll still use Word; if you're writing an email, you can use Outlook, or web mail, or whatever.

There is a Linux distro however that's geared specifically to the blind, as a distro you can install from scratch without sighted assistance and with all the normal built-in accessibility stuff turned on by default: Vinux. But Vinux gets installed on any Linux-compatible computer.