Powered by Blogger

Tuesday, November 13, 2007

JAWS 9 in Vista and more on Free Open Source AT

My new, Vista based desktop came back from the repair shop as good as new.  The first thing I did on it was to install the JAWS 9.0 public beta with its updates.  I can only describe the experience as a bit underwhelming.  Unlike my first series of articles about Vista, I did not do anything even approaching an objective comparison between JAWS, Window-Eyes and System Access but, instead, just lived with JAWS for a week or so.

 

As I wrote in my first article about the JAWS 9.0 beta, I think it is very good and even excellent in some places.  As I also wrote in that article, JAWS 9.0 is light on new features and even lighter on innovation.  JAWS works better in parts of Vista where it worked poorly or not at all in the past, a definite improvement. 

 

Because of my repetitive stress injuries, though, I had to switch to System Access to write this blog entry because JAWS, in its out-of-the-box form, still does not support Vista speech recognition, an important aspect of the Vista OS distribution.  JAWS also doesn’t work terribly well in the Windows Media Center software that came with my HP.  Both of these areas work pretty nicely with System Access.

 

For the most part, though, JAWS 9.0 is a big step up from 8.xx as its performance has increased noticeably and there are fewer really nagging bugs.

 

As I have written before, I do not think that it is fair that the most expensive screen reader on the market relies on volunteers to create the configurations and scripts for applications that fall outside the 90% rule.  On the other side of the coin, though, JAWS does work out-of-the-box with more programs and, unlike their competition, the ability to modify JAWS outside of FS puts a lot of power in the hands of the more nerdly inclined.

 

The three (or is it four) projects I currently work on will release their entire software open source under GPL.  Matt Campbell wrote an excellent comment the other day when I did an item on free (as in freedom with a lower case “f”) software.  Matt makes some very valid points but I think the following items express a number of reasons why AT should move to an open source model:

 

·         While Matt suggests that a free screen reader will only increase the sense of entitlement that many blinks already have, I counter with: if a sighted person goes to Best Buy, purchases a new PC, brings it home, attaches various cables and such, turns it on and after the OS installation and registration nonsense completes, do they feel “especially entitled” because they can use it without plunking down $500 for System Access or more for its competitors?  Following the civil rights argument, would a restaurant be considered integrated if they let minorities in but charged an extra $10 for every $3-4 item on the menu when serving black customers?  Why then should blinks have to pay all the extra cash just to use what our sighted friends get for free?

·         Screen readers, by their very nature, cause a lot of security problems (orca and VoiceOver may not as I haven’t explored how they gather data beyond the API.  At an ATIA sponsored AT/IT meeting a number of years ago, Madelyn Bryant McIntyre (then Director of MS ATG) and I brought a number of these issues to the table.

·        If a screen reader can function at the Windows login prompt, it can alsostore your user name and password and send it off to some nefarious hacker to use later.  Not coincidentally, a screen reader can read what you type into a user name and password field on a web page which may store your credit card information.  An open source solution can give organizations (government agencies and such) a way to look at the code and ensure that no one built in a back door.

·        In the past (this may no longer be true) some screen readers found ways to hack the Windows kernel without causing the OS to pop up a warning message.  This not only created a security hole, it prevented the user of being warned about it.

·             Many people in developing countries believe the best screen reader available is JAWS in its 40 minute mode.  Thus, no matter their task, they save their work and reboot every forty minutes.  To people who only earn 500 dollars per month in technology jobs, purchasing a full featured screen reader is out of the question.

·             Unless the software is open source, adding languages not determined to be profitable by the vendors of proprietary screen readers, is impossible.  Just take a look at the rather long lists of languages supported by orca and NVDA.

 

There are far more reasons for doing things under GPL, open source and leveraging massive collaboration than I can think of at this moment.  I will renew my JAWS SMA and continue to use it while working on GPL solutions for myself and the rest of our community to use.

 

-- End

1 Comments:

Anonymous Anonymous said...

I think there is another valid argument in the favour of open source free AT. The argument is that open source software is typically not driven by the desire to make a profit. Contrast this with the motivation of commercial AT vendors to make a profit. Commercial AT vendors are companies and the purpose of any company is to make money.

The motivation to make money can affect how people go about developing AT. If people are motivated by the need to make money then they are likely to develop software in a way that maximises their future potential to make money. One way in which this is done is by maximising the number of new features that can be added in future versions of a product.

This tactic of maximising the number of new features isn't just a superficial thing. It affects the fundamental concept of what a screen reader is. The current concept of a screen reader is something that not only reads what is on the screen but also something that simulates attention. This means that screen reader vendors have to produce features for two distinct areas.

The need to simulate attention is an artificial requirement that has arisen as a result of how screen reader vendors have conceptualised a screen reader so far. The reason why I consider it an artificial requirement is that attention is an inate human ability that spans the senses; therefore, attention is found in audition just as it is found in vision. The need to simulate attention has come about because of how screen reader vendors have chosen to present information to a user.

The need to make money may have affected how screen reader vendors chose to present information to a user. Choosing a presentation method that requires them to simulate attention gives them a wider scope for new features. This in turn gives them more ways to drive sales of new products and new versions of products. For example, the biggest feature to have come from the need to simulate attention is the quick navigation keys for web pages. I would be surprised if this feature wasn't responsible for having motivated people to upgrade their screen reader in the past.

Simulating attention rather than allowing a user's inate attentional mechanisms to do the job has a direct impact on a user. Simulation requires the screen reader vendors to update their simulation each time something changes or when a user wants new attentional abilities. So, a user has to wait for a screen reader vendor to make changes to their product in order to get these new attentional abilities. Of course, a screen reader vendors could always decide not to make the changes at all. If screen readers were designed in a way that permitted users to use their naturally occuring attentional mechanisms then this wait would be unnecessary and users would have a greater set of attentional abilities than the one that screen reader vendors choose to give them. More importantly, simulating attention is the cause of an entire class of accessibility problems. One feature of the way in which screen readers simulate attention is that they often tie attention to the keyboard focus. If the keyboard focus cannot be set to a control then a user has no means to easily switch their attention to that control. This leaves the user without the ability to find out what that control is, what it contains, and without the ability to interact with that control. This class of problems is an artifact of how screen reader vendors choose to present information to a user; a choice that may very well have been motivated by the need to make money.

6:32 AM  

Post a Comment

<< Home