Back to Windows...

[Eric S. Johansson]

Andy wrote:
> On 27/10/06, Eric S. Johansson <esj at harvee.org> wrote:
>> And as I said before, the only reason I use
>> Windows is because I need to use speech recognition.  I really need to
>> get on the new accessibility mailing list to advocate for a hopefully
>> easy to build tool that would make it possible to use Windows only for
>> speech recognition.
> Does speech recognition actually work?

It certainly does.  You get to 95% by training the system, you get to 
99% by the system training you.  Of my postings to this mailing list, I 
would say that all but a small handful have been written by speech 
recognition.  If you do careful analysis, you will find there are 
missing words, in proper tense, and things that sound right when you 
read them out loud but aren't the right words (see in proper tense).

I also have something in the ballpark of 6-7 thousand lines of Python 
code that I have written using speech recognition over the past five 
years.  That's just my open source project work and no, I don't code for 
a living otherwise I would not be able to speak anymore.


>> and the very last is tools for those who forced to use
>> unicorn sticks and other cruel tricks developed by accessibility advocates.
> ah obscure peripherals, that might be a problem.
> I suppose it all depends on whether the manufacture is releasing the
> specs for the devices or not.

No, it's that supporting unicorn sex (s.b. sticks)and joystick driven 
keyboard analogs are relatively easy to do and map to existing 
paradigms.  Unfortunately, they put a tremendous load on the user making 
them think they have accessibility and they do after a fashion.  and I 
guess it's better to be only 500 miles behind the starting line instead 
of 10,000.  But still, you are so far back you can't even see the 
starting line.

in fact, this is one of the bad things about handicapped accessibility 
treated by tabs.  They try to map much of the accessibility aid to the 
current paradigm for tab ( temporarily able-bodied) interfaces.

> Personally I think this would be a huge waste of developers time. Its
> not that we shouldn't be supporting accessibility, we really should
> but if a device vendor is actively working to prevent its use on our
> system should we really commit developers valuable time for a few
> users who are financing a company who is trying to prevent our OS
> existing?

As I've pointed out before, I'm not talking about a specific device in 
any one of these circumstances.  But in the case of speech recognition, 
expect to drop something on the order of $10 mil - $20 mil and five to 
10 years assuming you can navigate the patent law minefield for speech 
recognition.

By that time, I will probably be retired or working at Wal-Mart and 
using a computer probably won't be my highest priority.  So, the answer 
is to look for a shorter path.  I believe that path is creating a bridge 
application so that a speech recognition engine running on Windows can 
communicate with a gnome environment.  This could be interesting because 
it would require passing state information back and forth so that the 
Windows environment knows the state of the windowing environment on 
Linux and how to insert and correct text in text areas such as editors, 
terminal windows etc. on the Linux side, besides passing state 
information back, it should know how to display recognition correction 
dialog so that you could run Linux in a virtual machine full-screen on 
your computer and Windows sits in the background literally unnoticed.

The same concept could be applied to other handicapped accessibility 
devices.  And I understand your point that you really want to have these 
things running native on Linux but damn it all, my fucking hands and 
arms hurt all the time clean up to my ears and this persistence of pain 
in different parts of bodies exists for many handicapped people.  Could 
you really look me in the eye and say that you want me to suffer 
physically and with Windows for ideological purity?

If your answer is yes, that's consistent with the general attitude I've 
encountered in the Linux world.  The variant of no but... isn't much 
better because still, nothing changes.  But if that's your answer, I can 
live with it, because that's what I've been living with for 12 years and 
I will keep living with it adapting Windows slowly but surely to be more 
handicap friendly.

If the answer is no, then let's talk about how to make things better. 
It won't eliminate the physical pain.  Nothing but a minor miracle will 
do that.  But it will make things better for people like myself and 
hopefully eliminate one of the frustrations of our lives.

If you don't change things, they will remain the same and handicapp

> That would be why you need to work closely with development people.
> Although you may not be able to write the code yourself maybe you
> could hire a coder to do it for you? Its one of the many advantages to
> open source, you don't need to convince the vendor, you only need to
> convince a developer.
> (don't ask me to write the code though, although I have 8 computing
> programming books near me I am not an excellent coder *yet* :) )

Well, I have a couple of projects that could use some help including a 
friendlier wrapper to dtach which is almost done but understanding why 
it works would be good experience.

> 
>> Sucks to be human.
> Wouldn't want to be anything else

My dogs and cats have had pretty good lives.  And when the pain gets too 
much, they don't have to suffer.  On the third hand, they don't seem to 
be bothered by pain and disability as much as humans.  As long as they 
can bark at the mailman, UPS man, Federal Express man, anyone walking 
by... they seem happy.
> 
> Nearly forgot, are you on 6.06 or 6.10? I hear there are some
> accessibility improvements in 6.10, but I haven't installed it yet,
> doing a full system backup first (5.10 -> 6.06 left me with an
> unusable system, so I am being extra careful this time)

6.06 in virtual machines.  One of my projects will be delivered in 
virtual machine form in addition to the classic tar ball to give people 
a chance to try it out much more quickly than the bag of parts distribution.

As for the accessibility improvements, they don't even come close to 
what I need.  And if you're thinking of suggesting wine for the 
NaturallySpeaking environment, isn't there, will take money, will be 
very limited at first but not really usable except dictation within its 
own environment.  The full command control dictation plus selection by 
voice is going to require something like the bridge application I 
suggested above.

I swear to g*ddess, this crap gets you coming and going