PDA's

PDA's

[Buddy Brannan]

If I had to make a start of it, I'd say right off--not the Palm. The Palm
has to the best of my knowledge no expansion capability (I.E. no PC card
slot); the Visor might be something to consider, however, if their
Handspring module designs are available somewhere.

--
Buddy Brannan, KB5ELV
Email: davros@ycardz.com
Voice mail: 877-791-5298
All opinions are all mine!

Re: PDA's

[cpt.kirk]

I would agree that the Palm might not be the best place to start. The
visor seems a much better place to start, but has a drawback. They use USB
instead of serial for the standard IO. But, they do have an expansion
slot, and one can get the specs for the slot. I don't know if there is a
fee to get the specs or not. I would love to know what the performance of
the thing is when running linux.


Kirk Wood
Cpt.Kirk@1tree.net
------------------

Your fly might be open (but don't check it just now).

Re: PDA's

[Buddy Brannan]

Hmm. Well, USB won't be a problem when 2.4 comes out, I think; and it'll
have that schnazzified read-only file system. Should be interesting when
released.

--
Buddy Brannan, KB5ELV
Email: davros@ycardz.com
Voice mail: 877-791-5298
All opinions are all mine!

Re: PDA's

[cpt.kirk]

The more I look, the more I think putting together something might be the
way to go. The question would be what is the minimal requirements for a
usable system?

I know that a 386 can be made to boot and run. Along the same lines, I
think 2 MB can be done with 4 being the minimum for most distributions.
But that being said, what is the real minimum for us? Also, what price
range is acceptable. I am sure that the complexity of the app will have an
effect on the minimum requirements. Along those lines, if you go with
something put together ourselves from either DIMM modules of PC104
modules, we can allow for more options. The bottom unit might be a 386
with 2 MB of ram. The max would be about a 486 with 16 MB ram and 240 MB
of disk space (solid state).

Kirk Wood
Cpt.Kirk@1tree.net
------------------

Your fly might be open (but don't check it just now).

Re: PDA's

[Kirk Reiser]

<cpt.kirk@1tree.net> writes:

> The more I look, the more I think putting together something might be the
> way to go. The question would be what is the minimal requirements for a
> usable system?

Although I agree that designing a system built totally to our specs is
an intriguing-- idea; I don't think it is really a practical idea.
The object is, correct me if I'm wrong, to come up with a low cost pda
for blind people that will compete with what is available on the
market today.  The reason for this is to provide more flexability and
ease of use in a smaller less expensive package.  If this is an
appropriate evaluation then you will be required to find something
which is already available and modify it.  There is no way that we can
produce a custom made high end device at an affordable price.  We just
don't have the cost per user ratio necessary for that.  That is why
things like the old Sharp talking clock go by the wayside.  So, if we
want to accomplish anything, then I think we should concentrate on
what's available currently off the shelf.

  Kirk


Kirk Reiser				The Computer Braille Facility
e-mail: kirk@braille.uwo.ca		University of Western Ontario
phone: (519) 661-3061

Re: PDA's

[cpt.kirk]

We can certainly investigate using currently available PDA's. It would
seem to me that the easiest (and cheapest) route to start with is to setup
the Palm emulator and UCLinux. Actually, this might not be a bad route
anyway. If the platform will work, one can order a a simm module that
carries all that is needed. The distribution is available at
www.uclinux.org. They also offer a simm module with a working computer on
it that could be the basis for a customized PDA. I will track down the
Palm emulator. It is available free.

I will post the URL for the emulator after finding it. If someone needs
these on CD let me know. I can burn them and mail them for a few bucks to
cover my costs.

Kirk Wood
Cpt.Kirk@1tree.net
------------------

Your fly might be open (but don't check it just now).

Re: PDA's

[Geoff Shang]

On 11 Mar 2000, Kirk Reiser wrote:

> don't have the cost per user ratio necessary for that.  That is why
> things like the old Sharp talking clock go by the wayside.

Weren't they great.  <sigh>

Geoff.

Re: PDA's

[craig martin]

hi, guys.  You could use a tripple talk (if it's out then) for a synth
untill the system was more developed.  The only advantage you have is that
the tripple talk has its own power supply, and you don't have to worry
about software synths until you get going with a stable operating system
and input interface.

Re: PDA's

[Greg Keto]

I can't see a pda audio interface being done right in any O/S other 
than the palm O/S, CE is out of the question, why should an audio
interface have to run through a GUI? Palm/Visor I think they are both
Palm O/S based, with 3com Palm we had an "Open software architecture"
and with Visor we have an additional "Open hardware architecture"
to develop pda peripherals.

Software speech synthesis for the Palm is crude, Im not sure if its 
a limitation of the mc68328 dragonball processor or the simple 
sound circuitry in the Palm pilot, probably the sound circuit. I have
read stories of Soviet software engineers doing amazing things with 
early Intel processors during the 80's.
 
While the mc68328 is probably a cabable processor for speech synthesis,
its the architecture and storage of the palm pilot pda that limit it for
good software speech synthesis, speech recognition is probably out of
the question. Hardware speech synthesis is probably the best solution
for a good sounding audio interface for the Palm pilot. For input a 
single handed chorded key device would be optimal. Probably cumbersome
to learn, but once learned the combination of good speech output,
chorded key input and a well designed audio interface will provide an
extremly powerful man/machine schemata that would make J.C.R. Licklider
proud. A person can be in motion, eyes free not visually captive, palm
pilot tucked away with earphone output. 

I have heard of a scaled down linux kernel running on the the palm 
pilot but didn't hear anything about performance. A company called
Transmeta recently announced support for embedded Linux in its line
of low power processors. The unique capabilities of these processors
will probably make it possible to create speech recognition/synthesis
engines as intregral parts of the physical/virtual processor.

Good to see people are interested in this sort of thing, maybe this 
can lead to the next generation of operating system. I don't know
what happened in the last 15 years as far as computer interfaces, but
it looks like regression to me. Looks like its gonna take the free
software and open source movement to do what the commercial industry
has failed to do in the last 15 years.

Greg Keto


cpt.kirk@1tree.net wrote:
> 
> I would agree that the Palm might not be the best place to start. The
> visor seems a much better place to start, but has a drawback. They use USB
> instead of serial for the standard IO. But, they do have an expansion
> slot, and one can get the specs for the slot. I don't know if there is a
> fee to get the specs or not. I would love to know what the performance of
> the thing is when running linux.
> 
> Kirk Wood
> Cpt.Kirk@1tree.net
> ------------------
> 
> Your fly might be open (but don't check it just now).
> 
> _______________________________________________
> Speakup mailing list
> Speakup@braille.uwo.ca
> http://speech.braille.uwo.ca/mailman/listinfo/speakup

Re: PDA's

[Kirk Reiser]

Greg Keto <gorgotek@gte.net> writes:

> I can't see a pda audio interface being done right in any O/S other 
> than the palm O/S, CE is out of the question, why should an audio

Hi Greg: I would be interested in hearing your reasoning with respect
 to the above points. I don't know the Palm O/S but it seems to me
 that an audio interface or any other interface shouldn't be better or
 worse than any other OS.  I particularly don't understand why you
 think C is out of the question.  With a good compiler like gcc you
 can create better tighter code than you can in assembler if you're
 not an experienced assembler programmer.

> Software speech synthesis for the Palm is crude, Im not sure if its 
> a limitation of the mc68328 dragonball processor or the simple 

I think our software synth will probably be pretty crude to start off
with as well.  One of the aspects I am always fighting with is quality
vs size.  Although size isn't quite as critical in user-space, if you
want to use it for kernel code then you need to be very careful about
it's size.  The nice thing about publicly available code though, is
that it can be improved on over time.  Providing however, that you
have people willing to work on it.

> the question. Hardware speech synthesis is probably the best solution
> for a good sounding audio interface for the Palm pilot. For input a 

Once again, this is true providing you can design and implement a
hardware synth solution.  I know of two pcmcia speech synths: the
Accent Messenger and the Keynote one.  The Accent has no onboard
smarts so it needs a large driver to make it talk.  I know nothing
about the Keynote product.

> single handed chorded key device would be optimal. Probably cumbersome
> to learn, but once learned the combination of good speech output,
> chorded key input and a well designed audio interface will provide an
> extremly powerful man/machine schemata that would make J.C.R. Licklider
> proud. A person can be in motion, eyes free not visually captive, palm
> pilot tucked away with earphone output. 

Pretty cool imagery isn't it?  I am sure we can come up with
something, how elligant it will be will depend on a lot of facters.

> I have heard of a scaled down linux kernel running on the the palm 
> pilot but didn't hear anything about performance. A company called

There are Linux ports to a number of palmtops.  I have seen Linux
running on the Psion series 5 and it is quite impressive.  The biggest
difficulty is going to be with storage size.  I am not sure what the
largest flash-cards are these days.

> Good to see people are interested in this sort of thing, maybe this 
> can lead to the next generation of operating system. I don't know
> what happened in the last 15 years as far as computer interfaces, but
> it looks like regression to me. Looks like its gonna take the free
> software and open source movement to do what the commercial industry
> has failed to do in the last 15 years.

For the most part I agree with this statement.  I think the computer
future looks great for blinks providing they can give up their
MicroSoft habit.

  Kirk

-- 

Kirk Reiser				The Computer Braille Facility
e-mail: kirk@braille.uwo.ca		University of Western Ontario
phone: (519) 661-3061

Re: PDA's

[Greg Keto]

Hi Kirk, read an earlier post of yours, I also have a blind friend.
My statement is referring to the Windows C.E. O/S not the C language. 
I think the Palm O/S is well suited for an audio interface compared to
Windows C.E. O/S because the all applications on the Windows C.E. O/S
have to work through the Windows GUI.

Greg Keto
 
Kirk Reiser wrote:
> 
> Greg Keto <gorgotek@gte.net> writes:
> 
> > I can't see a pda audio interface being done right in any O/S other
> > than the palm O/S, CE is out of the question, why should an audio
> 
> Hi Greg: I would be interested in hearing your reasoning with respect
>  to the above points. I don't know the Palm O/S but it seems to me
>  that an audio interface or any other interface shouldn't be better or
>  worse than any other OS.  I particularly don't understand why you
>  think C is out of the question.  With a good compiler like gcc you
>  can create better tighter code than you can in assembler if you're
>  not an experienced assembler programmer.
> 
>   Kirk
> 
> --
> 
> Kirk Reiser                             The Computer Braille Facility
> e-mail: kirk@braille.uwo.ca             University of Western Ontario
> phone: (519) 661-3061

Re: PDA's

[Kirk Reiser]

Greg Keto <gorgotek@gte.net> writes:

> Hi Kirk, read an earlier post of yours, I also have a blind friend.
> My statement is referring to the Windows C.E. O/S not the C language. 
> I think the Palm O/S is well suited for an audio interface compared to
> Windows C.E. O/S because the all applications on the Windows C.E. O/S
> have to work through the Windows GUI.

Sorry about that!  It was one of those synthesizer things.  Se, C
seniore!

As far as Windows interfaces go, I don't think we're interested in
anything which isn't Linux based.  You have to know a good thing when
you find it.

  Kirk

-- 

Kirk Reiser				The Computer Braille Facility
e-mail: kirk@braille.uwo.ca		University of Western Ontario
phone: (519) 661-3061

Re: PDA's

[cpt.kirk]

I don't think we are looking at using either the winCE or the palm OS. We
are looking at using Linux because we can modify it and there are no
license issues. Besides if you run wince together you get the right effect
of the OS. The only possible advantage is that if you can pus the OS out
of the way the devices have more memory. The OS has major problems though.

There is a port of Linux that is aimed at WinCE machines. Haven't had the
chance to read the pages on it yet. But I don't see us having the
resources to chase an extra platform. Besides, the PalmOS devices are
available in a form that allows the complete displacement of the OS. such
is not true for WinCE. But the whole thing is we are not looking to move
to another OS. we are looking to adapt hardware to Linux.

Kirk Wood
Cpt.Kirk@1tree.net
------------------

Your fly might be open (but don't check it just now).

Re: PDA's

[Greg Keto]

Hello Kirk W.

The Palm O/S is "open", you may acquire and probably modify it to 
suit your hardware needs. If your stuck on Linux for a pda O/S you
should check out www.transmeta.com. IBM, Transmeta and others just
announced the establishment of the "Embedded Linux Consortium". The
Transmeta "Crusoe" Processor will be the premier processor for this
type of application.

Greg Keto

cpt.kirk@1tree.net wrote:
> 
> I don't think we are looking at using either the winCE or the palm OS. We
> are looking at using Linux because we can modify it and there are no
> license issues. Besides if you run wince together you get the right effect
> of the OS. The only possible advantage is that if you can pus the OS out
> of the way the devices have more memory. The OS has major problems though.
> 
> There is a port of Linux that is aimed at WinCE machines. Haven't had the
> chance to read the pages on it yet. But I don't see us having the
> resources to chase an extra platform. Besides, the PalmOS devices are
> available in a form that allows the complete displacement of the OS. such
> is not true for WinCE. But the whole thing is we are not looking to move
> to another OS. we are looking to adapt hardware to Linux.
> 
> Kirk Wood
> Cpt.Kirk@1tree.net
> ------------------
> 
> Your fly might be open (but don't check it just now).
> 
> _______________________________________________
> Speakup mailing list
> Speakup@braille.uwo.ca
> http://speech.braille.uwo.ca/mailman/listinfo/speakup

Re: PDA's

[craig martin]

the only problem with speech input is noise and memory
requirements.  Speech output was done on the Comodore 64  and amega series
back in the early ninties.  so we just have to solve the resource problem
on speech input.  C would lend itself to this since it works on Linux, and
the speaker and microphone are treated as files like any other device in
the /dev directory.