next »

[LesleyT]

I really am in love with EdWord from deafblind online

http://www.deafblindonline.co.uk/software.html

I wouldn't need something as dramatic as that for the XO and I wouldn't need the grids -- I think the small screen space would the grids less helpful.  But I love the talking word processor concept. 

I'd love to have a word processor that echos each keystroke and then echos each word as the spacebar is hit.  Maybe a button to push to have the whole document read to you.

There is quite a discussion on the accessibility list about screen readers, but I don't think my talking word processor would need all of that.  We already have the activities that do what I'd want a WriteandSpeak to do. 

Write -- obviously
Then for the echoing of the keystrokes, talkntype already does that.
And reading imput -- speak obviously does that.

Anyone up to marrying the code and making my talking word processor would be my best friend. I'm trying, but it's beyond me right now.

[moocapiean]

I'm willing to give it a shot, but I don't have much experience with python or programming for Sugar.  How important is it to you that it use Write?  For at least my first attempt, I think I'd rather just get a speaking text editor working (think Notepad+talkntype) and maybe later integrating it with Write.

[tdang]

moocapiean-

I'm not at all sure if the parts of Write that are relevant ARE in Python. I really don't know, but after LesleyT posted the message, I went poking around the Write code. It looks suspiciously like it uses a non-Python Canvas-kinda object for the text space.

Anyway...it's confusing. But maybe you're in luck that it's more up your alley than Python.

[moocapiean]

Before I posted, I poked around the Write source code a bit and that's the impression I got too.  But, not being very familiar with python, I wasn't sure.  Thanks for confirming!  I'll probably stick with python since it looks fairly easy to get a basic text editor working.

[LesleyT]

I can't code my way out of a cardboard box so ANYTHING anyone comes up with would be awesome!  I have a special needs classroom of my very own this year and I want to impliment the XO as AT -- not just play with it. But right now, I still am very much playing

[Tortuga]

I'm not sure if this is a viable option or not, but there are people working on getting speech-dispatcher working on the XO, if that might be of help?

Project page:
http://www.freebsoft.org/speechd

Several posting on the development board:
spd-say & speech-dispatcher

[moocapiean]

Attached is a first-version if anyone wants to take a look.  It'll repeat each letter or number typed, then say the word whenever you type whitespace (space, tab, enter) or a punctuation mark (comma, period, !, ?).  There are buttons for reading the entire document and for reading just a selection.

One problem you'll notice is that you can't type too fast, otherwise it'll fall behind and skip over letters. Also, there's no way to save and when you try to exit the activity, it'll warn you that any changes will be lost.

I'll continue working on this, but I wanted to release what I had so others could play around with it and offer suggestions.

Tortuga: Thanks for the link.  I've been using espeak from the Speak activity since I don't know if speechd comes with XO yet.

[Gabey8]

OMG! How excellent is this?

What do we do to download it onto the XO -- just click on the .xo file? Will it appear in the frame on Sugar?

[moocapiean]

Well, the way I've been installing it is to get it onto the XO, then using xo-get to install it.

Once xo-get is installed, I install WriteNSpeak by typing into Terminal:
xo-get install writenspeak-1.xo
where writenspeak-1.xo is the path to the file you downloaded.  Unfortunately, I'm not sure where Browse downloads files to.  What I've been doing is just transferring writenspeak-1.xo from my laptop to my XO using a USB drive.

[Gabey8]

I've done some experimenting with writenspeak, and I've gotten it to read off some pretty lengthy bits of text. I copied and pasted all the text from a webpage into the program, highlighted it with CTRL-A, and clicked the button to read selected text.

Oddly enough, if I click "read all", the voice gets all garbled as though it's trying to pronounce multiple things simultaneously and the words are stepping on each other. But if I highlight all the text and click "read selected", it works perfectly fine and every syllable is pronounced clearly.

I have a friend who's coming in to Philly next week with a BrailleNote. (See http://www.pulsedata.com/products/Notetakers/BrailleNoteQT.asp for details.) That device, along with having a braille display, has speech output for everything that a person types in. My friend normally has the speech output turned off, as she's deaf and cochlear implant technology doesn't render computer-synthesized voices clearly enough for her to understand the words. But I might ask her to turn the speech on just enough so I can hear what it sounds like to use the word processor on her device. I'm curious to know what additional features can be employed, now that there's a good basis to work with with writenspeak.

My cats, BTW, are not thanking you because now I've got the XO talking all over the place while they're trying to nap. And I keep doing things like typing in their names so it keeps saying their names. I think they're of the mindset that if the computer isn't going to pet them or feed them, then it should stop calling them.

[LesleyT]

OH MY GOODNESS!!  Awesome! 

I have been away while I've moved and so it's really neat to come back and see this!