1
Programming using the Ultra Hal Brain Editor / Brain swapping
« on: August 07, 2004, 07:22:57 pm »
Hi guys,
looks like you got speech recognition working. I've tried dragon for some time but still have not got the hang of it. Since I'm an foreigner I guess it might cause a problem with speech, maybe even when I use headset. It would be nice just to speak out load the conversion with hal. Can some of you tell me how to get a conversion send to hal, like how to get the cr+lf = execute command into the hal script window? I have tried but only got the "new line" printed in the chat window...
Next thing to investigate is how do you get hal looking for asyncronous answers? I mean you will ask him something, which he will be looking for and then he will be delivering answers (sometime later, or regulary) when he eventually found the answer(s)? I got a lot of services waiting for him to do this but I can't have him(the script) activated at <AUTO milliseconds> since that will take the cursor focus away from any application the user is currently working on (for example every 10 second). If I work on a word document, suddenly the hal script window will show up. I'll like him to act whenever an answer is awaiting attention instead, which can be solved by looking into a message queue. A new message would create a notification to the user, that something needs to be looked into.
I guess I can solve it by looking for messages while analyzing an answer from some input from the user. But I would really like him to initiate this on his own.
-"By the way Greg, there is some good news regarding Arsenal awaiting for you"
-Bring it on"
-"It seems Freddie is signed for another four years. Do ÿou want the full article?"
-For sure!
Rgds,
Greg
I have experimented with message queueing, it can be a powerful way to leave messages upon findings, one-time-searches or deliver subscriptions on timed intervals. All based on the user profile.
Any discussions with me are welcome...
looks like you got speech recognition working. I've tried dragon for some time but still have not got the hang of it. Since I'm an foreigner I guess it might cause a problem with speech, maybe even when I use headset. It would be nice just to speak out load the conversion with hal. Can some of you tell me how to get a conversion send to hal, like how to get the cr+lf = execute command into the hal script window? I have tried but only got the "new line" printed in the chat window...
Next thing to investigate is how do you get hal looking for asyncronous answers? I mean you will ask him something, which he will be looking for and then he will be delivering answers (sometime later, or regulary) when he eventually found the answer(s)? I got a lot of services waiting for him to do this but I can't have him(the script) activated at <AUTO milliseconds> since that will take the cursor focus away from any application the user is currently working on (for example every 10 second). If I work on a word document, suddenly the hal script window will show up. I'll like him to act whenever an answer is awaiting attention instead, which can be solved by looking into a message queue. A new message would create a notification to the user, that something needs to be looked into.
I guess I can solve it by looking for messages while analyzing an answer from some input from the user. But I would really like him to initiate this on his own.
-"By the way Greg, there is some good news regarding Arsenal awaiting for you"
-Bring it on"
-"It seems Freddie is signed for another four years. Do ÿou want the full article?"
-For sure!
Rgds,
Greg
I have experimented with message queueing, it can be a powerful way to leave messages upon findings, one-time-searches or deliver subscriptions on timed intervals. All based on the user profile.
Any discussions with me are welcome...