[MV] DNS on Mac on Parallels

[Gavin Wynford-Jones]

Chuck,

Your assertion is ridiculous and you ought to know it. Dragon has had  
the ability to successfully intermingle commands and dictation from  
the beginning. It does NOT cause problems since the commands are  
generally carefully defined and, in 99% of cases, consist of more  
than one word. The example you give would not result in the computer  
trying to go to sleep as there would not be a pause before the  
"command" and so it would be recognised as a continuation of the  
dictation.

I recall having long discussions about this sort of thing when we  
were trying to get a French version going. French requires good  
context sensitivity, which iListen lacks. I don't know whether that  
is a limitation of the engine or if it comes from a programming  
decision you guys made early on, but it is at the root of these  
problems. With good context recognition, commands can be integrated  
into dictation without problem.

Gavin

On 7 May 2007, at 16:21, Chuck Rogers wrote:

> Ian (and everyone else):
>
> As Joe pointed out, having commands intermingled in dictation mode  
> could cause huge problems. While there will be improvements, the  
> only way I see commands being integrated is if there is a trigger  
> word in front of the command. Even that could cause problems,  
> however. Imaging you are dictating the phrase "when I told my  
> computer to go to sleep" and the trigger word is "computer." You  
> wouldn't want your computer to go to sleep in the middle of a  
> dictation session!
>
> There will always be something to "get used to." If you have  
> commands integrated into dictation you will be even more restricted  
> as to what you can say for commands, since those commands will have  
> to be structured in such a way that they are less likely to be  
> mistaken for a phrase you want to dictate. This can present a very  
> unnatural way of using the computer, since those phrases would  
> likely be harder to remember.
>
> As it is, you can always say "one shot command" in iListen and it  
> will go into Command Mode for one command - so the very next thing  
> you say after "one shot command" will be interpreted as an action  
> instead of typed out. This may give you what you are looking for -  
> or at least get you closer.
>
>
>
>
> Best Regards,
>
> Chuck Rogers, Chief Evangelist
> MacSpeech, Inc.
>
>
>
>
> On May 6, 2007, at 6:12 PM, Ian Gilman wrote:
>
>> I understand there are technical challenges in creating a truly  
>> modeless speech interface, but we've got to get there somehow.   
>> Speech has the potential to be a much more natural interface for  
>> the computer than mice and keyboards, but always having to  
>> remember which mode you're in is a big stumbling block.  You may  
>> be used to it now, after you've been using iListen for a while,  
>> but that's not how most people talk.
>
> _______________________________________________
> MacVoice mailing list
> MacVoice at listserver.themacintoshguy.com
> http://listserver.themacintoshguy.com/mailman/listinfo/macvoice
>
> Listmom is trying to clean out his closets! Vintage Mac and random  
> stuff:
>         http://search.ebay.com/_W0QQsassZmacguy1984
>