Dragon NaturallySpeaking sadness

[deborah]

I love Dragon NaturallySpeaking. I couldn't be employed or a member of online communities without it. But that doesn't mean there aren't things about it that make me gnash my teeth. Here are my top three quality-of-life-affecting bugs which have been introduced in the past couple of versions:

1. The newly introduced inability to use "spell" commands in the flow of normal dictation. It used to be that an utterance could be "And the characters are named Bob, Lily, spell echo india lima oscar november whiskey yankee, and Ahmed." Now there has to be a pause-for-command on either side of the spelling command: "And the characters are named Bob, Lily, [BREAK] spell spacebar echo india lima oscar november whiskey yankee [BREAK], and Ahmed." Unless you dictate all the time you might not realize how incredibly disruptive it is to thought and flow to have to break like that. In my daily life there are any number of words I have to dictate with spelling commands which are just not worth adding to the dictionary, because I might dictate them three times total. The way it used to work was perfect. (Version 11 change, I think. It might have been version 10.)


2. The brand-new feature of the sleeping microphone to hear random background noise as a "wake-up" command. It's become comical how frequently I'm speaking with a friend or coworker and I suddenly have to indicate silence while I say "go to sleep. Go to sleep. Go to sleep," while meanwhile my computer processes whatever utterances at has heard in the interim, often to exceedingly disruptive effect. If you use a computer hands-free, using the "microphone off" command instead of the "go to sleep" command is pretty much impossible, at least without some kind of foot switch. I really hoped this bug was going to be fixed in the first NaturallySpeaking 11 patch, but instead they just introduced new commands for direct tweeting and Facebooking. (Version 11 change)


3. The dropped letters/doubled letters in non-Select and Say windows. This didn't used to happen. If you complain about it to Nuance, they just say they don't support non-Select and Say windows. It's really bad, and happens non-deterministically. Sometimes the first character of a dual-key sequence gets dropped, so, for example, "select all" turns into "a" (since "select all" is "control-a"). (Version 10, I think)

Aside from the bugs, here are my top Dragon feature requests:

1. A couple of versions ago, Dragon introduced the ability to operate (e.g. Select, bold, delete) on a word which is repeated multiple times in the window. You say something such as "bold Dragon", and a number appears next to all of the incidences of the word "Dragon". Then all the dictator needs to say is "choose 2" to choose the second incidence of the word "Dragon." This is fabulous. Now, Nuance, could you add this feature that when I say "choose 2" Dragon hears me uttering the command "choose 2", and not the dictated text "choose 2"? Four times out of five, Dragon thinks I am saying anything other than "please select the item with the big red 2 next to it". (I'm calling this a new feature, not a bug to fix, because it never really worked right.)


2. I would love a revamp of the advanced command editor and the Command Browser in general. It is, ironically, fairly difficult to navigate all of the features of the Command Browser via voice . It's also difficult to do any kind of nuanced sharing of commands between two different computers. And while revamping the Command Browser/advanced command editor interface, is there any way the documentation could get a revamp as well? Those "advanced command editing with Dragon version #" books are pretty basic, and don't get into the rich features in any depth.


3. Could there please be the ability to press the Windows key in an advanced command? The Windows key is not the same as CTRL-ESC, although both of them pull up the Windows start menu. It is a single key which sends its own key binding, and multitude of programs require the Windows key in particular.


4. Could you consider working with vendors who make other accessibility tools (e.g. JAWS, NVDA, Qwitter) for better interoperability? Right now, it's impossible to use any of those tools entirely hands-free with NaturallySpeaking, unless you buy the $900 J-Say package to interface with JAWS.