Yup -- that's based on ObjectStore -- I used to work at ObjectDesign, and was aware of MANY 3 tier projects with ObjectStore behind them, and X Windows in front -- but that's an object-oriented database, so doesn't count. And I started there in '94, so that's a little late.
I bet IBM would have done something with a relational engine, they were doing a lot of 3-tier stuff. And I bet ODI sales folks from the time would remember what else was out there -- after all, it was the COMPETITION.
Surely the big RDBMS companies could tell you what
their customers were doing at the time.
...and the BIOS on my old MICRON machine rolled over just fine. Whoops -- it wasn't supposed to. The NSTL Y2K test program said I'd have to crank it over by hand, and so did the MICRON homepage. Guess it was the DOS date program that had the bug, not the BIOS at all.
Only someone who has actually done it can tell us how bad a curve it is
Yup, that's me. Yes, there is a learning curve, the "zen" of using voice commands took about six months to arrive (how long did it take you to get automatic on keyboard, though?) And although there is a community of voice coders out there, and therefore some people to borrow from, and voice programming projects in process, the necessary macros and techniques are still pretty much do-it-yourself.
As a computer programmer who has been using speech since 1996 on account of repetitive strain injuries, I don't see speech as *the* user interface of the future, but I see it as part of the user interface of the future. I think it's a mistake to think that one mode will dominate -- I think in future we will see combinations of many modes. Personally, I use a combination of keyboard, stylus (as opposed to mouse), and voice. Voice is ideal for some tasks, such as bringing up tasks ("start calculator", "start emacs") and generating and editing text ("select text-you-want-to-edit...dictate-text-to-replace"). If you can see it, or think it *without strain*, you can say it, and speech recognition will do it. And anything that has a keyboard macro and a label (e.g., buttons, menu items, emacs commands) can be controlled by voice. Indeed, I prefer emacs by voice -- no chords, just "block forward", "block backward", "open file". The zen of voice plus emacs, if you will.
The watershed is, how easy (or painful) is it to navigate to a goal by voice... or by mouse... or by keyboard... or by wiggling your fingers, or by thinking at the computer? You tend to use the easiest interface for the current task. Voice moves more items into easy reach -- out of your penalty box, you might say.
It isn't so great for stupid little icons, or Web links, or applications without standard keyboard interfaces. Tom, I saw your critique of KDE, and the same lack of standard keyboard shortcuts that annoy the basic seasoned Unix hacker also makes it hard on the voice users. It's a problem with XWindows managers at the moment (I run an XWindows server on my Windows box, where I use speech recognition -- Dragon NaturallySpeaking, if you're curious). I hate to say it but Microsoft Windows really wins in this regard -- it has genuine keyboard shortcuts standards, which most applications respect. We need a keyboard standards evangelist on Linux...
As for punctuation in programming languages, a surprising amount of it can be automated, because a lot of punctuation is implicit in the constructs. Your standard for-loop comes supplied with "for(){}", and many editors take advantage of the fact already -- electric mode in cperl, for example. Python manages to dispense with a lot of the curly braces and semicolons -- mainly because they're already icing rather than essentials. Of course, the VoiceCoder email list has discussed this issue, and many others, (navigation, variable names) at length. --Patricia
Slashdot Mirror
Yup -- that's based on ObjectStore -- I used to work at ObjectDesign, and was aware of MANY 3 tier projects with ObjectStore behind them, and X Windows in front -- but that's an object-oriented database, so doesn't count. And I started there in '94, so that's a little late. I bet IBM would have done something with a relational engine, they were doing a lot of 3-tier stuff. And I bet ODI sales folks from the time would remember what else was out there -- after all, it was the COMPETITION. Surely the big RDBMS companies could tell you what their customers were doing at the time.
Moderate the previous up, please!!!
I know what Windows API's look like. I don'wanna see the source!
...and the BIOS on my old MICRON machine rolled over just fine. Whoops -- it wasn't supposed to. The NSTL Y2K test program said I'd have to crank it over by hand, and so did the MICRON homepage. Guess it was the DOS date program that had the bug, not the BIOS at all.
Yup, that's me. Yes, there is a learning curve, the "zen" of using voice commands took about six months to arrive (how long did it take you to get automatic on keyboard, though?) And although there is a community of voice coders out there, and therefore some people to borrow from, and voice programming projects in process, the necessary macros and techniques are still pretty much do-it-yourself.
--Patricia
The watershed is, how easy (or painful) is it to navigate to a goal by voice... or by mouse... or by keyboard... or by wiggling your fingers, or by thinking at the computer? You tend to use the easiest interface for the current task. Voice moves more items into easy reach -- out of your penalty box, you might say.
It isn't so great for stupid little icons, or Web links, or applications without standard keyboard interfaces. Tom, I saw your critique of KDE, and the same lack of standard keyboard shortcuts that annoy the basic seasoned Unix hacker also makes it hard on the voice users. It's a problem with XWindows managers at the moment (I run an XWindows server on my Windows box, where I use speech recognition -- Dragon NaturallySpeaking, if you're curious). I hate to say it but Microsoft Windows really wins in this regard -- it has genuine keyboard shortcuts standards, which most applications respect. We need a keyboard standards evangelist on Linux...
As for punctuation in programming languages, a surprising amount of it can be automated, because a lot of punctuation is implicit in the constructs. Your standard for-loop comes supplied with "for(){}", and many editors take advantage of the fact already -- electric mode in cperl, for example. Python manages to dispense with a lot of the curly braces and semicolons -- mainly because they're already icing rather than essentials. Of course, the VoiceCoder email list has discussed this issue, and many others, (navigation, variable names) at length. --Patricia