Mxsmanic said:
I'm not so sure, as it's extremely inefficient for certain
purposes.
Sorry it doesn't work for you.
That's true even of GUIs, actually.
I find the graphical user interface to be very efficient.
Speech-to-text does much of the rest, and does almost all of my
typing.
You can dictate C code faster and more accurately than you can
type it?
If there were much difference between slower and faster when I
dictate C++ code, dictating it like I dictate text would be faster
and more accurate, Yes. The only reason dictating code, including
command line stuff, might be slower is because the speech
recognition does not include orientation/optimization toward that
end. In fact, accuracy is another major benefit of speech-to-text.
You will find an occasional typo and maybe a grammar error, but
you will find zero spelling errors in my posts to Usenet.
Macroing can be voice controlled. Currently I use speech for
entering function block outlines. Speech is also used for any
ordinary words. The words wParam and lParam are by default
included in this vocabulary as I just spoke them. Few others are,
but they can be included. Any words, including capitalization, can
be simply added to the vocabulary.
Much is a matter of building a vocabulary.
Currently, the example that David Maynard gave might be difficult
to voice one term at a time, but it is remarkably easy to
voice as a macro.
As given, this is the linux command line to plot FastE packet rate
vs packet size.
echo 'pad=20; plot [64:1518] (100*10**6)/((pad+x)*8)' | gnuplot
-persist
That can be entered instantly/perfectly by simply saying "plot
FastE packet rate vs packet size" or just "packet rate versus
packet size".
echo 'pad=20; plot [64:1518] (100*10**6)/((pad+x)*8)' | gnuplot
-persist
That was just written instantly when I said "packet rate versus
packet size"
Macros are common, but speech makes them even easier to use and
the potential scope immense.
Actually, I can voice the first part of that command line quickly.
Echo space' pad equals 20; plot [64: 1518] (100*10**...
.... that was a first effort, light-years from the potential
.... that was spoken with an ordinary speech-to-text program
without any optimization
.... with optimization for command line symbols, spacing, and
punctuation, that will be faster and more accurate than by hand
I realize you are trolling both sides of this issue. But that's
the way it is. Maybe someone else will get something out of it.
By the way. Approached speech recognition with caution. The
potential for an ordinary PC user is much greater than using
Linux, but the frustration can be as intense.
Have fun.