On Mon, Jul 17, 2000 at 04:54:00PM +1000, Jeff Waugh wrote:
> In my original post, I even went so far as to shrug off voice recognition as
> far-fetched. I don't think it will have a huge impact on user interfaces for
> some time (primarily for the reasons outlined in Conrad's hilarious email).
bah. speech recognition is here already. it'll just be a race to see
whether we adapt to its stilted language or it to ours, first (just
like "graffiti"(sp?) on the palm pilots). and anything you can do with
speech, you can do with a cli ("books" having been doing it for ages).
for some tasks, the lack of tab-completion and "english baggage"
(pronounce "${PATH:-/bin:/sbin}") will make speech interfaces suck.
remember that the unique advantage of computers is that they are the
most flexible tool ever devised. why do we need a computer in
everything when a single computer can do everything?
embedded computers are just a normal computer with an interface more
suited to its task. its the prevalence of "we must stick a gui on it"
that has hampered useability to date. thank god photocopiers still
have a flat top and a big green button (usually).
i'd rather see houses have a single computation/storage "server" and
many, varied "interfaces" (terminals / thin clients / whatever -- the
distinction is fairly vague) embedded around the place. a nice
hot-swappable, high-speed device bus (eg firewire) brought out sockets
in the walls, and the rest is fairly simple (tho expensive at the
moment).
unfortunately this isn't what the vendors want to sell us, since the
profit margins on individual devices would be too low.
(and then again, i think we should all become borgs and embed high
speed communication directly in the back of our skull)
--
- Gus
--
SLUG - Sydney Linux User Group Mailing List - http://slug.org.au/
More Info: http://slug.org.au/lists/listinfo/slug
