Re: [NTLK] Palm LifeDrive

oh, boy,i didnt wish for this to happen...
I was just looking around for a pocket device that could do some multi
media.
I have given up on my search for now.
I have determined that its honestly not a top priority to replace the newt
any time soon.

I used it yesterday to write down a contact info. darn thing is for me
that the jaggies came up right when i was writing the last name and it made
things difficult. oh well.
I dont want to start a whole slew of comments on all this, i was just
getting opinions of anyone's experience with some of the newer pda's out
there.

Kevin

Original Message:
-----------------
From: Steven Scotten splicer@paroxysm.com
Date: Sun, 15 Apr 2007 21:43:50 -0700
To: newtontalk@newtontalk.net
Subject: Re: [NTLK] Palm LifeDrive

On Apr 15, 2007, at 5:27 PM, Lord Groundhog wrote:

> Kevin, with respect, there's a reason why some of us "spit fire and
> blood"
> about mere character recognition: for us, it's a retrograde step
> from the
> *handwriting* recognition of the Newton.

First, it was Steve (me) who wrote that, not Kevin.

Second, yes, I get it. I get that handwriting recognition allows for
input the way we've done things all along. It's intuitive and has
very little learning curve. It only requires a certain amount of care
to make certain that our words are well-formed enough for the HWR
engine to decipher. There's nothing new to learn, or at least very
little.

I get that handwriting recognition is the computer doing it our way,
it's how things are supposed to be. It has the romance and personal
attachment of it being our very own handwriting, as unique as a
fingerprint.

I get that handwriting is an art and deserves respect that typing or
semaphores or single-character recognition systems don't.

I get that real handwriting recognition is human and not mechanical.

> When I was a very little boy, I
> was taught to learn new words by sounding them out, letter by
> letter. That
> went along with learning to print, letter by letter. Before long
> though, I
> was expected to go beyond that, to recognize whole words at a time,
> and
> along with that I was taught to write whole words, using a cursive
> script.
> In my school it was the Palmer script. It was attractive, legible
> and most
> of all, faster than printing. Even though my handwriting has
> seriously
> degenerated over the years, it's still faster than printing. I've
> timed
> myself both with cursive script and with printing; it does NOT
> "double or
> triple" my speed to print even with the alphabet I already know,
> let alone
> some made-up set of symbols. And that's where the Newton comes in.

Seven years ago I made these very same arguments with great passion
and disgust for any fool who was stupid enough to get a Palm device
or do something blasphemous like install Graffiti on their Newt.

So imagine how I felt when I smugly challenged a co-worker to see how
fast we could write a sentence out, him on his lowly stupid pathetic
Palm and me on my beautiful kick-ass MP2100. The sentence was "The
time has come for all good men to come to the aid of their country."
I got to "all" and the idiot using the stupid single-character system
was finished. It wasn't even close. So who was the idiot really?
Well, as in most situations in life, the idiot was me.

Your analogy, cursive versus printing, has very little relation to
this. I'll admit that I printed when I was in my speed test, so
perhaps I could have been faster if newt would recognize my cursive.
It will not, and I've tried to get it to for over a decade now. So I
print on Newt.

I've just tried it again, and the fastest I can get the sentence out
on Newt is about 28 seconds. That's without corrections. I no longer
have a PalmOS device with a graffiti area so I tried again with
graffiti for Newton. It came in only a little faster---23 seconds.
But that time included all the corrections. Trying again with the
real HWR I slowed myself down so that I would make fewer errors and
made a concerted effort to make careful letter shapes so that Newt
would get it right. I came in at 36 seconds with only four
corrections to make. By the time I was done fixing the errors in the
sentence it was close to 75 seconds.

Certainly everyone else on the planet has more success with Newt's
HWR than me though. Four typos in one seventy-letter sentence is a
high rate of errors for a Newt user, no? But these were all errors
that I would say "don't count" if I were showing off Newt.... a
letter miscapitalized, an 'n' instead of an 'm', things like that. I
figure that's par for the course. I've never gotten it much better
than that.

A couple of things come into play here: first, with character-at-a-
time recognition, editing mistakes is fast and on the fly. It's a
backspace and do-over on the character that didn't get in right the
first time. With Newt I have to reposition the carat and go in
postmortem to edit because I don't even see the words appear until
I'm on the next (or the second or third next). The fastest thing I've
found is to select the whole word and try rewriting the whole word.
Otherwise I get spaces inserted that I have to delete.

The other factor is that (the original) Graffiti is a *simplified*
alphabet. This is why Xerox sued Palm. Graffiti2 requires more
gestures; original graffiti has a one-gesture shape for every letter,
which is what Xerox's patent is for. Anyhow, there's no stopping to
cross T's or dot I's or create the crossbars in A's or the vertical
stroke of a B.

> BTW, in this context comparing HWR to typing for speed is spurious
> even for
> people who type faster than they write. I can stand in a queue at
> the bank
> or post office, Newt in one hand and stylus in the other, and make
> notes
> without any need of a flat surface or anywhere to sit in order to
> type.

When did I compare HWR to typing? I'm only arguing that single-
character recognition, once learned, is faster.

I fully concede every other point: it's a dirty hack, it's not real
or pure handwriting, it's not intuitive and it's not comfortable for
someone new to it. It's not warm or fuzzy and it doesn't impress
anyone as wonderous magic.

The closer comparison between Newt's HWR and Graffiti's SCR would be
cursive writing to shorthand. Shorthand requires some learning and
it's certainly not real handwriting, but it is a hell of a lot faster
if you take the time to do it.

> ...just to reassure you we aren't simply irrational or crazy. :)

This isn't a you versus me thing. I am still crazy in love with my
Newt. But try as I might, and I have for years, I can't use HWR for
any extended writing, while I *can* pound out a few thousand words
with SCR sitting in a cafe.

By the way, this is still advantage to Newt. PalmOS devices sold
today either have no SCR or have CIC's Jot sold as Graffiti2, which
is "superior" to Graffiti only in that it's more like the Roman
alphabet. Which means crossing T's and dotting I's. The newest PalmOS
devices crash hard when putting the old Graffiti back on. So Graffiti
on a MessagePad is, in my opinion, the killer combination.

Steve

-- 
This is the NewtonTalk list - http://www.newtontalk.net/ for all inquiries
Official Newton FAQ: http://www.chuma.org/newton/faq/
WikiWikiNewt for all kinds of articles: http://tools.unna.org/wikiwikinewt/
--------------------------------------------------------------------
myhosting.com - Premium Microsoft® Windows® and Linux web and application
hosting - http://link.myhosting.com/myhosting
-- 
This is the NewtonTalk list - http://www.newtontalk.net/ for all inquiries
Official Newton FAQ: http://www.chuma.org/newton/faq/
WikiWikiNewt for all kinds of articles: http://tools.unna.org/wikiwikinewt/