On Wed, 11 Apr 2001 09:16:04 -0700, Rennie Allen <RAllen@csical.com>
wrote:
Lack of OS Support? If your ready to shell out money to write your
own handwriting recognition software, writing your own input stuff to
bypass the OS shouldn’t be an issue.
I didn’t suggest that one could not purchase the recognition technology.
Bypassing the OS can be a very big issue depending on the architecture
of the OS (in QNX it would be a breeze - but wait you don’t have to
anyway).
My understanding of PalmOs is that handwriting recognition is build in
(if not the hanwriting example should be replace by something else…)
What Palm has is not handwriting recognition; it is simple gesture
recognition (and it is completely useless - anybody who uses that on a
Palm is a geek, plain and simple - PDA’s are not going to become
ubiquitous with gesture recognition).
Well given the success of the Palm, it appears to me this gesture
recognition is rather popular. Most people I’ve seen using it
seem to enjoy it (the recognition). Personnaly I don’t own a PDA,
so I don’t really know what I’m talking about 
But this
recognition think is not important for that discussion. We could
be talking about support for DOC which came really late for NTO,
support from broadcom. (I could tell you more about the later…)
If you build a Palm like device (no real-time requirement) and you
need handwritting recognition, what OS are you going to choose.
PalmOS with the buildin stuff or QNX6 with the superior architecture.
I would choose PalmOS.
Again PalmOs does not have handwriting recognition. Your choice is
QNX6 with no handwriting recognition or PalmOs with no handwriting
recognition, or WinCE with bad handwriting recognition that crashes
every 15 minutes.
It did mention: “If the hardwriting is not in the palm replace it
by something else” so replace it by ". That gives QNX6 with no
gesture recognition versus Palm with gesture recognition.
As per the old MI series “your mission, should you choose to accept it,
is to provide handwriting recognition … Jim, you can have QNX6 as one
of your team members, or you can have PalmOs, your choice … this tape
will self destruct in 5 seconds”.
I would pick QNX6 that’s for sure, but you change the question
Handwritting on QNX6 may be better but today it aint there. That’s
It ain’t there on PalmOs either, and it ain’t really there on WinCE
either (try using it).
the case for LOTS of features nowadays. From my observation QNX6
definitely lack behind most other OS in this area and is not
improving. . In my day to day work this is getting to be more and
more of a problem for my customer. Sure QNX6 is comming out
with new stuff, USB for example. but I don’t feel that’s enough,
the gap isn’t closing.
It seems like your argument is that QNX6 should provide basically
everything you or your customer wants. QNX6 is not an end-user OS, it
is aimed at companies with the where-with-all to add real value.
No I’m not, I have no problem writting driver for devices ( that’s
what most of my work is). Most of the time QNX6 provides the
tools to do that.
My concern is with what are “mainstream” features, perhaps
i’m streching the target marget of QNX6 beyond what QSSL
really intend to do.
I know QNX6 is not an end-user OS (and I’m not the one
who’ll be pushing for that). But as a usable OS there are
certain thing I would expect that I find missing.
In conclusion my point is (no more not less), architecture isn’t
everything.
Well, that’s a difficult point to argue if taken literally, however, the
implied North American cultural semantics of “isn’t everything” are
“isn’t anything”, and that’s where the “implication” comes from. IMO
good architecture gets you 90% of the way there (which in North American
semantics equates to “architecture is everything”).
I’ll take your work about the semantic
But I’ve seen people through away LOTS of work (and money) because
they couldn’t deal with that 10%.
.