On Tue, Apr 25, 2000 at 04:25:53PM -0700, Steve Sabram wrote:
> Tom Zerucha wrote:
>
> > But to clarify what I posted before, a true MMU would be overkill,
> > though a restartable bus error and other CPU32 or 68010 core features
> > would be desirable and give a lot of bang for the amount of the
> > change. (Or an onboard dma copy unit, again for things like DmWrite -
> > to do ultrafast memcpy, memset, and memcmp calls and their str
> > versions).
> >
> I can see our difference here. My mindset is that the Palm has
> always been an embedded system with a realtime OS instead of a
> shrunken computer. Taking all of these things into consideration
> can slow down development cycles. Piling all of this on can kill an
> OS. Lots of us here have our battle scars from the overburdened
> Newton. I mean the thing even had its own inference engine built in
> and it died a well deserved death.
Actually I have the same mindset - most of what I do is embedded.
CPU32 was designed - optimized - for embedded applications.
If a DmWriteCheck would return an error instead of displaying
FatalError, it would help, and to maybe argue for the fuller MMU,
having a MemPtrWriteEnable type call wouldn't be outside the paradigm.
That and having more throughput wouldn't be a bad thing. An optimized
CPU core gets more done with fewer cycles, so you don't have to speed
up the clock rate.
And my point about sound is merely that it is as lame as the Pilot
generation of screens. A simple ad-lib type polyphonic sound
generator - even to bring it up to a commodore 64 or atari st sound
capability would be an improvement.
Arguing for wavetable synthesis might be like arguing for full text
font and format rendering, which is too much, but there are multiple
fonts available on the Palm simultaneously. I would like multiple
pitches.
> A good Palm app does "one thing" and that is it. If you need to do
> something else, you can turn on your Walkman or use you cell phone.
> Personally, I don't believe in the "consumer convergence" that is
> being hyped like crazy right now. Having that much value in a
> single handheld device that can be lost or stolen is too much of a
> risk the general buying public is willing to do. PDAs are carving a
> big nitch of its own. Most of all, there is not monthly service fee
> to use my Palm device.
You obviously don't have a Palm VII (at least not activated :).
Your point though is well taken. THE VALUE OF THE PALM PLATFORM IS
THE APPLICATIONS. Look at CE - there are few developers, but there
are apps, but nothing exciting and they aren't adapted to the PDA
platform. They shouldn't have called the new platform "Pocket PC" - a
more accurate name would be PeepHole 2000 since you are doing the same
things through a more narrow apeture. Palm apps use the limited
screen efficiently so the apps are actually more effective.
The apps do one thing, but I wouldn't limit what that one thing is.
Put another way, each app does just one thing, but there are thousands
of "things" done.
Palm will never do MP3s better than a stand-alone device (nor will CE
devices), so I don't think that will be a critical app. I would argue
the same thing for full voice recording. But that doesn't mean there
aren't uses for a better Audio subsystem.
The converse is that consumers won't want to buy two things that have
nearly identical hardware and take up "dilbert belt" space, but the
10% that doesn't overlap requires two devices. Consider Motorola's
Alphanumeric 2 way pager (lcd screen, keyboard, uses the dragonball),
v.s. a Palm VII. The former can't even do clipping, but the later
doesn't have notification.
If OmniSky gets the message light fixed, a lot of good things can
happen. Not push media per se, but notification of an external event.
You already have the text and other capabilities built into the palm,
and the wireless in the Palm VII or OmniSky.
--
For information on using the Palm Developer Forums, or to unsubscribe, please see
http://www.palm.com/devzone/mailinglists.html
