[HPCC-Conf] Coming up for air

[Frank Wales]

Joseph K. Horn wrote:
> Another reason is that many of us enjoy programming PPC's as such.  
 > PPC's can be programmed in simple ways and in goofy locations; nothing
 > else can satisfy the love of "couch hacking" like a PPC.

I have to admit that part of the thing that sucked me into computing was
the immediacy of my HP-41, and the ability to work on problems while sitting
on the bus, or while being bored at a lecture.  It had a wonderful balance of
programmability, immediate problem-solving tools and expandability, and I
really don't see that in modern machines like the Palm, which don't come
with any self-hosted programming tools, and which are overly-dependent on
a PC somewhere. (I say more about this in my 'RCL 20' piece, which is
also available here: http://www.limov.com/rcl20/chapter.lml )

> Richard Schwartz escriven:
>>Anyhow, Palm is truly the wave of the present...
 >
> Keyboard!  Keyboard!  Keyboard!  It's gotta have a KEYBOARD of some REAL kind!
 > Palms (and their ilk) require the user to have a steady hand and be an expert
 > at microscopic accupuncture just to use the so-called "keyboard".
 > A great way to save space, but NOT conducive to FUN long-term use!

Joe, I appreciate your sentiments, but I don't believe a physical keyboard
is *required*, although I will concede that it can improve certain aspects
of the user experience when done right.  I can still operate the RPN calculator
on my PalmOS machine with one hand, using the traditional rest-it-in-my-hand-
and-type-with-my-thumb approach.  If I need to enter a lot of information, I
can open it up, set it on a table, and type on the built-in keyboard.  It's not
quick (and Sony's engineers need to learn about keybounce and tactile feedback),
but it's not that bad either, especially for a first attempt in this form factor.
I used recently it to write an article, and I was also using it as a web
browser via my mobile phone to retrieve information for the article, and as
the to-do list manager for the items to cover in the article, which I'd
noted in odd moments over the preceding weeks.

 > In my oh-so-humble opinion, of course.  It has nothing to do with the fact
 > that I can't afford a Palm.  :-(

Actually, I think it might also be based on some misconceptions about input
methods on Palms.  Graffiti, Palm's pen-based input system (note: *not*
hand-writing recognition), is actually a very good way of entering information
fairly quickly, and the Sony implementation on machines like the NR70 (which is
my current machine) lets me see the trail of my pen on the input area, lets
me switch to an emulated keyboard for unusual characters whose strokes I don't
know too well, lets me hide the area completely to use more of the screen (for
example, for the HP-48 emulator that emulates the 48's whole keyboard and display
exactly as it appears on the real machine, albeit smaller and with no tactile
feedback), and lets me switch between pen-based input and physical keyboard
input as I please without the applications having a clue about where their
data is coming from.

I think we've only begun to scratch the surface (sic) of what high-res
touch-sensitive colour displays make possible in user-interface design
for portable computers, and even though I'll always the quality of the
HP-41 keyboard when using lesser ones, I also appreciate the limitations
of lettered physical keys as opposed to what's possible with on-the-fly
gesture- and touch-based input systems.

When I use the HP-48G emulator on my PDA and an input or option form comes up,
I'm always tempted to touch the option I want on the emulator's 'display',
rather than using the emulated arrow keys and 'OK' softkey to actually make
the choice.  This makes me realise that the direct-manipulation possibilities
of modern machines really do beat the older keyboard-only input model for many
routine tasks, and shouldn't be ignored just because it means that we might
have to lose the comfortable feel of buttons under our fingers as we work.
The abandoned Xpander was one piece of hardware that might have proved to
be a wonderful compromise in this regard, had it actually come to market
with a decent display (the one on the prototypes was terribly hard to read)
and a capable operating system (the prototypes used WindowsCE, which made
them seem sluggish).

I'd rather have excellent new calculators on touch-screens than rehashes
of the old stuff on keyboards, but so far I'm fairly disappointed by all
the PDA-based RPN calculators that I've seen.  They all seem to be stuck in
the rut of migrating designs limited by the technologies of the 1980s onto
modern computers with far more capabilities, presumably because they seem
familiar to fans of the great HP machines, and perhaps because it's easier
than trying to start with a clean sheet. So, we get programs that either
exactly emulate the HP-41 or the HP-48, or that present themselves as
a mismash of the HP-11, 12, 15 and 16C, but that still have four-level
stacks, one- or two-line displays, mnemonic-hell yellow-and-blue shifted
'keyboards', and things like 'P<>S' functions and limited precision, all
because the old 'real' calculators worked this way.

I'd like to see what new things are possible now that full-colour, touch-
sensitive displays, Internet connectivity, millions of bytes of storage and
200MHz processors are becoming commonplace for affordable shirt-pocket devices.
I'm also looking forward to the $30 Linux-based machine with a 25-key keyboard,
a small colour display, some standard I/O capabilities and a self-hosted
programming environment that would be the modern equivalent to the HP-41,
and that could make possible wholly new customised computing devices that
could be the 'personal programmable calculators' of the future.  I hope
I don't have to build the thing myself before I get one (although if someone
on the list will finance it, I'll have the business plan done by Christmas,
and the design by the Spring :-) ).

 > DANG, this is fun!  I can't WAIT to see what those guys at HP are concocting!
 > Think they're gonna resurrect the Xpander?

I fear they're not concocting anything very exciting, at least from our
perspective. You don't get great designs by asking students and consumers
what they want and then building it, nor by creating close functional and
stylistic clones of market-leading products, which is what it looks to me
that HP are doing.  HP are also sub-contracting the design and manufacture of
their new machines to an external company, and so it's *highly* unlikely
that the Xpander will ever re-surface in any form as an HP-badged product.

The new calculator group (remember when they had a division?) also has its
hands tied by not being allowed to make devices that could be seen to
compete in the PDA space, since that's Jornada/iPaq territory (remember when
HP divisions freely made products with overlapping feature sets?). Although
I wish HP well with their new attempt to stay in the calculator business, I
believe they are doing it out of a desire to make money in what they see as
a mature, low-end market segment by making products that appeal to the average
expectations of well-defined customers, rather than out of a desire to
discover what amazing things can be done for ordinary people with
visionary engineering and whiz-bang technology.

Consequently, I do not believe that I will ever again be excited by a new
HP calculator product, because I don't believe that HP see the market they're
in the way they used to, nor the way we want them to.  They see calculators
as a routine, stable market whose products have features that can be assembled
from market research, and sub-contracted for development and manufacture to
companies that also make nose-hair clippers, DVD players and mobile phones.
The HP-65, 41C or 48SX could *never* have been created in this way, but the
HP-10BII, the 6S, the 30S and even the 39G can be.  Expect more of the same,
perhaps with the odd RPN machine thrown out as a bone to those amazingly
stubborn and inscrutible old dogs who keep buying the HP-12C.

Of course, the path from innovation to maturity doesn't have to be relentlessly
downhill or down-market. I am as excited by the engineering of modern-day
Mercedes-Benzes as I am by that of classic HP calculators, and Mercedes-Benz
is operating in a very mature, fashion-conscious and price-sensitive market,
and able to make a good living despite nominally competing with Ford, Toyota,
BMW and others.  I can be just as boring about the virtues and design subtleties
of my ten-year-old Benz as my 20-year-old HP-41.  I really hope that, with some
financial successes under their belt, the calculator group at HP will get the
staff and the remit to try out some new stuff, and will try to grab the reins
of the market from TI, Sharp, Casio and, well, just about everyone else.
But presently, I am not hopeful that the huge company that the calculator
group are now approximately epsilon per cent of will provide long-term support
for that kind of adventuring. I cite the last ten years of HP senior management
decisions affecting calculators as my conclusive evidence for the prosecution;
that most senior HP management seems now to be from Compaq, does not mitigate
against this. (IMHO, HP didn't just acquire Compaq this year, they *became*
Compaq, with all the consequences that that would imply.)

At this point, I have almost no hope left for HP as the place I ought to be
looking for innovation or leadership in portable computing devices (or, quite
frankly, in anything else).  As a perennial optimist, I look forward to being
proved wrong, and being excited once again by a new HP calculator product.
But as a pragmatist, I'm no longer holding my breath.
-- 
Frank Wales [frank@limov.com]