|Re: HP's new calc page|
Message #6 Posted by Bill Wiese on 20 Aug 2003, 1:18 p.m.,
in response to message #5 by Fred
"(1) Ye olde quality. When 30+yr old calcs are mechanically outlasting those produced recently..."
>Gee, this never happens with those Hp41, classics,
>etc. (not, I have quite a few that exhibit all these
>issues) The connections in a 41 series and some of
>the Spice are pressure fits, etc.) Some of the good
>old days were good, but others certainly were not!
Yes, some truth there. But quality is also "Feel" (in automotive world it relates to "NVH" - noise, vibration, harshness, even though the engine may be fine and the doorknobs don't fall off). Having good click KBs that have double injection molded keytops that don't wear is a good start.
"True 4-level XYZT+L RPN with large ENTER key above numeric area. Non-RPN or half-assed RPN without the large [ENTER] key doesn't cut it. "
>>I do agree with the ergonomics of the classic
>>ENTER key placement, but the effort to maintain
>>RPN is commendable.
Yes. But KB layout and RPN entry are very tied together.
I am *FAST* number-cruncner on an RPN calc w/large ENTER key - whether it's a 41, 25, or 32S... doubt a small ENTER will help :)
"(3) Accuracy. When the new calcs (12C-Plat) don't offer the same accuracy for some functiosn as the orig 20yr old 12C design does, that's a poor sign. Regression in one area indicates likely problems with others"
>Certainly indicates room for improvement if
>it is in a practical analysis situation, but
>rarely are products without bugs even with the
>best intentions. Lets only deal with actual
>problems and not inferred problems!
The 12CPlat bugs I speak of are clearly *systemic* and *architectural* in design. They could EASILY have written a small 12C emulator that runs in the 6502-style CPU of the 12C-Plat without taking up much more (if any) ROM space. (In fact, such emulation may in fact allow some code compression given that the 'Nut instruction set is well adapted to BCD FP math ops - as opposed to a native rewrite in 6502 assy or C.) This would allow a perfect emulation of a 12C. And additional funcs could be added inside or outside of the emulation layer.
>HP has certain production facilities and capabilities
>they can draw on with their development partners. We
>may not like the reality, but going back to the days
>when calculators were originally developed is impossible.
>Patience will bring some rewards (not all) but what I
>see now occuring will only lead to having nothing at
Development partners?? That implies both do some work. Here it merely looks to me like HP issues a quick spec and has slave-labor Kinpo slightly rejigger an existing calc (33S). No 'value-added' with the HP name.
The ARM-based 49 does seem to be innovative. We'll see how good it is. Now if we could just get that large ENTER key on it we'd be set.
HP has largely turned into a 'rewrap' "badge engineering" house. The brains went away to Agilent (some stayed in the printer division, there still seems to be some innovation there) and they seem to wanna concentrate on building PC clones to sell at Costco - against Dell, who will eat their lunch (and I don't like Dell PCs).
San Jose, CA