Message #22 Posted by Karl Schneider on 23 June 2007, 1:14 a.m.,
in response to message #16 by Egan Ford
Hi, Egan --
Hmm, all without changing the form factor (which I assume to mean size and shape) or the keyboard of the HP-15C, incorporate the following list of improvements:
Add 2 line dot matrix display. User can select between classic single line or new dual line.
The only reasonable way that might be done is with a "single block grid" small-pixel display, a la the high-end Pioneer series, but more legible and less sensitive to viewing angle. Not impossible, but there would be a considerable cost in firmware development, as there would be two ways to display everything. That's why no such thing has been done before...
In program/debug mode display alpha RPN statements. Obvious enhancement if you have ever programmed any Voyager. 12C platinum should have done this (IMHO).
"Debug", I assume, means single-step execution, though it would be nice to display the command with a sustained press of a button, as the HP-41, HP-42S, and HP-32S/SII did. A dot-matrix display would be required, but the HP-12C Platinum doesn't have one.
Increase speed. Again user selectable between classic, 20-year battery life, and full speed 6-month battery life.
I certainly agree -- Saturn-processor speed as a minimum. I don't see a particular need for selectability. Higher speed is desirable for execution of programs, equations, and SOLVE/INTEG, but otherwise causes minor differences in battery life.
Sure, but without a more-advanced programming paradigm (e.g., HP-41/HP-42S), matrix storage, and matrix editing, a large amount of additional memory would be difficult to utilize effectively. For example, would you want to store a large amount of programming, but not be able to organize it into directories or packages demarcated by alphanumeric external labels and "END" statements? Would you want to navigate a huge matrix element-by-element using two keystrokes each time, but not be able to store more than five matrices in a large amount of RAM?
If the modus operandi of the HP-15C were not to be changed at all, about 32 more registers would have been very useful (perhaps one extra R2D2 chip, cost and space permitting). This would have allowed solution of a 9x9 real-valued linear system, or solution of a 4x4 complex-valued linear system without fancy maneuvers (and clearing memory beforehand), or inversion of an 8x8 matrix without clearing memory beforehand. However, the probable reason why 64 allocatable registers were provided was to make the latter two applications possible.
Add I/O. It is so cheap to add I/O. Bluetooth would be great. Industry standard I/O is what got me to switch to the 48GX in 1993. Look at the software selection and support available for the 41CX, 71B, and 48GX. I/O makes a difference. I/O builds communities.
That would certainly require a new list of alphanumeric commands. The 41CX, 71B, and 48GX are all alphanumeric.
42S complex number display (again user selectable).
I, too, like that display, but one should be mindful of the storage of variables in the registers, as well as the capability of the display to show abbreviated complex components having exponents, and complete mantissas as a minimum. Consider how "SHOW" (CLEAR PREFIX) would work.
The devil's in the details...