Negatives of HP (**edited) Message #2 Posted by bill platt on 16 Nov 2004, 11:27 a.m., in response to message #1 by Gene
Hi Gene,
15 minutes! It only took me 1 minute of confusion!
Seriously though, I think the whole handling of "minus" is the weakest, most poorly implemented part of the whole proud history of hp calculators.
That is, it seems that as long as HP was strictly RPN, it was nice. But all the algebraics, with exception of the RPL family, have a consternation factor near "10"
For instance:
32sii has a buga "unary minus" takes precedence over exponentiation in an equationexcept that it does not function as such if it is the first character in the equation list. (In fact, it does not even show up as a "unary minus" in this positionit looks like a minus sign, so maybe it isn't really a bug!) Further, the CHS key will type either a unary minus, or a minus, depending on the contextvery confusing!
33s: Improved on the 32sii by eliminating the precedence of "unary minus", yet, they retained the unary minus (a high placed minus) in the character setconfusing!
27s and 19b in my opinion are betterthere is no unary minusand in the equation list, it is a consistent systemif you want  (x^2) you write x^2, but if you want (x)^2 you write (x)^2. In the 27s, the CHS key is simply thata Change Sign key.
30s: the CHS key is a unary minus keyexcept it is infix, never postfix. However, this machine does not use the unary minus any more than the 33sthere is no good reason for itexcept that they require it for the negative exponent. At least the 30s will give you a "syn error" if you do it the "wrong" way.
My absolute favorite machine for simplicity is the 48gx and its family. Here, there is no "unary minus" and the CHS key is merely thata postfix change sign key. You can use the minus sign anywhere you need a minus signand you can change the sign of the last + or  in the string by using the CHS key. Even better, the 48g will parse the command line to the stack, and in so doing, simplify extra minuses.
Intersting to note that the 27s/19b do not parse them out but allow extra minuses, e.g. 677 = 60. THe 30s does not allow extrasit returns a syntax error. The 32sii will allow double minuses, but not triples, however, you can get a double minus, followed by a "unary minus" to parse in the 32sii.
Note that where I say CHS above, I am referring to the key that says "+/" Interesting that the old voyagers all used CHSit also wold have been a better choice for the 48.
Perhaps the most surprising thing in a good way is the 33s ALGEbraic mode. When using the calculator line, you do all "one number functions" postfix (like RPN) and you add, subtract etc infix. Yet, the upper line display is parsed to proper algebraic notation, e.g. 30 sin parses to sin(30). Even better, the potential confusion involving the "unary minus", and also negative exponents, is resolved as well. If you do this on the 33s in ALG:
4 + 3 CHS ^ 2, the machine will parse it to:
4 + (3)^2 so you can see what you are asctually doing.
**{Someone with a 33s, please confirm thisI don't have it at the office and I don't want to mislead).
Best regards,
Bill
Edited: 16 Nov 2004, 11:40 a.m.
