Here's another calculator with Aplets. I wasn't exactly enthusiastic about this concept when I first saw it with the HP-38G, and I am still not enthusiastic today. But if it helps Hewlett Packard reestablish its market leader position (and produce some decent engineering calculators afterwards), I am all for it!

I guess that calculators like the HP-39G and its all too many counterparts from Casio, Sharp, and Texas Instruments all prove an interesting point: easy to use is not the same as easy to learn. Many people complain that RPN (not to mention RPL) are fiendishly difficult to learn; in fact, I've heard numerous stories of HP calculators not being stolen because they're RPN and no-one knows what to do them.

Formula entry calculators represent the other end of the spectrum. They're easy to learn alright; in this respect, they could only be improved, I guess, if they came with an optical wand with which to scan in equations from paper. As for ease of use, however, well, why don't you judge it for yourself? The following two keystroke sequences accomplish the same thing: taking the square root of the most recently displayed number (i.e., the result of a previous calculation):

HP-39G		Shift		√		Shift		Ans		=
HP-48G		√

If this is what they call an improvement, no wonder engineers are less than enthusiastic, and are willing to pay a king's ransom for a decade old, battered HP-41CX or HP-42S!

Functional integration in the HP-39G is as good as it is in the HP-38G; for instance, support for complex numbers is complete, or nearly so. One exception is the built-in generalized factorial function, which works only for real arguments.

As is the case with the HP-38G, the programming model of the HP-39G is somewhat inflexible. A "program" is something that can only be invoked from the command line (or from within another program) via the RUN instruction. Sounds good, except that it makes it impossible to create programs that can be incorporated into user-defined expressions. It is, however, possible to accomplish this using function expressions in the Function Aplet. So here is an example, yet another variant of my Gamma function example, one that works for complex arguments as well, so as to supplement the built-in generalized factorial. (Actually, this expression computes the logarithm of the Gamma function, thereby also extending the function's range.)

F1(X)=IFTE(RE(X)<5,F1(X+2)-LN(X*(X+1)),X*LN(X)-X+LN(2*π/X)/2+
      ((((1/99/X²-1/140)/X²+1/105)/X²-1/30)/X²+1)/X/12)