The Handheld Calculator Turns 40

Ian Lamont writes "The handheld calculator turns 40 years old this year, and the Smithsonian Institution's National Museum of American History has officially added to its collection examples of the first two programmable calculators, the TI-58 and TI-59. The museum already has the original 1967 'Cal-Tech' prototype, which weighs three pounds. At a ceremony at the Smithsonian yesterday, Jerry Merryman, one of the members of the TI team which developed the calculator, said that the project was started without a set budget and was something that 'we did in our spare time.' Antique calculators have a devoted following; news of a contest celebrating the 35th anniversary of the HP-35 slide rule calculator brought hundreds of fans out of the woodwork to reminisce about the pros and cons of various 70s' era calculators. There are a lot of Web resources devoted to these devices, including the Old Calculators Web Museum, where you can see pictures of everything from the Bohn Contex Model 10 Mechanical Calculator ('apparently the design of the machine caught the attention of the Soviets') to TI's first scientific calculator, the SR-20 ('keyboards were prone to bounce even when new')."

Comments

Two of TI's First, They Mean.

[TaleSpinner]

The first programmable hand-held calculator was Hewlett Packard's HP-65. The SR-52 came a year later. HP then brought out the HP-67, and TI followed a year later with the TI-59. HP then came out with the HP-41 handheld programmable with slots for adding interfaces including HP-IL allowing the calc to handle all kinds of control and data-collections chores in labs. TI followed suit with the TI-88 the following year. I mean the year after that. No, it was the next year. The year after? As a matter of fact, TI never did come out with competition for the HP-41.

But there is no doubt that the first programmable handheld was the HP-65. If they don't have that in their collection then they ain't got the first.

Re:Two of TI's First, They Mean.

[TaleSpinner]

RPN was the only reliable way to be sure both you and the calculator agreed on the order of the computation. For a long time you could not walk up to an algebraic notation calc and expect to be able to use it to produce known-good results. Many of them were only "semi-algebraic" where you would enter 2+2= to get 4 but 30 SIN to get sin(30) - which is RPN. It was a long time before you could do "SIN(30)". Calcs also differed in the number of pending operations they would support, and because of implied priorities these did not match up with the "levels of parens" number. Only with RPN could you know exactly how to structure a problem and feed it in, how to deal with intermediate results, and how to get a reliable result that could be replicated on the same or other RPN models. TI machines weren't even consistent within their own calc line, never mind anyone else's.

RPN also required fewer keystrokes, and the advantage mounted with increasing problem complexity. Also, stack machines were more amenable to programming because the state of the calc could be known exactly, whereas with a TI the state was encoded in the pending ops stack, the paren stack, and then the program area. Jumping into such a mess was an adventure, to say the least.

Re:toys

[Constantine XVI]

The HP50g, TI89, and TI-Nspire CAS all have CAS systems (though the basic form factor is mostly the same as their 48g, 84+ and Nspire kin)

PS: I reccomend the 50g myself. It's definitley a bit on the advanced end (and has RPN as an option, which is what attracted me to HPs in the first place), but shouldn't be a problem for you judging by your homepage URL