Posted on 26 August 1998

The following text was written by S.T.L.:

Let's see, it's late at night (or really early in the morning), and I'm reading Wired (a good magazine). September 98, pg 212, if anyone's interested. Nicholas Negroponte writes:

"But the potential for very low cost computers is wildly more than we have made of it. Why? Because inexpensive computing is a crummy business. [description of low profit margins, etc] US companies just do not know how to tackle the low end. And by 'low end' I don't mean the much vaunted sub-$1000 computer - I mean PCs that cost less than $100."

I argue here that Intel, and the other big chipmakers, are good at the very high end (can't wait for the Merced to arrive), but their attempts at "Network PCs" and (for example) the Intel Celeron, are laughable. Thus we shouldn't expect very low end computers to come from them. I say that the "PCs that cost less than $100" that Negroponte speaks of are ALREADY HERE, in the form of TI calculators. Just as with PCs that range from $2000 to $5000, one can pay less than $100 for a low-end TI (like an 81), pay $100 for a TI-85 (good general purpose calc) or splurge and get a $200 TI-92. Or $275 TI-92+.

Calculators used to be very simple. Add 2 and 2, please. They were electronic abaci. Nowadays, I would say that the TI calculators aren't abaci anymore, they are small computers in their own right. This isn't like saying a car is a computer, because it contains microprocessors. TIs (my experience is with the 85 and the 92) can do most anything an Apple computer can do (the old ones), and sometimes faster. PDAs will never be successful until they can act as a universal computer. I also would say that TIs and their decendants may replace PDAs as the small electronic devices of choice. I take notes on my TI-85 when I don't want to remember something. And small videogame systems, like the old 8-bit Game Boy. For example:
In the "olden days", someone could get an Apple to do math for them. TIs do this much better, and they fit in your pocket. (Well, large pockets.) In the "olden days", someone could take notes on a PDA. TIs do this, and more. In the "olden days", someone could play games on a Game Boy. TIs have a wealth of (mostly free) games.
And like a PC, they can connect to a large network of new programs (via a PC), and communicate directly with one other TI. (Like a modem-modem connection.) And remember - this was all done WITHOUT Texas Instruments really intending this to happen. The first assembly shells exploited loopholes that let them get to the core of the processor. Texas Instruments saw how people loved the computing capabilities this gave them, and then put assembly support into the 86 and others. The low-end future of computing is probably within TI's reach, if it can keep costs under (say) $300 max, and be even MORE versatile than the current TIs. If anything, the history of computing shows that *open architecture is the road to success*. IBM let everyone clone their computers and make OSes for it. Apple was hissy and only allowed itself to build their computers and write OSes. (Allowing cloning a decade later didn't help any). Result: you're probably reading this on a system that is IBM-compatible. Even more so: the current success of the "Wintel" systems is because Microsoft made a shell for MS-DOS that made it friendly, and powerful.

Texas Instruments probably planned to just make a calculator when it made (say) the TI-85. Yet because it left a loophole, people made shells, and OSes for it, and programs. And they did all that for NO pay at all. Just people hacking up programs in their free time. TI didn't even make it convenient to exploit that loophole, and yet the various assembler shells have large bases of support. (How large, I don't know). As processor costs go down, Texas Instruments will be able to either reduce the price of their calculators, or add new features. Weasels have developed memory expanders and figured out ways to get TI-85s to make sounds. And that's just with one link port. Imagine a TI-X that had the following features:

• 1 MB user RAM, 3 MB archive RAM
• 20 MHz processor
• Headphones port
• Calc/PC Link port
• Calc/Calc Link port
• Memory Expander port

My opinion is, that Texas Instruments should also finally ACKNOWLEDGE that TI calculators can do other things than calculate, and in future TIs make it easier for people to add features to TIs. Look what's happened so far. What's your opinion?