Embedded Systems Conference, Boston, Mass.—Teardowns of current products are insightful, but teardowns of older products that failed can provide broader perspective. Sometimes it's the concept, sometimes it's the execution, sometimes it's the market, sometimes it's a combination of factors, and sometimes, well, who knows?

At the Embedded Systems Conference, Allan Yogasingam and Steve Bitten, technical analysts for EETimes and TechInsights, did live teardowns on the Apple Newton MessagePad and Bandai-Apple Game Console of the 1990s. Both products combined leading-edge ICs with older ones (some of which are still made!), and their failures in the market have lessons that still apply--and some that don't.

The Newton, developed in 1992 and released in 1993 was the first personal digital assistant (PDA), and combined calendar, notepad, a 9600 baud fax/modem, and similar basic features with notepad-like handwriting recognition capability. By our standards, it was quite a handful, approximately twice the size of a modern PDA.

Inside the unit was an Am610 processor, a cutting-edge 32-bit RISC unit. It was supported by 4 MB of AMD ROM, Epson RAM, an Intel 8 MB flashfile memory (used in lieu of additional static RAM), and a 12-bit analog/digital converter from Analog Devices. The back of the single PCB was mostly analog circuitry, while the front was primarily digital. An RS-422 line driver and a dc/dc switching regular, both from Linear Technology Corp., rounded out the bill of materials.

But a product is more than its BOM. Although there are Newtons still in use, and even user groups for it, the PDA was a failure, most likely for these reasons, according to Yogasingam:

• Apple tried to re-invent computing, by saying the Newton would replace the PC (or Mac) rather than position it as an adjunct device
• Even as an adjunct of the Mac, it had synchronization problems
• The high price of $700 to $1200 was a real obstacle
• The large and bulky form factor, due to technology at the time, contradicted the PDA product message
• And most critically, the handwriting recognition simply did not work well. Yet this feature was the focus of the marketing and pushed as a key feature.

Development of the Newton took between three and four years, and Apple suffered a serious "black eye" in the market and red ink as a result of it. But there were some positive outcomes: its operating system was sold to Palm, and became the basis of the Palm Pilot OS, and it put end-users on the road to non-QWERTY portable devices as a product line. Since it consumed so much of Apple's R&D resources, set such high (and unrealistic expectations), and fell so flat, it also cost John Scully his job as Apple CEO.