Fast Eproms CY7C291 Data?

I was an engineer at Data I/O from 1981 to 1997 and made numerous visits to semiconductor companies. The semiconductor companies had a good business reason for limiting access to the programming algorithms. The customers expected the programmed PROMs to work over the entire temperature and voltage range for the life of their products. They also expected a minimal number of chips that failed programming. To achieve this, the semiconductor companies would work with leading PROM programmer companies so they could certify the programmer met there specs. If a customer programmed part with one of these devices, the parts should meet all the specs and the semiconductor would stand behind them. Data I/O would publish a "Wall Chart" listing the thousands of different parts their programmers would handle and what revision of software and hardware was needed. Sometimes the programming algorithm would change after the PROM was released. Some process change would require a longer pulse width or a voltage change. I remember one time National Semiconductor had several production lots of bipolar PROMs that had a poor yield with the current algorithm. National asked the leading programmer companies to rush an update to customers. National could do this with 4 or 5 programmer companies, it couldn't happen with an update to a datasheet and expect every programmer to be updated. Michael Holley