I've heard about doing the multipule copy trick, thou ive yet to try my hand at spliceing the code. The reason i try and stick with Eprom's is a few years ago, when i was getting into prom burning, an old tech stressed to me the importance of "If it was designed to take xxx part, then use xxx part" He claimed that with newer, faster flash chips that sometimes there can be read issues analogous to the same concept of why with older MFM drives, you needed to set the interleave.
Of course, he's not here to fully explain why, so i should probabily discount his line of reasoning
Sometimes that line of thinking can lead to really expensive decisions with no apparent advantage, other than being a CYA move.
For example, I once knew of a fellow who paid a premium for 200 nsec DRAM chips when 150 nsec chips were half the price. His reasoning was that he was completely safe in using the slower part. This is the same guy, BTW, who refused to use Hynix parts because the originals were branded Hyundai.
In a sense, this is lazy thinking. A competent service person would sit down and compare datasheets and actually attempt to understand the circuit. Replacing a TMS2716 EPROM, for example, with an MM27C16 is not likely to create any problems.
Saying that "an interleave of
x" is best for a given hard drive and controller betrays a certain amount of ignorance. The more considered statement is that "application
y works best with an interleave of
x". Indeed, there were some people smart enough to partition and format their hard drives accordingly, with different partitions having different interleaves, keyed to the application. If CORETEST says that the interleave for a given drive-controller combination is 3:1, you take that as a starting point. In fact, the behavior of the application may be that data requests are better serviced with a 4:1 interleave.
You could see this kind of thing on some of the better (commercial) CP/M implementations, where the boot tracks were formatted at 1:1 because they were simply streamed into memory; the directory tracks were interleaved at 2:1 because a directory search could function at that interleave with no skipped revs, and the data area was formatted at 3:1 because that was a good compromise.
