Re: An introduction to MI by example II (part B)...

 >> MI (old and new) is what defines the virtual machine that *is* the AS/400.

You are taking this way too seriously. The AS/400 is brilliant - agreed 100% but
MI does not define the virtual machine.

It was, as I said, an intermediate language output by the compilers that IBM
allowed (or not at different times) people to use directly. I suspect that if C
had been available back on the S/38 that users would never have been allowed to
use MI, but that is speculation on my part.

Even if I were to accept the argument that MI defined the VM - that would _only_
be true of CISC machines with the two level microcode. MI is no more native to
the RISC boxes than COBOL or RPG. On RISC, the closest to "real" MI you can get
is MI-prime, which as I noted before is a derivative of W-code. IBM will make
this available but only for a fee - last time I heard they wanted $1,000,000 !!!

Let me reiterate my point. There is nothing wrong with programming in MI, just
as there is nothing wrong with programming in RPG, COBOL, Java or anything else.
Just don't kid yourself that you are coding to the machine. If this were the
case don't you think that IBM would still be having their compilers generate MI?
The fact that IBM's own tools no longer do so should give you the hint.
+---
| This is the MI Programmers Mailing List!
| To submit a new message, send your mail to MI400@midrange.com.
| To subscribe to this list send email to MI400-SUB@midrange.com.
| To unsubscribe from this list send email to MI400-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: dr2@cssas400.com
+---