The Soul of Barb's New Machine 1149

That's an aspect of a successful evolution of the technology. The problem, as I see it, is that IBM didn't force old code to "convert" to new the way we did. For instance, with the obsoletion of KA floating point instructions (they just simply didn't exist), we did NOT put the emulation in the KL-microcode. That would have kept all programs using these instruction unmodified. What JMF did was write an emulation in a monitor module and blasted each use of the instructions with a warning to the job. If a customer didn't want to change his code, he could build his monitor with KASER but it came at the price of "unsupported" once we put the module on the Customer Supported Tape.

hmmm...I'm trying to recall if TOPS-10 has any "original" code. I suspect the "oldest" as a monitor module would have to be DECtape...but that module was renamed, too, from DTASRC to DTASER. I don't know why it was renamed. Renaming usually meant a rewrite, but I know TW didn't do a rewrite.

Another module that didn't get rewritten was SCNSER; even though everybody dearly wanted it to be rewritten, nobody wanted, or rather, dared to do the job.

Oh, and the code that was dumped to PDP-8s and PDP-11s for comm frontends should be pretty much in its original form.

BAH

Subtract a hundred and four for e-mail.