[Info-vax] OpenVMS.Org quick pool
Bob Koehler
koehler at eisner.nospam.encompasserve.org
Tue Aug 21 09:47:30 EDT 2012
In article <k0udhm$48m$1 at dont-email.me>, Stephen Hoffman <seaohveh at hoffmanlabs.invalid> writes:
> And why would there be a whole pile of effort expended back-porting
> stuff to VAX? That architecture was deemed to be on the way out, just
> as soon as the Alpha systems were released. Whether a task is
> technically feasible collides - once again - with business
> considerations and the financial feasibility of the task.
Exactly. Every time you make a copy of something, you double the
maintenance effort. So just to keep up with changes in one copy
takes a huge effort in the other. And that requires huge business
justification.
We're told that the Alpha and IA64 share a source pool. We've seen a
few things in IA64 that aren't in Alpha, but it's mostly support for
new hardware. It makes sense not to build adapters to hang the
lastest peripherals off old Alphas, just like it made sense not to
add the latest graphics adapters to VAX 11 series. So it makes sense
not to build or test Alpha copies of that code.
But most of wht is on IA64 is on Alpha. And I expect that's mostly
write once, compile twice, certify for each supported model, with a
few tweaks. So the work to support m models of Alpha and n models
of IA64 is only slightly larger than to support m+n models of one
architecture.
More information about the Info-vax
mailing list