Subject: Re: SRM vs ARC
To: David Seifert <email@example.com>
From: Chris G. Demetriou <firstname.lastname@example.org>
Date: 08/11/1997 13:39:17
[ These are not the words or thoughts of my employer... 8-]
> It is not encouraging to see that even DEC's own people don't like SRM.
What, you think everybody within Digital likes VMS, UNIX and NT?
or likes the Alpha? or likes the VAX? or ...?
There are a lot of technical arguments both for and against a lot of
pieces of hardware and software. SRM is a just piece of software,
albeit one which is critical for certain operating systems to work on
> I like SRM, although I'd like to see more documentation for it.
There actually are some pretty good user docs...
> There are clearly a lot of untapped powerful features in there.
Yes, and the development of those powerful features have had a large
cost, both in firmware size and in stability and reliability.
Many, in fact I'd say _most_, of the features in the newer versions of
the SRM console are ... useless for the functions that console
firmware is supposed to perform. I'd trade multiple processes, lots
of interactive help, 'more', pipes, a shell-like language, etc., away,
if that meant that I could have firmware which would reliably boot the
machines i've had to use over the past couple of years. The old
TurboChannel-system SRM was great: it wasn't very intelligent, but it
worked. The new SRM is much more "intelligent," but the fact of the
matter is, (in my experience, with Digital UNIX and NetBSD) when you
try to reboot with it sometimes it just doesn't reboot, and that's