[Simh] TOPS-20 4.1 BUGHLT

[Cory Smelosky]

On Mon, 11 May 2015, Timothe Litt wrote:

> The first part is self-explanatory.  QUASAR looks for SPOOL: (usually
> ps:<spool>)
> early in initialization.  It's not there.  It's possible that the
> logical name is not/mis
> defined.  Either RCDIR% or GTDIR% returned "Structure is not mounted".  If
> you believe the successful "crash saved in ps:<spool>quasar-nsd-crash.exe"
> message, it's the logical name.  Perhaps the TSU that you applied updated
> QUASAR from the hard-coded ps:<spool> to the logical, but didn't define
> it.  Or perhaps the message is optimistic and/or ps:<spool> is corrupt.
>

I repeated the install, fresh structures, repeated TSU install.  No issues 
this time.

> The bughlt could be anything.
>
> Together, probably a corrupt disk - perhaps an unclean shutdown caused a
> page
> write to be incomplete.  You can try running CHECKD.  You can debug &
> fix with
> some mixture of eddt, mddt & filddt.  Or you can build a new PS:.  If
> necessary,
> there's an excellent chance that your local software specialist can
> recover data
> from the old one.  If you can find her.
>

Sounds like just a corrupted disk.

> As with any corrupt disk, the more you try to run it before it's fixed,
> the worse
> the chances of complete recovery as data will be reused.
>

It was an install done just that day so no major data loss.  Just an hour 
of my time.

> Disk space is now so cheap that the easiest way to backup these systems is
> to simply copy the SIMH disk images (with the simulator stopped after a
> clean
> OS shutdown.)  That's my recommendation for all SimH machines... These disks
> are smaller than your typical video (or a couple of hundred photos).
>

How well do they compress?

> This communication may not represent my employer's views,
> if any, on the matters discussed.
>
> On 10-May-15 21:42, Cory Smelosky wrote:
>

-- 
Cory Smelosky
http://gewt.net Personal stuff
http://gimme-sympathy.org Projects