[Simh] Formally documenting PDP tape drive behavior (and other facets of SIMH and the systems it emulates)

[Rich Alderson]

> From: "Armistead, Jason" <Jason.Armistead at otis.com>
> Date: Thu, 21 Mar 2013 13:55:18 +0000

Sorry to drag this back up.  At the time it was posted, I was busy trying to
get KI10 diagnostics from a disk on the museum's Toad-1 system (TOPS-20)
freshly compiled from the MORDOR sources (derived ultimately from the backups
preserved by Tim Litt) onto DECtapes for the KI.  My colleague Ian King was
looking at alternatives to the method by which I was proposing to accomplish
this, which is what began this thread.

> There has been a lot of interesting discussion about tape drives on PDP
> systems in the last week or so.  It's great to hear those familiar with the
> inner workings of these devices, or with access to the manuals, share their
> wisdom.

> One question though - after the "old timers" are gone, would a newbie to SIMH
> be able to figure these same things out ?

Yes, although the route to knowledge is a bit complex.  I had to figure it out
for myself several years ago, not for SimH but for a forensic analysis of raw
bit dumps of the contents of some important DECtapes (2MB DOS files with the
real data hidden somewhere at random inside).

What the eager learner needs to do is to read all of the manuals describing
DECtape controllers for the various DEC architectures.  (The TD8E can be
skipped, for all the reasons Johnny, Bob and Tim have made clear.)  Yes, much
of the information is redundant, but I found that there were things mentioned
in one manual or another that were repeated nowhere else.

So you would want to look at the files in http://bitsavers.org/pdf/dec/dectape
as well as those for various controllers in the pdp-<N> directories.  You're
looking for TC01, TC02, TC08, TD10, and TC11.

> Or, in other words, do the accompanying comments in the code and the SIMH
> documentation adequately capture the "nuggets" of the conversations that take
> place on this list ?  And do we provide enough in the way of "breadcrumbs" to
> allow a newbie to re-discover this same information ?

> Capturing this knowledge is a hard thing to do, and with retrocomputing we
> are presently at a point in time where we can still rely on folks who
> designed, built, tested, operated and programmed these systems, and in some
> cases they still even have running systems available for experimentation and
> measurement.  But what happens when that is no longer the case ?

> Just food for thought ... let's make sure SIMH and its fellow emulators like
> MAME / MESS can still be enjoyed and understood by future generations.

Amen!

Rich Alderson
Living Computer Museum
Seattle