[Simh] PDP-11 Mag Tapes

[Shoppa, Tim]

Under some circumstances DOS-11 tapes have the last record of a text file, shorter than 512 bytes. And it can be an odd length. I've seen some tools that deal with tape images, getting confused by tape records with odd lengths. Note that DECUS TPC for example pads out the tape record with an extra byte etc. (well really that's probably RMS-11). With you saying that not all files are listed on some tapes... I wonder if the tape image is "out of sync" after an odd length block.

You might try running RT-11 DIR/DOS (aka FILEX) against the same tape images. If it's the odd record length thing then it probably gets out of sync at the same place (although surely with a different error message.)

I myself am interested in old DECUS stuff from that era. E.g. the early DOS-11 and RSTS-11 stuff that DECUS published on machine-readable media. Generally DECUS sent that stuff around on paper tape but you might run across copies of it on 9-track.

With regards to early RT-11, we already have copies all the way back to V1. But DOS-11 stuff from that era is spotty at best in comparison.

From: simh-bounces at trailing-edge.com [mailto:simh-bounces at trailing-edge.com] On Behalf Of B M
Sent: Tuesday, October 11, 2011 12:29 AM
To: simh at trailing-edge.com
Subject: [Simh] PDP-11 Mag Tapes

I am using SIMH with some mag tape images (TM device) under DOS-11 in the PDP-11 simulator.

Several of the tape images do not work.  DOS-11 doesn't seem to process the file correctly.
Executing 'MT0:/DI' in PIP under DOS-11 frequently fails with an error code (eg. S257), or
does not have all files listed.

The format of the tape images is as follows:
      <label><record1>...<recordN><eof>                           file #1
                 ...
      <label><record1>...<recordN><eof>                           file #x
      <eof>
      <eof>

The <label> record is a DOS-11 file entry (filename in RAD-50, ...) and is 14 bytes long.

Each of the data records has a 4 byte record length at the beginning and end of the
record.  The length is a 4 byte integer in little endian form.  A 512 byte record ends
up being 512 bytes + 2 x 4 bytes or 520 bytes in total.

An <eof> is 0x00000000 (4 x 0 bytes).

End of data is 2 <eof> back to back.

I have written a program that verifies the content of the tape image.  I see the expected content of
files under my verification program, but the content frequently does not match what DOS-11 gets.

I am looking for suggestions on how to tackle this problem.  I have several 800bpi 9-track tapes which
contain DEC source code for really early versions of RSTS, RT-11, etc.  I am looking to image the files
and make them available for SIMH.

I can supply copies of the tape images as necessary.  Note that I am using 40 year old equipment
to retrieve the tape images.  The tapes are being read correctly without read errors popping up.

Is anyone interested in RSTS or RT-11 for really early PDP-11's such as PDP-11/20 or PDP-11/10??
I am trying to get the early versions of various OS'  running on my PDP-11/10 and PDP-11/20.
Eventually I also want to get Unix running on my PDP-11/20 (just for the geek factor!!)

Thanks!

--barrym

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.trailing-edge.com/pipermail/simh/attachments/20111011/6f2127d4/attachment-0002.html>