MULTOS/8 progress, and a question...

[Josh Dersch]

Annnd, top-posting with an answer to my own question.

One has to patch PIP to modify the size of the partitions; this is actually spewed out during assembly of MULTOS.BN:

  2.32. 43                           NOTE:   SYS DEVICE SIZE = 2720
  2.32. 43                           NOTE:   SET PIP LOCATIONS 13623, 13656,
  2.32. 43                           NOTE:   AND 13657 TO 5060

(Values will vary depending on how your system is configured.)  That'll teach me to ignore the output of the assembler when there are no errors :).

Use ODT to hack up  PIP, SAVE it and you're good.  J1: through J4: are now 1432 blocks in length.

The SYS: device is still reporting as ~3000 blocks, though.  And since it comes with over 2700 blocks filled, I made sure to free up ample space, and did a SQUISH so things won't collide (hopefully.)

Hopefully this saves someone some time in the future...

- Josh


> -----Original Message-----
> From: cctalk [mailto:cctalk-bounces at classiccmp.org] On Behalf Of Josh
> Dersch
> Sent: Tuesday, June 14, 2016 10:14 AM
> To: 'General Discussion: On-Topic and Off-Topic Posts'
> <cctalk at classiccmp.org>
> Subject: MULTOS/8 progress, and a question...
> 
> Hi all --
> 
> Having overcome my earlier issues with SIMH, I now have MULTOS/8 happily
> running on the real hardware, with two terminals.  (Yay!).  I'd like to get the
> system up to four terminals and here's where I'm running into trouble.
> 
> I've modified PARAM.PG appropriately (set JOBS to 4, TERMS to 4, and filled
> in the device codes for the extra terminals).  I've rebuilt JOBS.BN and
> MULTOS, ran BUILD and inserted J1 through J4 and everything runs and the
> system comes up with 4 terminals (yay!).
> 
> What isn't working properly are the job devices J1 through J4 -- each of these
> provides each terminal/user his/her own "virtual" drive on the system's RK05
> pack -- effectively it partitions the RK05 into four smaller drives, J1: through
> J4:.  With the two user setup, J1: and J2: are approximately 3000 blocks in size
> (so each basically gets one side of the RK05 pack).  With the four-user setup,
> I'd expect J1:, J2:, J3: and J4: to each be about 1500 blocks, but they're still
> reporting as 3000 blocks each.
> 
> What's more, the actual starting offsets of J1: through J4: appear to be what
> I'd expect (that is, J1 starts at 0, J2 starts at ~1500, J3 starts at ~3000, etc.).  So
> if I start filling these drives, eventually they clobber each other.
> 
> The only manual I've found (http://www.pdp8.net/os/multos8/) suggests
> that these should be appropriately sized (see section 2.1.1) automagically,
> but this does not seem to be the case; doing a ZERO on them creates a
> filesystem 3000 blocks in length.
> 
> I'm digging through the source code to look for clues, but I haven't found
> anything yet.
> 
> Anyone fooled with this before?
> Thanks,
> Josh