Paper tape and 8th bit?

[Tony Duell]

> 
> Hi Guys,
> 
> I'm attempting to recover some software from paper-tape.
> 
> Never actually had a PT reader in my "altair days", I fooled around
> with magnetic tape systems (audio and 9-track), and later went to
> the NorthStar disk system... Did a bit of paper-tape stuff with the
> university systems, but I don't recall any real details.
> 
> Anyway - I'm using an OP-80A which is a very simple manual feed
> reader - you position a light over the unit and pull the tape through
> wire guides over an optical sensor and it provides parallel data.
> 
> This all appears to work OK.
> 
> The tape I've been testing with is a Processor Technology "BASIC
> VDM DRIVER" - it contains driver software for the PT VDM-1 video
> board.
> 
> My question - Does anyone know what they are using the 8th bit
> for?  - I get nicely readable ASCII BASIC source out of it, except

Well, AFAIK few punches [1] or readers ever did anything with the 8th bit 
other than just transfer it to/from the system. So what it means depends 
on the device that punched that tape.

[1] I say 'few', not none because the serial adapter board for the 4070 
can be set up to make this a locally-generated parity bit. In other words 
you can send it 7 bit characters (or 8 bit with the top bit ignored) and 
it will punch them with the 8th level an even parity bit. All 
link-selectable.

> that the 8th bit seems to be somewhat randomly set on certain
> characters. If I strip the 8th bit I get what appears to be legit BASIC
> code.
> 
> I thought it might be parity, however this does not appear to be

That would have been my first guess too. Normally even parity, since then 
totally blank and all holes are valid characters (the latter can be used 
to overpunch any other character, and was origianlly simply ignored by 
the system).

Of course on a binary tape (true binary, not Intel-Hex or something), 
it's just anothe bit.

> the case - the codes 0A (00001010) and 0D (00001101) both
> appear with the 8th bit clear - If the 8th bit were parity, one or
> the other should have it set. Other characters always have it set,
> for example 'T' (54) seems to always appears as (D4).

Doe any character exist in both forms (with the top bit both clear and 
set)? If so, could it be something like a 'start of statement' marker or 
'start of keyword' or soemthing like that?

-tony