1" paper tape buy ?

[CSquared]

> Date: Wed, 04 Mar 2009 00:41:07 -0800
> From: Brent Hilpert <hilpert at cs.ubc.ca>

> CSquared wrote:
>> 
>> On Mon, 2 Mar 2009 11:41:14 -0800 Rich Alderson <RichA at vulcan.com> said:
>> 
>> >> From: CSquared
>> >> Sent: Friday, February 27, 2009 2:53 PM
>> 
>> >> I also seem to recall that Univac or some other computer manufacturer
>> >> actually used a 90 (?) column card for even better packing density.
>> >> I'll leave that improved calculation to someone who actually remembers
>> >> how many columns those cards were blessed with.
>> 
>> > Univac 90-column cards are Hollerith cards just like the "IBM" cards.
>> > They consist of upper and lower groups of 45 columns of 6 round punches
>> > (instead of the rectangular punches favoured by IBM).
>> 
>> Interesting.  I don't know how I remembered the 90 columns as I never
>> worked with a Univac computer - only IBM's and various minis which mostly
>> didn't even have punched card readers at all.  I'm not real sure I've
>> ever even seen one of the Univac punched cards.  Your description
>> of the hole grouping sounds like maybe they encoded characters as
>> 6-bit codes vs. the "12-1 is an A", "12-2 is a B" scheme used by IBM.
>> I'm not real sure that's even correct either - it's been a right good while.
> 
> It doesn't present the full encoding but there is a picture and some more info here:
>  http://www.fourmilab.ch/documents/univac/cards.html
> 
> The labelling on the card suggests the numeric digits were punched in biquinary
> format. I'm going to guess that that goes back to the origins of that card type
> when only numeric data where utilised, and that alpha & punctuation were added
> later, using the remaining 6-bit binary patterns.

I was just about to say I had never seen one of those, but then I seem to recall the Dallas-Ft. Worth turnpike also used Univac punched cards as tickets now that I think about it.  That former turnpike has been a freeway (just another segment of I30) for quite a few years now.  Perhaps that is the source of my hazy memories of those cards.  If the coding is bi-quinary as the web site suggests, that brings back other unrelated memories of the adder in the IBM 1401.  Of course I'm sure those show up in other machines as well, but that was my first encounter with them in about 1962 or so.

That looks to be a very interesting web site for many reasons, though those cards were not what I visualized.  I have another vague memory of a card with oval holes (longer in the vertical dimension) with the 90 holes all in 1 row rather than in 2 rows as the web site shows.  Now I'm going to have to do a bit of searching to see if I can find something that matches that memory.