On Fri, Jun 27, 2025 at 10:49 PM Frank Leonhardt via cctalk
<cctalk(a)classiccmp.org> wrote:
I'm not sure they didn't - it was something I should have found out
fifty years ago. And the same readers coped with 5 and 8 hole tapes,
where the sprocket is in a different position.
Actually it isn't. It's 3 data holes from one edge of the tape in both
cases. So take that edge as one reference and have an adustable guide
for the other edge
Where there were no holes
other than the sprocket they flew through. I don't think it was
necessary to read a NUL, but it might have been. And there may have been
different handling for the eight and five bit. I shall have to ask Mr Onion!
My guess is that the machine it was connected to ignored nuls so it
could run the reader flat out, but other caracters had to be
processed.
If anyone wants to theorise, the roller was controlled by a solenoid and
could move the tape forward one character at a time, or apparently
free-run. In other words it didn't seem to unload and load on each
character. At the time I just used it - I didn't think that much about it.
This sounds like an Elliot optical reader, or maybe a Trend. Those can
run continously if you engage the pinch roller solenoid, alternatively
you can control the solenoid (and the brake) to step the tape a
character at a time.
And FWIW both of those readers need sprocket holes.
-tony