Am 21.12.2014 um 12:57 schrieb Orchid Win7 v1:
> Don't ask me how I found this, but:
>
> http://thesis.library.caltech.edu/2441/1/Knuth_de_1963.pdf
>
> Anybody have any ideas how this was typeset?
>
> It *looks* like it was written on a typewriter. But that obviously can't
> be true, because typewriters don't have Greek letters and other
> mathematical symbols on them. So... how?
>
> (Incidentally, I bought a book on Galios Theory from Amazon. It was only

> to read complicated formulas...)

There is some evidence in the scanned document that special symbols like 
greek letters were added later, after the bulk of the document had been 
typed already; some of those special characters appear to be drawn, 
while others appear to be typed.

So it's probably /not/ one of these:

http://blog.modernmechanix.com/mags/MechanixIllustrated/12-1958/double_keyboard.jpg


Sub- and superscript were of course done by moving the paper up or down 
manually by a detent(*), typing the letter, then moving the paper back.

(*On a typewriter, the cylinder would usually have detents at fractions 
of a nominal line feed; operating the carriage return lever would 
advance the paper by multiple such small steps at once, with the exact 
number configured via a lever. The cylinder could also be operated 
manually via a knob at either side.)

I suppose underline was achieved by overtyping with a special underline 
characters; IIRC common typewriters lacked such a special character, and 
you'd use the minus instead.

Some special characters were certainly achieved by overtyping as well, 
such as the not-equal sign (equal sign and slash) or the "end of proof" 
sign (opening and closing square brackets).

Post a reply to this message