In message <Pine.LNX.4.62.0507050047280.20412@Psilocybe.Update.UU.SE>, Johnny B
illquist writes:
>On Mon, 4 Jul 2005, Petar Bogdanovic wrote:
>
>> Mike Parson wrote:
>> > The vi-keys are the vi-keys cuz that's where the arrows were on the
>> > terminal Bill Joy was using at the time he wrote it (adm3).
>>
>> I can't see any arrows on the 'h', 'j', 'k' & 'l' of the adm3[1].
>
>I very much suspect that this is an urban story without truth. :-)
>My only guess at why hjkl would be that ^H is backspace, and ^J is
>linefeed. So you already got two movements placed there. The other two
>might just be extrapolated from the placement of the first two.
>
>But that was a very personal guess, and personally I also think that vi
>sucks bigtime. :-)

It's not an urban legend; there are a fair number of us who used them, 
way back when.

There was no standard for escape sequences or function keys then; 
low-priced terminals, in particular -- and the ADM 3 was about as cheap 
as they got -- used as little circuitry and as minimal a microprocessor 
as possible.  That's why termcap was invented -- control codes differed 
wildly.  Not all terminals had arrow keys; of those that did, not all 
transmitted something.  Two-character sequences were hard to deal with, 
because machines didn't have high-resolution timers.  

I should add that I used the ADM 3 as little as possible -- it used a 
5x7 pixel matrix for characters, and that made lower-case characters 
really ugly -- the "descenders" on some letters (g, y, etc.) didn't go 
below the line.

		--Steven M. Bellovin, http://www.cs.columbia.edu/~smb