Re: [TUHS] [TUHS} cat -v and other complaints

Re: [TUHS] [TUHS} cat -v and other complaints

[Doug McIlroy]

> you can't tell me
> this system was designed with the idea of running it using text terminal
> and no mouse.  There is also no cursor addressing, no curses.

The well named Curses and the associated vi were an ugly outgrowth
of glass screens--an outgrowth many of us in the Unix lab never
adopted. That branch of evolution was unrelated to the Blit branch that
essentially preserved the old TTY interface, except that one could have
multiple terminals on screen and a mouse was available to give mechanical
help for manual cut/paste/edit activities. Plan 9 terminal-handling
smoothly continued that evolutionary branch.

Mouse support could have been used to take off in a radical direction,
but it wasn't. To my mind, the primary innovation in Plan 9 was not
terminal support, nor everything-is-a-file. Rather it was an advance in
what Vyssotsky called "distributable computing", where components can
collaborate in a uniform way, no matter where they are. The key was the 9P
protocol that unpacked the notion of file type--a unifying principle
that brought simplicity and generality to a diversity of particulars.

Re: [TUHS] [TUHS} cat -v and other complaints

[Will Senn]

Doug et al.,

I’m interested in learning about this curses vs blit business. Is there a writeup or book chapter out there that covers this in any detail?

Thanks,

Will

Sent from my iPhone

On Sep 8, 2018, at 7:02 AM, Doug McIlroy <doug@cs.dartmouth.edu> wrote:

>> you can't tell me
>> this system was designed with the idea of running it using text terminal
>> and no mouse.  There is also no cursor addressing, no curses.
> 
> The well named Curses and the associated vi were an ugly outgrowth
> of glass screens--an outgrowth many of us in the Unix lab never
> adopted. That branch of evolution was unrelated to the Blit branch that
> essentially preserved the old TTY interface, except that one could have
> multiple terminals on screen and a mouse was available to give mechanical
> help for manual cut/paste/edit activities. Plan 9 terminal-handling
> smoothly continued that evolutionary branch.
> 
> Mouse support could have been used to take off in a radical direction,
> but it wasn't. To my mind, the primary innovation in Plan 9 was not
> terminal support, nor everything-is-a-file. Rather it was an advance in
> what Vyssotsky called "distributable computing", where components can
> collaborate in a uniform way, no matter where they are. The key was the 9P
> protocol that unpacked the notion of file type--a unifying principle
> that brought simplicity and generality to a diversity of particulars.

Re: [TUHS] cat -v and other complaints

[Ralph Corderoy]

Hi Will,

> I'm interested in learning about this curses vs blit business. Is
> there a writeup or book chapter out there that covers this in any
> detail?

https://en.wikipedia.org/wiki/Blit_(computer_terminal) is a jumping-off
point.  And I suppose the same goes for curses(3):
https://en.wikipedia.org/wiki/Curses_(programming_library)

-- 
Cheers, Ralph.
https://plus.google.com/+RalphCorderoy

Re: [TUHS] cat -v and other complaints

[Arthur Krewat]

On 9/8/2018 10:22 AM, Ralph Corderoy wrote:
>
>> I'm interested in learning about this curses vs blit business. Is
>> there a writeup or book chapter out there that covers this in any
>> detail?
> https://en.wikipedia.org/wiki/Blit_(computer_terminal) is a jumping-off
> point.  And I suppose the same goes for curses(3):
> https://en.wikipedia.org/wiki/Curses_(programming_library)
>

In my opinion (as retarded as I can be sometimes), this is an 
apples-and-oranges comparison.

Blit is a completely new terminal type, with specific operating 
system/software support.

Curses is a way to control various already-existing terminal types. DEC 
terminals, Hazeltine, etc. A recent termcap on my Solaris server has 472 
entries. The wide-ranging support was quite important.

Many people/institutions had a variety of terminals already, usually 
recycled from previous systems. I remember one instance when I was 17 
years old working at BOCES/LIRICS on Long Island, and an office worker 
in a local high-school looked at me like a deer in the headlights when 
they could no longer use their current-loop terminal and acoustic 
coupler. Sorry, the leased-line mux in the other room can't do that. It 
has to be RS232. We gladly gave them a new LA36. Which invoked another 
set of "how do I..." questions. Ah, progress. (This was to support 
TOPS-10 on DEC KS10's, but the same thing happened many times over my 
early career. People just didn't want to give up what they already had)

ak