From: David Jeske <jeske@fiction.isdn.uiuc.edu>
Message-Id: <199509281620.LAA14639@fiction.isdn.uiuc.edu>
Subject: Re: TTY monitor
To: vsta@cisco.com
Date: Thu, 28 Sep 1995 11:20:20 -0500 (CDT)
In-Reply-To: <Pine.3.89.9509281012.A9255-0100000@zeus.polsl.gliwice.pl> from "Filip R. Zawadiak" at Sep 28, 95 10:18:04 am
Reply-To: jeske@uiuc.edu
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 1178      

> FALSE! There should be simply more TTY monitors... And something like 
> virtual keyboard server - like todays virtual consoles. Then at first 
> level I could select to which vkeyb I'm now typing, and another keys 
> could select virtual screen of current TTYmon.
> 
> PS. Why kbd + cons = cons2 ? PC doesn't have console, it has keyboard and 
> one or more graphic cards...

I'm hoping to change this around. My original idea was to go with a 
totally virtualized design where one "virtualizer" would talk to a kbd 
server and a cons/vga server. Where those servers wouldn't be virtual at 
all, but instead would implement a "save/restore state" protocol. 
However, I realized that the "virtualizer" was really a "serializer" and 
that data could not be output to the screen you were not looking at. 

So I want to come up with a clean I/O concept which can cleanly 
virtualize the connections but also allow the servers to be "dumb".

-- 
jeske@uiuc.edu   + David Jeske(N9LCA)<A HREF="http://www.cen.uiuc.edu/~jeske/">
NeXTMail accepted + CompEng Student/OS,uP design/Call Gtalk at (708)998-0008
    User of Linux/NeXT/DOS/WIN/OS.2/VSTa (all coexisting on one system) </A>