From: utzoo!decvax!ucbvax!info-vax
Newsgroups: fa.info-vax
Title: Re: SCRFT handling of attributes on braindamaged terminals
Article-I.D.: ucbvax.528
Posted: Mon Dec 27 23:59:28 1982
Received: Wed Dec 29 00:53:37 1982

>From EGK@MIT-OZ  Mon Dec 27 23:50:24 1982
Mail-From: ARPANET host SANDIA rcvd at 27-Dec-82 0053-PST
Mail-From: ARPANET site MIT-MC rcvd at 27-Dec-82 0151-MST
To: EPS@MIT-MC
Cc: EGK@MIT-OZ, info-Vax@MIT-MC
In-Reply-To: Your message of 25-Dec-82 0114-EST
Remailed-Date: 27 Dec 1982 2008-PST
Remailed-From: the tty of Geoffrey S. Goodfellow  <Geoff5 at SRI-CSL>
Remailed-To: Info-VAX@SRI-CSL: ;

I have written much display hacking stuff recently and have run across
the problem you mentioned.  Most terminals that handle atributes DO
*NOT* have this problem.  there are a few people making terminals with
this brain damage, noteably, ADDS, some teleray models, ADM5 (these are
all that come to mind imeadeatly) (if you can supply more, I'd like to
see them listed!)  The reason they do this it seems, is that people who
designed the terminals are "old-timers" who were designing terminals
back in the days when memory was expensive and in order to keep the cost
down, they skimped on memory by taking up a screen memory location with
atribute info.  Idealy there should be no terminal sold today with this
gross misfeature.  It is difficult to write any decent algorithm to
account for both normal terminals and brain damaged ones.  The bigest
problem is when you used atributes on the screen semi randomly, the
columns are all off on the lines you dont touch.

My solution to the problem is to ignore atributes on brain damaged
terminals.  Hopefully, if enough people do likewise, terminal
manufacturers will get the message and update their terminal designs.

			  
			    0
		      +  @  +  @  +
			\ * | * /
		    % + -- EGK -- + %
			/ * | * \ 
		      +  @  +  @  +
			    0
-------