TERMDEF.TXT contains notes on the effective use of TERMDEF,
the AUTODEX Terminal Definition program.

1. General Information
======================

TERMDEF (Terminal Definition) lets you edit any existing terminal definition
in AUTODEX.  The program is menu driven, and "almost" self-explanatory.

You will need your terminal hardware manual handy.

Before running TERMDEF, you should have run INSTALL to define a terminal
to AUTODEX that is "close" to yours.  This terminal definition
will be the starting point for the edit process.  If you are unsure
of a terminal that is "close", try the Osborne 1.

After entering

     A>TERMDEF<RETURN>

you will be given an opportunity to change disks and read this documentation.
After that the master menu is displayed.  Start by typing

     1

to list your current terminal definition.  Compare the information with your
hardware manual, and when you see discrepancies, enter a

     2

to begin the edit process.  After making edits and returning to the master
menu, type a

     3

to update the AUTODEX terminal definition.  If you are dissatisfied
with your new terminal definition, type a

     4

to exit without changing AUTODEX.

Now run AUTODEX by entering

     A>AX<RETURN>

and looking at the results.  Repeat this process until you are satisfied
with the results.

2. Terminal Editing
===================

When editing a terminal definition, TERMDEF displays about 13 different
categories of terminal information to edit.  All but a few are
self-explanatory.  The ones that are not are Position Cursor,
Move Cursor, Highlight a Field, and Pad Character.

3. Editing: Position Cursor
===========================

AUTODEX's Terminal Definition  supports six different conventions for
positioning the cursor to a specified row and column. You may
specify your sequence explicitly if it is not on the standard list.

The first, and by far the most common, is descended from
the ADM 3A, and requires a four character Escape sequence:
          ESC = R C
The ESC and = are transmitted as is.  The R is the desired
row number + 32 (where rows start at 0).  For example, row
20 would be the ASCII character 2, with a decimal value of
20+32 = 52, which is hexadecimal 32.  The C is the desired
column, using the same convention as for the row.  With
the exception of the terminals below, this is 
universally used.

The second is for the DEC VT-100 and other ANSI standard
terminals.  The sequence is:
          ESC [ Row ; Column H
The Row  and Column values start at 1 and are represented
as ASCII text strings.

The third is for the entire line of Vector Graphic
computers.  The sequence is:
          ESC Column Row
The column and row start at 0 and are represented as
ASCII numeric values.  Some implementations add 128 to the
values to turn on the high bit, but this is not required.

The fourth is for the ADDS Viewpoint and DEC VT-52.  The sequence is:
          ESC Y R C
The R and C are just like the ADM 3A.

The fifth is for the Hewlett Packard 125.  The sequence is:
          ESC & a  Row r Column C
The Row  and Column values start at 0 and are represented
as ASCII text strings.

The sixth is for the Hazeltine 1500.  The sequence is:
          TILDE CTRL-Q C R
The column and row start at 0 and are represented as
ASCII numeric values.

Finally, you may specify exactly how the cursor is addressed.
You must carefully study your hardware manual, and then respond
to the following questions:

	1. The row bias, i.e. the row number to specify relative
        to zero.  For example, if the row bias were 32 (very common),
        then the topmost row (row 0) would be specified as a 32.
	2. The column bias; the same principle as the row bias.
	3. Whether rows or columns are transmitted first
	4. Whether row and column values are in binary
	   or ASCII text
	5. The ASCII sequence before the first of row or column
	   are transmitted (called The Leadin Sequence)
	6. The ASCII sequence (if any) separating the
	   row and column values (called The Middle Sequence)
	7. The ASCII sequence  (if any) after the latter of row
	   or column are transmitted (called The Ending Sequence)


4. Editing: Move Cursor
=======================

AUTODEX's Terminal Definition  supports cursor movement sequences
(i.e., up, down, left, right) of one to three characters in length.
By far the most common, but by no means universal convention,
is for one character sequences.

You are first asked to define the output cursor sequences, that is, the
sequences that are transmitted by the computer to move the cursor up, down,
etc.

You are then asked if the input cursor sequences should be the same.  In most
cases, the answer to this is Yes.  There are two notable exceptions
to this rule:

     * When the computer/terminal uses different codes for input and output.
       The Xerox 820 II is the only one we know of that does this, but
       there are no doubt others.

     * When you are used to using specific control keys that do not match
       the arrow keys to move the cursor around.  If someone has used
       WordStar, for example, for several years, they will probably want to
       customize their cursor input sequences to match WordStar.

One notable exception to this process is the Hazeltine 1500, which has the
unique distinction of having cursor input sequences of different length.
Further, the right arrow key (on the models that have one) generates
a CTRL-P, which is both AUTODEX's and CP/M's print toggle!
The move cursor routines will ask you if this is a Hazeltine 1500.  If
so,  respond y.  This will get you the output sequences fine, but
you must use single character control keys for up, down, left, right
on input!


5. Editing: Highlighting
========================

Terminals support highlighting in one of two ways:

     1. Active: the field to be highlighted must be transmitted to the
        screen, preceded by a sequence to turn highlighting on, and
        followed by a sequence to turn it off.
        Most terminals are this way.

     2. Static:  the field to be highlighted is already on the screen;
        the program transmits a highlight stop character at the
        end of the field, and a highlight start character at the beginning
        of the field. The TeleVideo line and portions of the
        Lear-Siegler line operate this way.

It is often not clear from the hardware manual which of the above is the
method used.  If in doubt, try the Active method and look at the results!

6. Pad Character
================

Some terminals need pad characters after Clear Screen, Clear End of
Screen, or Clear End of Line.  Pad characters are non-displayable characters
that give the terminal time to finish the operation (e.g. Clear Screen)
before the displayable characters are received.  This is a characteristic of
older terminals without buffers for the characters being transmitted.  Newer
terminals and the modern integrated computer/terminals (Xerox 820 II,
Osborne 1, etc.) rarely (if ever) need these pad characters.  If you sometimes
see one to three characters of gibberish on the screen, go into the pad
character function and set the number of pad characters for all operations
to zero.


 start character at t