regan@gvax.UUCP (03/22/87)
This message is prompted by recent postings on both scientific
word-processors for the ST series and on the way ST keyboard
information is handled. I use the T^3 ("T-cubed" or "T-three")
scientific text-processor on MS-DOS systems, and have
corresponded with its designers (TCI Software Research, 1190-B
Foster Road, Las Cruces, NM, 88001) for several years.
PLUG ON
T^3 is great; two major advantages are the ability to see
mathematics and alternate font styles on-screen as you type, and
to create new symbols for both printer and screen via an
uncomplicated font editor. Multiple-layer lines are treated as
single units, and any number of keystrokes may be saved as a macro.
I've written a 400-page doctoral dissertation and several papers on it.
Even for just ASCII text entry it has the best system
for getting around a document that I've seen. There are no 'modes';
one puts down a visual mark to insert text. To jump forward [backward]
to the next occurrence of a given character, be it a letter, space,
period, or carriage-return, simply hold the right [left] arrow key
down and type the character. This leads right into my main point,
so
PLUG OFF
Two years ago I contacted TCI about the not-yet-out ST and its
CGA-slaying mono graphics screen. The obstacle they mentioned wasn't
any difficulty in porting the code (p-system & IBM Pascal) or working
with GEM/TOS, but the keyboard: then (as now?) it would only store
two (2) key-presses simultaneously. The program uses three-key
combinations copiously (and some fours), so this was no-go.
I called Atari technical staff about this last December, since I'd
like to buy the 1040ST + planned MS-DOS box come June and my tax refund,
and their response was, in two words, "Good point!". Let me add
some tips on what one can do with "n-key rollover" and separate
recording of press and release (the latter I understand the ST's have):
(1) Making sensitive commands difficult to type by mistake:
T^3 uses /DEL/Shift/End (held down in that order) for "delete to end
of document". If you don't release the keys right away,
it will show you what it's about to do in inverse-video,
and if it's a mistake you can cancel the command by hitting <ESC>
(so a fourth key down) before releasing. Such conventions are
universal in the system, applying also to...
(2) Avoiding the display of pop-up menus:
Menus only pop up when the "Menu" key <F9> is released (GEM mouse programmers
take note), but are "there" when it is pressed. E.g. to mark the whole
paragraph the cursor is in and move it to the end of a document, one
can type the "chord sequence" /Left-Arrow/Shift/CR/, /F1/CR/,
/Shift/End/, /F9/+\/m\/+\. No menus appear, saving one screen-write
time, and the "mark from previous shift-CR (paragraph start) to next CR",
"jump to document-end", and "move block to cursor position" commands
are executed quickly and "silently". I won't explain the keys' meanings
(except the numeric-keypad <+> means "Accept")--the point is
that one gets all the virtues of both menu-based and command-based WP
systems combined. (Record the whole thing as a macro, calling it "pmove"
if you wish, and invoke it by typing /Ctrl/pmove to save more time later.)
(3) "Chording" has lots of other applications: music from QWERTY input,
combining key-commands with the mouse buttons in arcade-style games,
you name it.
So, aside from saying "Atari--please take note", let me ask some specific
questions:
I. Do currently-made ST's have the n-key feature (under TOS)?
II. If not, can the keyboard hardware be addressed in a fashion
that will still make programs portable within the ST line?
Will the Megas have the feature?
III. Will the planned 8088 box emulate the full IBM BIOS means of
handling the keyboard?
Neither I nor the TCI people have been able to try T^3 with Paradox's
MS.EM to see whether the key-combos work.
I, and also the T^3 people, will appreciate and consider all replies.
Mail me at 'regan@gvax.cs.cornell.edu' (Arpanet) or for Bitnet,
'regan%amvax.tn.cornell.edu@CRNLCS.BITNET' whatever you may
not wish to post. Thanks,
Dr. Kenneth W. Regan,
Center for Applied Math.,
211 Sage Hall, Cornell Univ.
Ithaca, NY 14853-6202.