mark@cbosgd.UUCP (Mark Horton) (05/02/86)
Are you tired of incompatible keyboards that require you to stand
on your head to type, or to use two hands to type Control S? I am.
I've watched the most recent array of keyboards on IBM PC's and
AT&T PC's, and instead of converging to a common standard, the new
keyboards are diverging wildly and recklessly. I'd like to propose
a standard to solve this problem.
First off, you may ask, why doesn't ANSI or ISO already have a standard
keyboard layout? Well, they do. In fact, there are several of them.
Someone once said "the wonderful thing about standards is that there are
so many of them to choose from." I hate to add to the mess, but the
fact is that nobody is paying any attention to the ANSI standards.
I suspect this is because the standards are so old and so bizarre that
they don't meet today's needs.
ANSI X4.14-1971 is probably the most appropriate standard for "Alphanumeric
keyboard arrangements accommodating the character sets of ASCII and ASCSOCR."
I'll briefly mention ANSI X4.23-1982 "for office machines and supplies -
alphanumeric machines - keyboard arrangement", from which the IBM PC
and DEC VT200 seem to have derived, with the extra key between left shift
and Z, and ANSI X4.7-1973 "for typewriter keyboard arrangement". If there
are other related standards that our library doesn't have, I'd appreciate
it if someone would bring them up. I'll also ignore the bit-paired keyboard
in X4.14, since most keyboards these days are typewriter-style. (Bit-paired
keyboards are based on the Teletype Model 33, with shift 2 = ")
ANSI X4.14 Typewriter Keyboard
------------------------------------------------------------------
| ! | @ | # | $ | % | ^ | & | * | ( | ) | _ | + | ~ |
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 0 | - | = | ` |
------------------------------------------------------------------
| Q | W | E | R | T | Y | U | I | O | P | ] | | |
| | | | | | | | | | | [ | \ |
------------------------------------------------------------------
|LOCK| A | S | D | F | G | H | J | K | L | : | " | } |
| | | | | | | | | | | ; | ' | { |
------------------------------------------------------------------
|SHFT| Z | X | C | V | B | N | M | , | . | ? |SHFT|
| | | | | | | | | , | . | / | |
-------------------------------------------------------------
| SPACE |
| |
------------------------------------
Note that it doesn't include any keys like CTRL, RETURN, TAB, ESC, and
the like, and that it requires a SHIFT LOCK key to the left of A. The
semantics of SHIFT LOCK aren't defined anywhere, so I suppose the CAPS
LOCK keys you see today don't violate this, but it would seem that they
intended a typewriter style SHIFT LOCK. The bit paired keyboard requires
a SHIFT LOCK key in the same place.
It does require a CONTROL FUNCTION key, but doesn't say where. It specifies
that this CTRL key works as we expect for bit paired keyboards, but seems
to leave it up in the air for typewriter keyboards. You get the impression
that they never expected typewriter keyboards to catch on for ASCII in 1971.
There are provisions for leaving off some of the keys at the far right
(the ~, \, and } keys) if you aren't going to support those characters,
but not for moving them or rearranging them.
Note that there are three required keys to the right of 0 before you can
put in a backspace key, even though a selectric has only two. Note that
the popular practice of putting [ and ] unshifted and { and } shifted to
the right of P is forbidden. And note that there are three required keys
to the right of L before the RETURN key, compared to two for the selectric.
I only know of one keyboard that conforms to this standard: the HDS
Concept 100. It puts the CTRL key to the left of CAPS LOCK. The more
recent Concept AVT swapped CTRL and CAPS LOCK, so it doesn't conform
any more.
So let's ask: what should a keyboard look like? Well, if you ask a human
factors person, the answer is "all keyboards should be consistent so that
a user won't be confused by them." If any given user only uses one particular
keyboard, there's no problem. But for users who switch between different
keyboards all the time, standards would be nice. Since the people out there
selling computers and terminals would like to create a fairly high turnover,
consistency is important.
I've used lots of keyboards over the years, starting with an ASR 33 and
an old manual typewriter. The worst keyboard I ever used was a Hazeltine
2000, where shift space created a zero. The best is an IBM Selectric.
It's unfortunate that the Selectric isn't an ASCII keyboard. But industry
has come to the rescue, converting the Selectric layout to ASCII. Too bad
they can't agree on how to do it. And too bad they can't reproduce the
marvelous mechanical feel of the Selectric in an electronic keyboard.
(Some keyboards, like the Sun-2 or the AT&T PC 302, come close, however.)
In the process of using lots of keyboards, I learned to type with heuristics
that seem to work on most keyboards. Lately, these heuristics are falling
apart, as new keyboards get wilder and wilder. Let's examine the heuristics.
(1) Regular touch typing of letters, typewriter style, should work. This
means the QWERTY arrangement of letters, also , . ; /
should be in the normal places, with shifts < > : ?
respectively, and that there should be shift keys directly to the
left of Z and to the right of /. This worked nicely until the IBM PC
came out, breaking all attempts at touch typing. It's especially serious
to have the left shift key broken, since some of us used keypunches a lot,
and only the left shift key works on keypunches. I find myself overusing
the left hand shift key to this day because of this.
(2) The CTRL, TAB, and ESC keys are very heavily used in computer work.
Control characters are critical for most applications, especially in UNIX,
most DEC operating systems, and MS DOS. The CTRL key must be to the left
of A. I've discovered that my fingers don't seem to mind if there is a
small CAPS LOCK key between CTRL and A, such as the Concept 100 or Ann
Arbor Ambassador use, and even a large one like the VT100 or Keytronics
doesn't get in the way too much. But any keyboard with CAPS LOCK to the
left of CTRL, such as the HP 2621 or Concept AVT, breaks my ability to
type completely. Tab is used as a text character a lot, and ESC is very
heavily used in VI, EMACS, and in DOS as an exit character. TAB must go
directly to the left of Q, and ESC directly to the left of 1.
(3) The return key should be as large as possible, with exactly two
keys between it and L. These keys are generally ;: and '". It should
be at least two keys wide, and an L or 2x2 square are even better.
In practice, however, I never seem to hit the top part of the L.
(4) The numbers 1-0 should be in the standard places. The shifted numbers
should be !@#$%^&*(), typewriter style. (Bit paired keyboards, where ( and
) are shift 8 and 9, are very difficult to use.)
(5) The location of backspace doesn't matter much. It has never been very
well standardized, sometimes there are two keys between it and 0, sometimes
three, and sometimes it's put in a strange place, like on the right side of
the QWERTY row in the Ambassador. I find that CTRL H is more dependable,
and always seems to work as backspace, so I use it instead. (This means
I depend on CTRL even more.) Nonetheless, it seems reasonable to me that
backspace should be within touch typing distance, which means only two
keys between it and 0. Keyboards with three keys in between force the
user to remove the right hand from the home row to hit it.
(6) The strange ASCII punctuation ( [] {} \| `~ ) varies so much from
keyboard to keyboard that I can't touch type these keys, so I'm used to
having to hunt for them. The same applies to DEL and LINEFEED and BREAK
keys, which wander all over the keyboard and sometimes aren't even present.
Many operating systems, such as DEC and ITS, use DEL as backspace. X4.14
suggests that
3.2.1. In Fig 1, key D 13 produces underline in both the shifted
and unshifted modes, but may optionally produce DEL (delete) in the
shifted mode and underline in the unshifted mode.
in the section on bit-paired keyboards. Terminals that do this are quite
painful to use on such systems. Fortunately, this practice has been dying
out. Unfortunately, PC keyboards usually don't have DEL or BREAK keys on
them at all. Xenix on a PC uses the DELETE (as in DELETE CHAR) key to
generate DEL, but DOS doesn't, and this practice hasn't caught on.
Here is the minimum "sane keyboard" layout required to be able to touch
type on a keyboard:
Horton's Minimum Sane Keyboard
------------------------------------------------------------------
| ESC| ! | @ | # | $ | % | ^ | & | * | ( | ) | _ | + |
| | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 0 | - | = |
--------------------------------------------------------------------
| TAB | Q | W | E | R | T | Y | U | I | O | P |
| | | | | | | | | | | |
--------------------------------------------------------------------------|
|CTRL| ** | A | S | D | F | G | H | J | K | L | : | " | RETURN |
| | | | | | | | | | | | ; | ' | |
--------------------------------------------------------------------------|
| SHIFT | Z | X | C | V | B | N | M | < | > | ? | SHIFT |
| | | | | | | | | , | . | / | |
-----------------------------------------------------------------------|
| SPACE |
| |
----------------------------------------------
** This position may contain either CAPS LOCK or the right half of a
wide CTRL key.
It is almost, but not quite, possible to conform to X4.14 and still have
a sane keyboard. The problems are the RETURN and BACKSPACE keys, which
X4.14 would have you move one more space to the right. To conform, a
CAPS LOCK key must also go between CTRL and A.
Some examples of widespread keyboards and why they are insane:
IBM PC: The \ | key moves the left shift key over too far, and the
RETURN (enter) key is one space to far to the right. Keytronics
has done nicely by picking up the pieces of this mess, even though
their CAPS LOCK key is awfully large.
IBM AT: They fixed the SHIFT and RETURN problems, producing mostly
very nice keys. But they put ESC in a separate pad off to the right,
and put ` ~ where ESC should be.
IBM RT: CAPS LOCK and CTRL are interchanged. Fortunately, their 4.2BSD
port has an option to reverse them in software.
NEW IBM PC and AT: The latest keyboard from IBM has yet another random
rearrangement of the keys on the left.
AT&T 6300 301: "Bug for bug" compatible with the IBM PC - left shift
and RETURN are in the wrong places.
AT&T 6300 PLUS 302: CAPS LOCK and CTRL are interchanged. AT&T seems to be
doing this on many new keyboards lately.
HP 2621: Many felt this was HP's best keyboard. CAPS LOCK was to the
left of CTRL, otherwise this was a sane keyboard. (Putting DEL on shift
ESC is unusual and slightly annoying, but not really insane.)
HP 150: ESC is to the left of LEFT SHIFT, and DEL is shift ESC. Also,
CAPS LOCK is to the left of CTRL. HP seems to be using this keyboard
on all new products.
You'll notice one trend: putting CAPS LOCK to the left of A and CTRL
to the left of SPACE. This is especially alarming, since it makes
25% of the ASCII character set unusable. (Ever try to quickly lunge
for CTRL S on such a keyboard? It's a two handed operation.) AT&T's
Teletype division tried this a few years back, but Bell Labs spoke out,
and Teletype fixed the problem. The 5425 and 5620 keyboards now have
an excellent layout. (If only they could reproduce the feel of the 302
while keeping this layout.)
A sane keyboard such as the above will meet with general acceptance from
the current user community, and allow them to touch type. It is also a
useful property for a vendor to claim their existing products conform to.
But it prevents them from gaining the degree of proficiency they should
have, because some of the extra keys aren't specified. This lack of
specification in the ANSI standards is precisely what got us into trouble
in the first place. So let's finish the job and create a complete standard
keyboard arrangement.
Horton's Standard Keyboard
------------------------------------------------------------------------++++++
| ESC| ! | @ | # | $ | % | ^ | & | * | ( | ) | _ | + | BACK + DEL+
| | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 0 | - | = | SPACE+ +
------------------------------------------------------------------------++++++
| TAB | Q | W | E | R | T | Y | U | I | O | P | { | } | | | ~ |
| | | | | | | | | | | | [ | ] | \ | ` |
------------------------------------------------------------------------++++++
| CTRL | A | S | D | F | G | H | J | K | L | : | " | RETURN +LINE+
| | | | | | | | | | | ; | ' | +FEED+
------------------------------------------------------------------------++++++
| SHIFT | Z | X | C | V | B | N | M | < | > | ? | SHIFT +BRK +
| | | | | | | | | , | . | / | + +
------------------------------------------------------------------------++++++
|CAPS| SPACE |
|LOCK| |
---------------------------------------------
This keyboard is sane, it uses the popular convention of [ ] unshifted and
side-by-side, with { } shifted, and it's complete. It puts CAPS LOCK out
of the way, yet easily reachable. (CAPS LOCK isn't a key that is used a
lot, even by secretaries who make up the largest demand for IBM Selectric
style keyboards.)
The only problem here is that I had to make up placements for the \ | and
` ~ keys, which seem to be different on every keyboard, and for DEL, LINE
FEED, and BREAK, which are different or missing on every keyboard. Perhaps
one of the readers can do something better with these keys. I've put the
printing characters in the arrangement used by the AT&T/Teletype 4425/5425,
which seems as good as any, and put the nonprinting ones in +'s to indicate
that they are optional.
Comments and suggested improvements are welcome. Vendors are especially
encouraged to follow this discussion. Input from international sources as
to what can be done to easily internationalize this is welcome. Many of
the opinions in here represent my own experience, although I believe that
much of it applies to the community at large. I would like this to result
in a standard layout which is useful to all of the community, of course, not
just me.gemini@homxb.UUCP (Rick Richardson) (05/02/86)
So far, the best keyboard I have used is a modified IBM PC/AT keyboard.
I move the ESC, tilda, and backspace keys by rotating them to the
right (so ESC ends up to the left of 1, tilda is to right of backslash,
and backspace is to right of tilda). I use software to adjust the
BIOS and drivers for the new placement. This gives me a keyboard
with a great feel, a huge return key, control and escape in the right
places. The backspace is still a little too far in right field, and the
function keys are still off in left, but there's nothing you can do to
fix that.
I believe, however, that no matter what layout you come up with as
"standard", there is somebody out there who needs something different.
Not to mention differences in keyboard feel. (Everytime I get a wet
noodle handshake I can't help but wonder if this person was the inventor
of mashed potatoe tactile feedback).
Don't we all have our own favorite keyboards? Wouldn't the world be
a better place if I could get *MY* keyboard to plug into any computer.
Don't bother standardizing the layout. Standardize the (serial) interface.
Have keyboard, will travel...
Rick Richardson, PC Research, Inc. (201) 922-1134, (201) 834-1378 @ AT&T-CP
..!ihnp4!castor!{rer,pcrat!rer} <--Replies to here, not to homxb!gemini, please.mrl@oddjob.UUCP (Scott R. Anderson) (05/03/86)
In article <2071@cbosgd.UUCP> mark@cbosgd.UUCP (Mark Horton) writes: >Are you tired of incompatible keyboards that require you to stand >on your head to type, or to use two hands to type Control S? I am. Me too! > [The Concept 100 ] puts the CTRL key to the left of CAPS LOCK. The more > recent Concept AVT swapped CTRL and CAPS LOCK, so it doesn't conform > any more. Our Concept AVT does conform, so it's either been changed for the better or the worse, depending on which of ours is more recent. > (6) The strange ASCII punctuation ( [] {} \| `~ ) varies so much I strongly object to the practice of pairing [] and {}, as opposed to [{ and ]}, which is something that I have never quite gotten used to. The latter follows in a long tradition of parentheses () being on the same keyboard level. > Some examples of widespread keyboards and why they are insane: On the Macintosh, DEL is SHIFT-BS. The CTRL key (referred to as COMMAND) is directly to the left of the space bar. The `~ key is to the left of 1, and the ESC key is nonexistant (some programs, such as terminal emulators, will map `~ into ESC). Yecch. And on the new (Mac Plus) keyboard, the right shift key is small and hard to hit. However, the Mac does have an OPTION (i.e. META) key, which I believe is an up-and-coming feature (it is placed to the left of CTRL). The vt200 has a ",," and ".." key, as on many typewriters, and puts <> on a key between Z and SHIFT; as you might expect, this is a real pain, and no one around here likes it. Also, ESC, BS, and LF only exist when the vt200 is in vt100 mode, and then they are up amongst the function keys (`~ is again where ESC should be). >Comments and suggested improvements are welcome. In the best tradition of ANSI standards, I would like to include in my preferred keyboard a feature that is not now available in any keyboard that I know of. What I think would really be nice is if there were an additional CTRL key on the *right* side of the keyboard, for the same reason that there are two SHIFT keys. This would be a boon to people who use Emacs and other editors that make great use of control sequences. However, this would necessitate moving the RETURN key, in spite of Mark's requirement: > (3) The return key should be as large as possible, with exactly two > keys between it and L. It couldn't be moved too far, so the keyboard might look something like this: -------------------------------------------------------------------------------> | ESC| ! | @ | # | $ | % | ^ | & | * | ( | ) | _ | + | | |BACK|BRK> | | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 0 | - | = | \ |SPAC| > -------------------------------------------------------------------------------> | TAB | Q | W | E | R | T | Y | U | I | O | P | { | } | DEL| R > | | | | | | | | | | | | [ | ] | | E > -------------------------------------------------------------------------| T > <CTRL| ~ | A | S | D | F | G | H | J | K | L | : | " | CTRL | R > < | ` | @@ | @@ | @@ | @@ | | | @@ | @@ | @@ | ;@@| ' | | N > <------------------------------------------------------------------------------> <CAP| SHIFT | Z | X | C | V | B | N | M | < | > | ? | SHIFT | LF | <LCK| | | | | | | | | , | . | / | | | ------------------------------------------------------------------------------ | META | SPACE | META | | | | | ------------------------------------------------------------- I don't think that the RETURN is much more difficult to reach here than it is in the standard position. This is the same position it has on the vt200 keyboard, and is similar to the position on the Sun and Concept AVT. It also can be as large as a 2x2 square, as Mark suggested and as I have indicated. The BS key has ended back where it usually is; an alternative is to rotate BS, DEL and LF by one key in a counter clockwise direction. This would put DEL close to where it is on Televideo keyboards and make LF less accessible (I don't think it is used too much anyway). I moved the `~ key between CTRL and A; this is the image of '" on the opposite side, and fills the standard CAPS LOCK position. This key is probably the most likely of any to float around, so this should not be a major change, even though I have never seen it in this position. I also included the META key in a position that is close to where it is often put (Sun, Macintosh, etc.). Again, there are two for both left- and right-handed use. So, are you listening, ANSI? -- * * * * * * * Scott Anderson * * ** ihnp4!oddjob!kaos!sra * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
gwyn@brl-smoke.ARPA (Doug Gwyn ) (05/03/86)
Gee, my 5620 keyboard is a "Horton Standard Keyboard", except for the placement of the BREAK key, which is well to the side so that it can't be bumped into by accident. The only funny feature is having to type CTRL plus COMMA to generate a NUL, instead of CTRL plus SPACE. I have to agree that this is one of the most comfortable keyboard arrangements I've encountered, if you discount the misfeature of having a QWERTY arrangement instead of Dvorak. (It should be at least user-selectable as on the Apple //c.) The absolutely WORST keyboards I've had to use have been touted as "ergonomic", such as the DEC VT220 humongous flat keyboard with control functions all misplaced. IBM PCs come a close next-to-last. Why do the biggest companies make the stupidest keyboards? Is it because they use correspondence typists rather than computists for user feedback? In any event, we need to identify the source of the problem before we can expect to get it fixed. Would some terminal vendors please explain how they came up with their keyboard layout designs?
ccrdave@ucdavis.UUCP (Lord Kahless @ Imperial Propoganda) (05/04/86)
> The absolutely WORST keyboards I've had to use have been touted as > "ergonomic", such as the DEC VT220 humongous flat keyboard with > control functions all misplaced. I *LIKE* the 220, better than any other keyboard I've used. You get used to the layout quickly, if you give it a chance. You naturally start looking for the <> brackets on the left, and, soon enough, you know to use your right hand to press ESC. If your hands are of reasonable size, you don't even have to remove your hand from "home" position on the keyboard. The same applies for ^h and ^j. By the way, I am NOT a DEC employee, nor do I have any interest in the company, other than the fact that I *LOVE* the 220.
kwh@bentley.UUCP (KW Heuer) (05/05/86)
In article <2071@cbosgd.UUCP> cbosgd!mark writes: >Note that the popular practice of putting [ and ] unshifted and { and } >shifted to the right of P is forbidden [in ANSI X4.14]. I've never understood that stupid convention of putting [ and ] on the same key. They are used the same amount, after all (except in LISP). Can you imagine how hard it would be to type on a keyboard where ) was a shifted ( ? >(1) Regular touch typing of letters, typewriter style, should work. This >means the QWERTY arrangement of letters ... Dvorak arrangement would be better, of course, if enough people used it. I wish IBM had introduced it with the Selectric, since they were changing the non-alphabetic keys anyway. (I know that some older typewriters had the bit paired arrangement, but I'm not sure if the Selectric was the first of the modern style.) >(2) The CTRL, TAB, and ESC keys are very heavily used ... TAB must go >directly to the left of Q, and ESC directly to the left of 1. I once used a terminal that had BACKTAB to the left of 1. The first time I ran TECO on it, I couldn't find ESC. No problem, I figured, I'll just use CTRL-[ instead. Unfortunately that also generated a BACKTAB (^Y) because of the way the keyboard was wired! >(4) The numbers 1-0 should be in the standard places. The shifted numbers >should be !@#$%^&*(), typewriter style. (Bit paired keyboards, where ( and >) are shift 8 and 9, are very difficult to use.) It just depends on what you're accustomed to, I think. I started with bit paired terminals, and had lots of trouble when I started using typewriter style keyboards. If all keyboards were alike, it wouldn't much matter where the special characters are (as long as all forms of bracket, () [] {} <>, are on adjacent keys), or even if they moved 0 to the left of 1 (which would make more sense anyway). >Let's finish the job and create a complete standard keyboard arrangement. [picture deleted] I mostly like it. I'm a little worried about BREAK; that's often a dangerous key and should perhaps be further from RETURN and LINEFEED. Or else make it a two-key operation. >[This keyboard] puts CAPS LOCK out of the way, yet easily reachable. (CAPS >LOCK isn't a key that is used a lot, even by secretaries who make up the >largest demand for IBM Selectric style keyboards.) I use SHIFT LOCK (or CAPS LOCK) so infrequently that I wouldn't mind if it were moved to a side keypad, or removed entirely. If both it and CTRL are to the left of A, in either order, people will have trouble with it. Next the space bar isn't too bad, though I've still managed to hit it by mistake. Here's a thought. How about a keyboard that toggles CAPS LOCK when the user hits both shift keys together? (And while I'm thinking of it: why doesn't somebody make a keyboard without explicit function keys, but where CTRL-1 thru CTRL-0 generate an appropriate escape sequence?) >Comments and suggested improvements are welcome. You neglected to specify the semantics of the CTRL key. The ASCII character set has 32 control characters (33 if you count DEL), but only 26 letters. One of the others (ESC) has its own key. On a keyboard where ] and \ are unshifted, I'd expect CTRL to generate ^] and ^\ in the obvious way. That leaves three unspecified: ^@ (NUL), ^^ (RS), and ^_ (US). How do you propose to generate these? Some implementations include: [0] You can't generate them. ("None of our programs require those control characters, so they can't be important.") Clearly a bad choice. [1] Since "@" "^" "_" are the shifts of "2" "6" "-", use CTRL-SHIFT with those keys. This is mnemonic, but slightly awkward. [2] Same except SHIFT not required. [3] Other related keys. CTRL-SPACE for ^@, CTRL-~ for ^^, CTRL-? for ^_ make sense if you think of CTRL as "0x1f &"; I believe this is what the VT100 uses. (Because of which there's an annoying inconsistency between the TTY 5420 and the "VT100 compatible" TTY 5425). The TTY 5620 uses ",<" ".>" and "/?"; the latter two make sense if you think in terms of the shifted forms, but the ",<" key must have been chosen for position. [4] Other arbitrary keys. I think it was the IBM 3101 that was so brain- damaged that we had to use special labels in order to use the local screen editor. (They also put a "<>" key between Z and left SHIFT.) I think this should be part of the standard. (Sigh, this was so *simple* on the bit paired keyboards.)
cmt@myrias.UUCP (Chris Thomson) (05/05/86)
In <2071@cbosgd.UUCP> Mark Horton writes: > Are you tired of incompatible keyboards ... ? You bet I am. There seems to be a rampant opinion that "since engineers and casual computer users can't touch type anyway, keyboard layout is arbitrary". Grrr. When I went through high school a typing course was mandatory. I think it still should be. One pet peeve I have (that Mark missed) is that most new keyboards do not set off the home key positions (or at least the F and J keys), a practice that I find particularly upsetting, since it makes touch typing unreliable, and is a zero-cost thing to fix. Responding to Mark's suggested keyboard (reproduced below), I have a few comments: 1. Overall, I like it. I'd be happy to have one of these on my desk. 2. No meta key is specified. It should be optional, and below the left shift key. This is the ALT key on IBM PC's, the S2 key on Ann Arbor XL's, and the LEFT key on Sun-2's. 3. The placement of caps lock could just as well be as on the Sun-2, where it is under the left part of the left shift key. 4. While I hold the opposite view to Mark regarding the relative placement of caps lock and control, his double-wide control key satisfies both points of view. > Horton's Standard Keyboard ------------------------------------------------------------------------++++++ | ESC| ! | @ | # | $ | % | ^ | & | * | ( | ) | _ | + | BACK + DEL+ | | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 0 | - | = | SPACE+ + ------------------------------------------------------------------------++++++ | TAB | Q | W | E | R | T | Y | U | I | O | P | { | } | | | ~ | | | | | | | | | | | | | [ | ] | \ | ` | ------------------------------------------------------------------------++++++ | CTRL | A | S | D | F | G | H | J | K | L | : | " | RETURN +LINE+ | | | | | | | | | | | ; | ' | +FEED+ ------------------------------------------------------------------------++++++ | SHIFT | Z | X | C | V | B | N | M | < | > | ? | SHIFT +BRK + | | | | | | | | | , | . | / | + + ------------------------------------------------------------------------++++++ |CAPS| SPACE | |LOCK| | --------------------------------------------- -- Chris Thomson, Myrias Research Corporation ihnp4!alberta!myrias!cmt 200 10328 81 Ave, Edmonton Alberta, Canada 403 432 1616
dave@uwvax.UUCP (Dave Cohrs) (05/05/86)
Wow! The Horton Standard Keyboard is almost the same as my uvaxII keyboard.
Remapping the keys in software and a magic marker do wonders...
A few comments on the Horton keyboard
1) I like my RETURN key big. Really big. Putting RETURN and LF next
to each other helps, but I still like the RETURN a double-height key.
2) The CAPS LOCK could be moved over so it's harder to hit. To the left
of the left SHIFT key, perhaps? The best thing to do with a CAPS LOCK
key is remove it.
3) The BRK key should also be moved out of the way. Perhaps to the right
of DEL...
4) Some people like a META key. This is usually under the SHIFT or 'Z'
keys. Perhaps this should be next the SPACE bar.
Other than that, I find the layout a good one. Now, if we could only
get the "touch" of a keyboard standardized...
--
Dave Cohrs
(608) 262-1204
...!{harvard,ihnp4,seismo,topaz}!uwvax!dave
dave@rsch.wisc.edulamy@utai.UUCP (Francois Lamy) (05/05/86)
In article <245@myrias.UUCP> cmt@myrias.UUCP (Chris Thomson) writes: > 2. No meta key is specified. It should be optional, and below the left > shift key. This is the ALT key on IBM PC's, the S2 key on Ann Arbor > XL's, and the LEFT key on Sun-2's. Meta keys are nice. Agreed. This being net.internat, and being stuck quite often with my native tongue, I think that a "compose character" key MUST be present, and should sit under the <shift>. Ideally, that key should be "hardwired", and distinct from a "meta" key. Why? So that the first unilingual programmer around does not get tempted in using the key for something else... How are other ISO Latin 1 characters entered? I'd prefer only one strike, using "compose" as a modifier key. <compose-[> might yield "e'" on a french-canadian "programmer's layout". A "secretary's layout" with "e'" on the keyboard might still be useable by a programmer if <compose-e'> yielded a "[". This is not a trivial issue. Programmers won't use "secretarial layouts" if getting "[" requires 2**n key strokes, n -> infinity. National keyboards are a necessity. I believe that there should be a standard keyboard interface standard, so that you could buy a keyboard with your favorite layout, and hook it to your favorite computer/terminal. This presumes that information like modifier keys depressed be transmitted as well > 3. The placement of caps lock could just as well be as on the Sun-2, > where it is under the left part of the left shift key. I personally despise <caps lock> and see no need for it to be reachable from the home keys (how often is it used in a normal session? even once?). > 4. While I hold the opposite view to Mark regarding the relative placement > of caps lock and control, his double-wide control key satisfies both > points of view. I prefer a full sized control key. What follows is extracted from a Macintosh layout, with Control replacing "caps lock", "compose" replacing "clover", and "meta" replacing "option" >------------------------------------------------------------------------+ >| CTRL | A | S | D | F | G | H | J | K | L | : | " | RETURN + >| | | | | | | | | | | ; | ' | + >------------------------------------------------------------------------+ >| SHIFT | Z | X | C | V | B | N | M | < | > | ? | SHIFT + >| | | | | | | | | , | . | / | + >------------------------------------------------------------------------+ >| META| COMPOSE| SPACE | >| | | | >-------------------------------------------------------+ -- Jean-Francois Lamy Department of Computer Science, University of Toronto, Departement d'informatique et de recherche operationnelle, U. de Montreal. CSNet: lamy@toronto UUCP: {ihnp4,utzoo,decwrl,uw-beaver}!utcsri!utai!lamy EAN: lamy@ai.toronto.cdn ARPA: lamy%toronto@csnet-relay
lav@mtsbb.UUCP (L.A.VALLONE) (05/05/86)
I'm glad that someone has finally voiced the disgust I have felt over the past few years. In the last 6 months I have continuously swapped between : an hp 2621, an Ann Arbor Ambassador, an AT&T PC 6300, an IBM PC and an AT&T UNIX PC. I used to consider myself a reasonably proficient touch typist (approx. 80 wpm text) but recently my fingers have gotten so confused that even my typing of basic text has slowed by at least 10 wpm. The above degradation is probably not all due to "finger confusion". Like Mark Horton, I long for the feel of the old Selectric keyboard. The streamlined feel of the "new" electronic keyboards just doesn't cut it. Of the above, I found the hp 2621 keyboard the closest to the Selectric and the least tiring. Alas, I am down to using only the IBM-PC and the AT&T UNIX PC. This combination has only been made bearable by using Paul Fox's installable device driver for swapping the ctrl and CAPS-LOCK keys on the UNIX PC. I loath the return keys on both, but the placement of the Enter key on the UNIX PC definitely helps. I join in Mark's plea for a sane keyboard arrangement and hope it comes before my fingers are committed. (-: Following are some additional comments on Mark's suggestions: > > Horton's Minimum Sane Keyboard > ------------------------------------------------------------------ > | ESC| ! | @ | # | $ | % | ^ | & | * | ( | ) | _ | + | > | | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 0 | - | = | > -------------------------------------------------------------------- > | TAB | Q | W | E | R | T | Y | U | I | O | P | > | | | | | | | | | | | | > --------------------------------------------------------------------------| > |CTRL| ** | A | S | D | F | G | H | J | K | L | : | " | RETURN | > | | | | | | | | | | | | ; | ' | | > --------------------------------------------------------------------------| > | SHIFT | Z | X | C | V | B | N | M | < | > | ? | SHIFT | > | | | | | | | | | , | . | / | | > -----------------------------------------------------------------------| > | SPACE | > | | > ---------------------------------------------- > ** This position may contain either CAPS LOCK or the right half of a > wide CTRL key. > > It is almost, but not quite, possible to conform to X4.14 and still have > a sane keyboard. The problems are the RETURN and BACKSPACE keys, which > X4.14 would have you move one more space to the right. To conform, a > CAPS LOCK key must also go between CTRL and A. > I prefer NOT to have the CAPS LOCK key between the 'A' and 'CTRL' keys. I rarely ever use the CAPS LOCK key (typical, as you mentioned) and find it especially dangerous for vi users (j to go down, J joins - multiple inadvertant J's cost dearly). > > Horton's Standard Keyboard > ------------------------------------------------------------------------++++++ > | ESC| ! | @ | # | $ | % | ^ | & | * | ( | ) | _ | + | BACK + DEL+ > | | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 0 | - | = | SPACE+ + > ------------------------------------------------------------------------++++++ > | TAB | Q | W | E | R | T | Y | U | I | O | P | { | } | | | ~ | > | | | | | | | | | | | | [ | ] | \ | ` | > ------------------------------------------------------------------------++++++ > | CTRL | A | S | D | F | G | H | J | K | L | : | " | RETURN +LINE+ > | | | | | | | | | | | ; | ' | +FEED+ > ------------------------------------------------------------------------++++++ > | SHIFT | Z | X | C | V | B | N | M | < | > | ? | SHIFT +BRK + > | | | | | | | | | , | . | / | + + > ------------------------------------------------------------------------++++++ > |CAPS| SPACE | > |LOCK| | > --------------------------------------------- > > This keyboard is sane, it uses the popular convention of [ ] unshifted and > side-by-side, with { } shifted, and it's complete. It puts CAPS LOCK out > of the way, yet easily reachable. (CAPS LOCK isn't a key that is used a > lot, even by secretaries who make up the largest demand for IBM Selectric > style keyboards.) > > The only problem here is that I had to make up placements for the \ | and > ` ~ keys, which seem to be different on every keyboard, and for DEL, LINE > FEED, and BREAK, which are different or missing on every keyboard. Perhaps > one of the readers can do something better with these keys. I've put the > printing characters in the arrangement used by the AT&T/Teletype 4425/5425, > which seems as good as any, and put the nonprinting ones in +'s to indicate > that they are optional. > > Comments and suggested improvements are welcome. Vendors are especially > encouraged to follow this discussion. Input from international sources as > to what can be done to easily internationalize this is welcome. Many of > the opinions in here represent my own experience, although I believe that > much of it applies to the community at large. I would like this to result > in a standard layout which is useful to all of the community, of course, not > just me. I have no great feeling with the placement of the \ | ` ~ DEL LINE FEED and BREAK keys. Although, it seems to me that ~ gets a lot more use than ` and might be better if unshifted (even though I have never seen a single keyboard with ~ unshifted). -- Lee Vallone AT&T Information Systems Merlin {... ihnp4, mtuxo}!mtsbb!lav
sas1@sphinx.UChicago.UUCP (Stuart Schmukler) (05/05/86)
In article <2071@cbosgd.UUCP> mark@cbosgd.UUCP (Mark Horton) writes: >Are you tired of incompatible keyboards that require you to stand >on your head to type, or to use two hands to type Control S? I am. Yes I am tried of tring to use incompatible keyboards, but the manufacturers seem to to rely on the pliablity of computer users. None of the manufacturers I have spoken with seem to be able to (or are willing to) justify their positioning of keys on a keyboard. In article <1314@oddjob.UUCP> mrl@oddjob.UUCP (Scott R. Anderson) writes: >The vt200 has a ",," and ".." key, as on many typewriters, and >puts <> on a key between Z and SHIFT; as you might expect, this >is a real pain, and no one around here likes it. I have even seen some one from DEC strongly defend this positioning in a discussion of standard keyboards for the APL community. In general that keyboard seems to be motivated by an attempt to court the secretarial user and the europeans. My biggest problem is with the right-handed designers of numeric keypads. Grafting the keypad to the right side of the keyboard seems to put 20% (I think) of the population disadvantage. Putting the keypad on a separate attachable panel seems to be a good solution for both left and right handers. >>Comments and suggested improvements are welcome. (I wish that manufacturers would put this on their keyboards:).) SaS
lee@cyb-eng.UUCP (Bill Lee) (05/05/86)
> Are you tired of incompatible keyboards that require you to stand > on your head to type, or to use two hands to type Control S? I am. > I've watched the most recent array of keyboards on IBM PC's and > AT&T PC's, and instead of converging to a common standard, the new > keyboards are diverging wildly and recklessly. I'd like to propose > a standard to solve this problem. > I agree 100% with Mark. Keyboards are seemingly being designed by people that haven't ever used a real computer, and certainly not a Unix system. I just had the misfortune of ordering a Wyse 85 terminal (the one that does VT220 emulation). Since I was familiar with the WY-50 and the WY-75, I assumed I was safe ordering a WY-85 sight unseen. Big mistake. The keyboard is a complete mess. To save any poor souls considering purchasing a WY-85, let me mention a bit of the keyboard braindamage. To start with, there is no DEL key. The correction key (marked by the large X) can be programmed to send a two character sequence that begins with DEL. I'm making do with that for now. The ESC key is up on the top row of function keys (it is shifted F11). Try that it vi or your favorite screen editor. In the normal ESC position (left of 1) is "~/`" which makes using ~ a real pain. One of the biggest loses is the ">/<" key (yes, they are on the same key). It is to the left of the Z, right next to SHIFT. I/O redirection is very easy to screw up with this arrangement. In summary, I do not recommend the WY-85 to anyone since the keyboard is so wrong. Mark's suggested keyboard looks right on to me. Everyone of the keys is in the "right" place. I hope his posting finds its way to terminal vendors and particular to their design engineers and human factors people (if terminal vendors have human factors). In fact, I'm going to print out his posting and mail it to WYSE along with a letter explaining how they went wrong with the Wy-85 keyboard. P.S. To be able to "use" the WY-85, you have to order the "WY-85 Programmers Guide" for $40.00. Escape sequences, how to use the "Compose Character" key, how to get the information needed to write your own termcap (since the VT100 termcap I'm using doesn't always work with vi), etc. require the $40 manual. Bill Lee Edge Computer Corp. Austin, Texas (512) 499-1586 mot!edge!lee
bobr@zeus.UUCP (Robert Reed) (05/05/86)
Tektronix has had a keyboard standard which has been employed through the 402X, 411X, 410X, and 412X families of terminals (more or less) and which I find quite comfortable for touch typing and using editors which require the use of control and escape keys: ---------------------------------------------------------------------------- | { | ! | @ | # | $ | % | ^ | & | * | ( | ) | _ | + | } | DEL| | [ | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 0 | - | = | ] | | ++++--------------------------------------------------------------------------- +ESC | ~ | Q | W | E | R | T | Y | U | I | O | P | ` |BACK |LINE| + | | | | | | | | | | | | | \ |SPACE |FEED| +++++-------------------------------------------------------------------------- +TAB | CTRL | A | S | D | F | G | H | J | K | L | : | " | RETURN | + | | | | | | | | | | | ; | ' | | ------------------------------------------------------------------------------- |CAPS| SHIFT | Z | X | C | V | B | N | M | < | > | ? |SHIFT |BRK | |LOCK| | | | | | | | | , | . | / | | | ------------------------------------------------------------------------------ | SPACE | | | ----------------------------------------- Mark Horton's suggested layout suffers from an imbalance, placing most of the keys under the right hand. From the normal resting place of ASDF and JKL; (as can be seen below), the left hand has one horizontal extension to reach the edge of the keyboard whereas the right hand must extend three keys from the home row to reach the edge. The keyboard above balances this at two and two. Caps lock does not obscure the space bar, but is not intervening between CTRL and A either. Both TAB and ESC are closer to the little finger, since it's easier to do a horizontal extension than a vertical one (these keys are 1 and a half wide, as indicated by the plusses). The {[ and }] keys have been moved to the top row so that the second row of special keys can be brought in closer. Backspace is moved to a more accessible location, at the expense of pushing }] to the limits of reach. With this arrangement, the entire keyboard can be learned by touch, although admittedly, the extra characters required make the task of the little fingers prone to error. ------------------------------------------------------------------------++++++ | ESC| ! | @ | # | $ | % | ^ | & | * | ( | ) | _ | + | BACK + DEL+ | | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 0 | - | = | SPACE+ + ------------------------------------------------------------------------++++++ | TAB | Q | W | E | R | T | Y | U | I | O | P | { | } | | | ~ | | | | | | | | | | | | | [ | ] | \ | ` | ------------------------------------------------------------------------++++++ | CTRL | A | S | D | F | G | H | J | K | L | : | " | RETURN +LINE+ | | | | | | | | | | | ; | ' | +FEED+ ------------------------------------------------------------------------++++++ | SHIFT | Z | X | C | V | B | N | M | < | > | ? | SHIFT +BRK + | | | | | | | | | , | . | / | + + ------------------------------------------------------------------------++++++ |CAPS| SPACE | |LOCK| | ---------------------------------------------
mrl@oddjob.UUCP (Scott R. Anderson) (05/06/86)
In article <132@sphinx.UChicago.UUCP> sas1@sphinx.UUCP (Stuart Schmukler) writes: >In article <1314@oddjob.UUCP> mrl@oddjob.UUCP (Scott R. Anderson) writes: >>The vt200 has ",," and ".." keys, as on many typewriters, and >>puts <> on a key between Z and SHIFT; as you might expect, this >>is a real pain, and no one around here likes it. > >I have even seen some one from DEC strongly defend this positioning in a >discussion of standard keyboards for the APL community. In general that >keyboard seems to be motivated by an attempt to court the secretarial user and >the europeans. It would seem so, except that if they were trying to exactly reproduce the standard typewriter keyboard, they did a lousy job of key positioning, what with the left shift key shifted left and the return key shifted right. -- * * * * * * * Scott Anderson * * ** ihnp4!oddjob!kaos!sra * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * *
faunt@spar.UUCP (05/06/86)
Where do the META keys go? I want my bucky bits!
jr@inset.UUCP (Jim R Oldroyd) (05/06/86)
I quite like Mark's final Standard Layout, but make the following
points:
- The location of the DEL key is very important. This is because it
is quite common over in Europe to use DEL for <delete> and
BS for <backspace> which are quite different concepts. (Consider
line editing, for example). However, Mark's choice of position is
acceptable. HP's use of SHIFT-something is NOT!
- The positioning of keys like \ | ` ~ [ ] { } is a major problem.
Mark admits to placing some of these at his own discretion.
But, in many countries, these characters are not `back', `pipe',
`subs', `tilde', brackets or braces; but appear as `c-cidilla',
`e-acute', `angstrom' etc. The location of such keys is, of course,
defined in the typewriter layouts adopted by these countries.
[Sorry can't quote any refs here.] As an example, in Denmark,
the keys `angstrom', `o/' and `ae' are arranged thus:
--------------------------
| P | ang | + | |
| | | ; | |
-------------------| <CR>|
| L | ae | o/ | |
| | | | |
------------------------------
[Note also the size of the <CR>.]
In Norway, they have the SAME three characters. But the `standard'
layout is:
--------------------------
| P | ang | + | |
| | | ; | |
-------------------| <CR>|
| L | o/ | ae | |
| | | | |
------------------------------
- Then of course, there are major problems with Mark's proposals
in countries where QWERTY is not the standard layout for the
alphabetic characters. In France, AZERTY is used; in parts
of Germany, QWERTZ is popular.
One of the problems facing the designer of a keyboard is not only
how to position the keys of the US ASCII set, but also how to
locate them in such a way as to be convienient for users of non-US
ASCII keyboards. For the user of such keyboards who ALSO has
to work with several types of such keyboards, the time spent
hunting for keys is ridiculous.
Currently, two solutions to this problem seem to be in use.
1. Provide an `ASCII' keyboard with the local characters where
they'll fit.
2. Provide a local keyboard.
The first solution has the advantage that it is cheaper, more
flexible and the `international' user can quickly locate the
majority of the keys.
The second solution is preferable to the End-Users of Applications
and also to a large number of Software Developers who (rightly)
ask "Why should WE use a US keyboard?".
Jim.
PS: I understand that in French secretarial colleges, the AZERTY
layout is being phased out. Could someone add more input on
this?
PPS: What about the Dijkstra(?) keyboard? I understand there's now
a version for use with computer systems.
--
"The Software did it".
--
++ Jim R Oldroyd
++ jr@inset.co.uk
++ ..!mcvax!ukc!inset!jrkwh@bentley.UUCP (KW Heuer) (05/06/86)
In article <859@uwvax.UUCP> uwvax!dave (Dave Cohrs) writes: >A few comments on the Horton keyboard ... >3) The BRK key should also be moved out of the way. Perhaps to the right > of DEL... That's fine if you're assuming a UNIX* system where BRK and DEL both send a SIGINT, but it would be a major nuisance on a "DEL is erase" system. Karl W. Z. Heuer (ihnp4!bentley!kwh), The Walking Lint
chris@columbia.UUCP (Chris Maio) (05/06/86)
There's nothing wrong with the Horton Standard Keyboard, but I would suggest adopting the (similar) DEC VT100 keyboard as a standard instead, with the option widening the control key and moving caps lock out of the way. Among other things, a significant number of the keyboards out there would already be compliant. >In any event, we need to identify the source of the problem before >we can expect to get it fixed. Would some terminal vendors please >explain how they came up with their keyboard layout designs? I think Human Designed Systems, makers of the old Concept 100's and AVTs, does it right. After I ordered a couple of batches of AVTs with non-standard keyboards, the president of the company himself called me up to ask what I didn't like about their keyboards. I didn't ask for anything odd, and sure enough the next keyboard they brought out, for the HDS200, was exactly what I wanted -- basically a VT100 layout but with a wide control key, caps lock under linefeed, meta-key under left-shift, plus the keypad and dozens of pf keys on an "ergonmetric" keyboard with a good feel. If you're picky, you can redefine keys like BS, DEL, LF, CR, etc to generate whatever you want (you can even download your own font if you don't like the standard one, or if you want, say, italics in standout mode). I won't go into the many other great features of the HDS200, but I would definitely recommend it to anyone who cares what kind of terminal they use. Chris
dbw@ariel.UUCP (DAVE B. WOOD) (05/06/86)
All this has me curious: what do other non-Unix systems do with async terminals, and what might this mean for the keyboards? The IBM 3270 terminals have a rather elegant architecture wherein the keyboard layout is what the cluster controller (read "computer") agrees to use, and the physical scan code is what is actually sent to the cluster when a key is pressed or released. (How would you like to have a switch to denote your keycap set, and have terminfo stuff actually download your keyboard layout?) Don't get me wrong- for Unix I'll take Mark's keyboard every time. Dave Wood
keyboard@cbosgd.UUCP (05/06/86)
I appreciate the interest in this topic from the community. I'm
saving up comments and will make appropriate modifications. From
what I've seen so far, changes will probably include:
(1) Tightening up the spec a bit, to mention that the caps lock key,
if between CTRL and A, must be single width (not VT100 style),
that the control keys must do the right thing, that the strange
punct keys [ ] \ ` must be somewhere in the upper right, and the like.
(2) Rearranging the right keys a bit, e.g. BREAK at far upper right.
(My preliminary picture grew to past 80 chars wide, sorry.)
(3) Including a third "Horton's Standard Extended ASCII keyboard"
which includes META/ALT, arrows, function keys, arrow keys,
and numeric pad, along with what they generated ala X3.64.
Right now, I'd like to take a survey on some unclear issues. Please
reply by mail for this, your 'r' key will probably generate something
reasonable, but mail to ...!cbosgd!keyboard or keyboard@seismo.CSS.GOV will
work if that fails.
In answering these, please include the codes (e.g. BS-1) with each answer;
this will help me tabulate the responses. One easy way to do this is to
use your R command (from rn) or equivalent option to include the body of
this survey, then delete all inappropriate lines. Or you can write down
the codes (there are 7 questions a-f) on a scrap of paper and then reply
by hand, typing them in on separate lines.
Please try to keep your answers to one of the choices given, unless you
strongly feel a different answer is needed. I will read the responses
for human-readable comments, but I'll use grep -n to count the results.
Thank you in advance for taking a few minutes to respond to this survey.
(a) Backspace key. Please answer ONLY if you use your backspace key
regularly - if you're a CTRL H user, respond that you are. Do you prefer:
BS-0) I don't use my BS key much, I use CTRL H instead.
BS-1) Two keys between BS and zero.
BS-2) Three keys between BS and zero.
BS-3) Either is fine - I use both and it doesn't matter.
BS-4) I don't know - my keyboard has 2 keys between BS/0 and I like it.
BS-5) I don't know - my keyboard has 3 keys between BS/0 and I like it.
(b) CR, Return or Enter key. Do you prefer:
CR-0) I never use this key, I generally use ________ instead.
CR-1) Two keys between L and CR.
CR-2) Three keys between L and CR.
CR-3) Three keys between L and CR, except the IBM PC is too far.
CR-4) Either is fine - I use both and it doesn't matter.
CR-5) I don't know - my keyboard has 2 keys between L/CR and I like it.
CR-6) I don't know - my keyboard has 3 keys between L/CR and I like it.
(c) How big does the Return key need to be?
CRW-0) I never use this key.
CRW-1) Double wide is fine.
CRW-2) Single wide but double high (IBM PC) is better.
CRW-3) L shaped (VT100/Sun) is minimum acceptable.
CRW-4) 2x2 key is minimum acceptable.
(d) How do you want NULL generated? (Note - I have a problem here. I tend
to be a bit slow getting off the CTRL key, and hit CTRL SPACE sometimes
when I wanted a space, and get a NULL I didn't want. This motivates
this question.)
NU-0) I don't care, I never type nulls.
NU-1) I don't care, but I want CTRL-SPACE to generate SPACE.
NU-2) CTRL 2 should be the only way.
NU-3) CTRL SHIFT 2 should be the only way.
NU-4) CTRL SPACE should be the only way.
NU-5) Either CTRL 2 or CTRL SHIFT 2 should work, but not CTRL SPACE.
NU-6) Either CTRL 2, CTRL SHIFT 2, or CTRL SPACE should work.
NU-7) Other (specify.)
(e) When CTRL and CAPS LOCK are both to the left of A, and both keys are
single width, which best describes your experience?
CL-0) I've never used such a terminal.
CL-1) CTRL must be to the left of CAPS LOCK.
CL-2) CAPS LOCK must be to the left of CTRL.
CL-3) I am happy with either arrangement.
CL-4) I hate both arrangements - CTRL should be double-wide here.
CL-5) I hate both arrangements - CAPS LOCK should be double-wide here.
CL-6) I've only used CTRL to left of CAPS LOCK and I like it.
CL-7) I've only used CAPS LOCK to left of CTRL and I like it.
(f) When comparing keyboards with CTRL to the left of a single width CAPS
LOCK key (e.g. Ann Arbor or Concept), to keyboards with CTRL to the
left of a double width CAPS LOCK key (e.g. VT100 or Keytronics) which
do you prefer?
CLW-0) I've never used such a terminal.
CLW-1) CAPS LOCK must be single width
CLW-2) CAPS LOCK must be double width
CLW-3) I am happy with either arrangement.
CLW-4) I hate both arrangements - CAPS LOCK should be elsewhere
CLW-5) I hate both arrangements - CTRL should be elsewhere
CLW-6) I've only used single width CAPS LOCK and I like it.
CLW-7) I've only used double width CAPS LOCK and I like it.
g) Here are some proposed alternate locations for CAPS LOCK. Please bear
in mind that political reality suggests using a location that is commonly
used today, such as to the left of A or on one side of the space bar.
CLL-0) I don't care, I never use it.
CLL-1) I prefer to the left of A.
CLL-2) I prefer under the left shift, e.g. to the left of ALT or META.
CLL-3) I prefer directly to the left of space, to the right of ALT or META.
CLL-4) I prefer directly to the right of space, to the left of a 2nd CTRL.
CLL-5) I prefer to the far right of space, to the right of a 2nd CTRL.
CLL-6) I think CAPS LOCK should go way off in the boonies somewhere.barmar@mit-eddie.MIT.EDU (Barry Margolin) (05/07/86)
In article <824@cyb-eng.UUCP> lee@cyb-eng.UUCP (Bill Lee) writes: >I just had the misfortune of ordering a Wyse 85 terminal (the >one that does VT220 emulation). Since I was familiar with the WY-50 and >the WY-75, I assumed I was safe ordering a WY-85 sight unseen. Big >mistake. The keyboard is a complete mess. ... [description deleted] > In summary, I do not recommend the >WY-85 to anyone since the keyboard is so wrong. The description Bill Lee gave of the WY-85 sounds like what I remember of the VT220. As he says, the WY-85 is supposed to emulate a VT220, so it is hardly Wyse's fault that the keyboard sucks (it is DEC's fault, and partly ANSI's). >P.S. To be able to "use" the WY-85, you have to order the "WY-85 Programmers >Guide" for $40.00. Escape sequences, how to use the "Compose Character" key, >how to get the information needed to write your own termcap (since the VT100 >termcap I'm using doesn't always work with vi), etc. require the $40 manual. I think this is also true of the DEC VT-220. I think the assumption of the terminal manufacturers is that most of the terminal users are not programmers, so they do not need to ship a programmer's manual with every terminal. As for the $40, don't fool yourself -- if they had included the manual with each terminal, the price of the terminal would probably have been at least $20 higher. If you were buying 100 terminals for 98 secretaries and 2 programmers, your savings are at least $2,000 because of the unbundling of the manual; assuming the $20 price hike for the manual, the customer saves money as long as there are more non-programmers than programmers. As the major customers of most terminal manufacturers are offices running packaged software (and do no programming of their own), this is probably a good marketing move (akin to AT&T's unbundling of many programming tools in the Unix(tm) PC). -- Barry Margolin ARPA: barmar@MIT-Multics UUCP: ..!genrad!mit-eddie!barmar
danhart@aluxp.UUCP (HART) (05/07/86)
If a really sane keyboard is impossible and a standard is all that is sought, please try to include symetric pairs of cntl and esc/meta keys. My knuckles are constantly sore from bumping in the control crossovers. -- ________ -------- Dan Hart AT&T:Bell Labs 1247 So. Cedar Crest Blvd. Allentown PA aluxp!aloft!danhart 215-770-3657 My opinions are mine alone, usually.
macrakis@harvard.UUCP (Stavros Macrakis) (05/07/86)
In response to the proposal for keyboard standardization, several people lamented the absence of a Meta key. (See below) Although I certainly sympathize (being an Emacs user from way back), the use of the top bit to transmit additional control information is a bad idea: on the one hand, this bit is currently sometimes used for parity checking; on the other, all extended character set proposals use it -- as should be particularly clear to those on net.internat. Of course, this all started because the original ANSI control characters were being unused for their standard functions. Perhaps the way to go is to standardize on the notion of `bucky bits' (that is, characters modified by keys depressed simultaneously), but to define a protocol within Ascii for transmitting them as multi-character escape sequences (e.g. instead of transmitting 210 for control-meta-i, transmit ESC CMP I, where CMP is an escape code meaning control-meta-prefix). -s cmt@myrias.UUCP (Chris Thomson) > 2. No meta key is specified. It should be optional, and below the left dave@uwvax.UUCP (Dave Cohrs) > 4) Some people like a META key. faunt@spar.UUCP (Doug Faunt) > Where do the META keys go?...I want my bucky bits! mrl@oddjob.UUCP (Scott R. Anderson) >... the Mac does have an OPTION (i.e. META) key, which I believe > is an up-and-coming feature -- -s Stavros Macrakis Macrakis@Harvard.{Harvard.EDU,ARPA,uucp,csnet} Harvard Aiken Lab 111 @Harvunxh.bitnet 33 Oxford Street Cambridge, MA 02138
pete@valid.UUCP (Pete Zakel) (05/07/86)
> One pet peeve I have (that Mark missed) is that most new keyboards do not set > off the home key positions (or at least the F and J keys), a practice that > I find particularly upsetting, since it makes touch typing unreliable, and > is a zero-cost thing to fix. > -- > Chris Thomson Why do almost all keyboards with specially indented home keys put the indents on the F and J keys? This really irritates me! I prefer the indents on the D and K keys because: 1) My middle fingers are longer than the others and 2) My index finger has to hit a lot more keys and it is easier to keep the middle finger on the home keys and orient the other fingers to the middle finger than to always have to find the indent with the index finger and orient the other fingers to that. -- -Pete Zakel (..!{hplabs,amd,pyramid,ihnp4}!pesnta!valid!pete)
sommar@enea.UUCP (Erland Sommarskog) (05/07/86)
I'd like too add some comments om Mark Horton proposal to a keyboard
standard.
Jim R Oldroyd points at an important point, the need for national
layout. MH has e.g "}" as shift "]". On my keyboard they are interchanged,
since "]" is a captial letter and "}" the small version of the same.
Of the same reason I don't have "@" on shift-2 but as shift-"`". The Swedish
standard also contains some ofter differences to the American, bur they are not
important here. (But I like it, and don't like using American keyboards.)
Another thing that anyone hasn't mentioned is the arrow keys. They usually
sit somewhere on separate keypad. Sometimes in row, somtimes in a square.
But how, differs from keyboard to keyboard. E.g. the keyboards I use
the most are those of Facit Twist and Tandberg TDV 2230. (The Facit is
a Swedish terminal, the Tandberg Norwegian. Both of them are very good.)
Let's how they have arranged the arrow keys:
Facit Tandberg
Backtab Up Tab DEL Up CSI
Left Down Right Left Home Right
Backtab Down Tab
Notes:
1) Both keyboards have function keys above these. I never use these keys.
2) Both keyboards also have Tab in its normal place.
3) CSI on the Tandberg generates <ESC>-[
The most confusing is of course that Left/backtab and Right/Tab are inter-
changed between the two keyboards. The fact that Tandberg has an obsolete
"Home" between Up and Down doesn't trouble me that much.
I also like to add that I find this type of arrangement far more logical than
the VT100-like
Up Left Right Down
(With excuses for the order).
Another annoying differnce between the keyboards are the arrangements of the
left shift key and "No scroll"
The Facit has (like VT100 I believe)
NoScroll LeftShift Z with LeftShift 1.5 the size of a normal key.
The Tandberg has
LeftShift </> Z </> means "<" unshifted and ">" shifted
The No-scroll key is placed at far the right , three rows over PF1.
Although the Facit places the shift better than Tandberg, according to Horton,
The placement of NoScroll is stupid.
More on the subject could be said, but I stop with this.
Erland Sommarskog
ENEA Data, Swedenkludge@gitpyr.UUCP (05/07/86)
In article <1138@ariel.UUCP> dbw@ariel.UUCP (DAVE B. WOOD) writes: >The IBM 3270 terminals have a rather elegant architecture >wherein the keyboard layout is what the cluster controller >(read "computer") agrees to use, and the physical scan code >is what is actually sent to the cluster when a key is >pressed or released. (How would you like to have a switch >to denote your keycap set, and have terminfo stuff actually >download your keyboard layout?) I used to have an IBM 3277 terminal with a homebrew protocol converter that allowed it to connect up to ASCII devices. It had the crispest screen, best-feeling keyboard, and it can be configured to whatever layout you desire... I moved the ALT (control) key over to where it belonged, fixed up the tabs, and set up Televideo emulation. The terminal is just too intelligent for its own good. Friends have modified 3274 display controllers so that they can use AZERTY and DSK keyboards. It's a powerful system.... if only it used ASCII... -- ------- Disclaimer: Everything I say is probably a trademark of someone. But don't worry, I probably don't know what I'm talking about. Scott Dorsey " If value corrupts kaptain_kludge then absolute value corrupts absolutely" ICS Programming Lab (Where old terminals go to die), Rich 110, Georgia Institute of Technology, Box 36681, Atlanta, Georgia 30332 ...!{akgua,allegra,amd,hplabs,ihnp4,seismo,ut-ngp}!gatech!gitpyr!kludge
george@mnetor.UUCP (George Hart) (05/08/86)
In article <824@cyb-eng.UUCP> lee@cyb-eng.UUCP (Bill Lee) writes: >I just had the misfortune of ordering a Wyse 85 terminal (the >one that does VT220 emulation). Since I was familiar with the WY-50 and >the WY-75, I assumed I was safe ordering a WY-85 sight unseen. Big >mistake. The keyboard is a complete mess. I really don't think it is that bad, especially considering what it is trying to be compatible with. >To save any poor souls considering >purchasing a WY-85, let me mention a bit of the keyboard braindamage. To >start with, there is no DEL key. The correction key (marked by the large X) >can be programmed to send a two character sequence that begins with DEL. >I'm making do with that for now. Actually, the "X" key only sends DEL. The second character is CANCEL which if you type shifted "X". Trust me, this is being typed on a WY85. You have the optional of making the "X" key send (unshift/shift) DEL/CAN or BS/DEL. Choosable from setup. Also choosable from setup is whether you want the compose key active or not. When I had a VT220, I finally pulled the compose keycap off because I kept hitting it. >The ESC key is up on the top row of function >keys (it is shifted F11). Try that it vi or your favorite screen editor. In >the normal ESC position (left of 1) is "~/`" which makes using ~ a real pain. >One of the biggest loses is the ">/<" key (yes, they are on the same key). >It is to the left of the Z, right next to SHIFT. I/O redirection is very >easy to screw up with this arrangement. In summary, I do not recommend the >WY-85 to anyone since the keyboard is so wrong. By setting the WY85 in to VT200 7 bit mode, you get the function keys plus all the VT100 control modes (which are a subset of VT200). Right now, I have both the unshifted and shifted definitions of F11, F12, and F13 defined to be ESC,BS, and LF respectively. This leaves me with (un)shiftedF6-F10 and F17-20 to program however I need to (which is very convenient) and I don't have to shift anything to generate ESC in vi, etc. As to the position of the ESC/F11 and the >/< keys, they could be better but the VT220 is no winner either. My biggest beef about the keyboard is the relative position of the lock and ctrl keys. I often inadvertently hit lock. >P.S. To be able to "use" the WY-85, you have to order the "WY-85 Programmers >Guide" for $40.00. Escape sequences, how to use the "Compose Character" key, >how to get the information needed to write your own termcap (since the VT100 >termcap I'm using doesn't always work with vi), etc. require the $40 manual. The reply card included with my WY85 didn't say anything about charging for the Guide ... of course, 4 months later, it still hasn't arrived. We use basically a extended vt100 termcap with no problems. About the only thing different is the number of lines (25 as opposed to 24). Our terminals run at 9.6 or 19.2 kb, 7 bits even but with parity checking off in the terminal. Other nice features of the WY85 (compared to the VT220) are the status line displays, the tilt/swivel base, the slightly larger screen size, the fact that it seems to handle 19.2 kb well, and the nicer (I feel) touch to the keyboard. -- Regards, George Hart, Computer X Canada Ltd. UUCP: {allegra|decvax|duke|floyd|linus|ihnp4}!utzoo!mnetor!george BELL: (416)475-8980
albert@kim.Berkeley.EDU (Anthony Albert) (05/09/86)
Have there been studies done to obtain key-usage statistics? This would seem to be a necessary first step to developing standards. Perhaps then we could have keyboards optimized for UNIX, for text input, etc. Of course, that could lead to proposals for keyboards optimized for EMACS, vi, lisp, etc. At any rate, it would give some objective data to use in making designs. Anthony Albert ..!ucbvax!kim!albert albert@kim.berkeley.edu
dave@uwvax.UUCP (Dave Cohrs) (05/09/86)
In article <163@mtsbb.UUCP>, lav@mtsbb.UUCP (L.A.VALLONE) writes: > The streamlined feel of the "new" electronic > keyboards just doesn't cut it. Of the above, I found the hp 2621 > keyboard the closest to the Selectric and the least tiring. And herein lies the biggest problem. I *loath* the hp2621 keyboard. I hate it so much I mutilated one last night (it was an accident, really!). There are funny keys in funny places that I bump with my hands, not my fingers. The CAPS and CTRL are reversed. The feel is terrible. The timing for the BREAK key crashed my microvax (thus, the mutilation). Anyway, the problem is that we all like the keyboard we use the most. Mark Horton and I use similar keyboards and like them. Lee Vallone likes a somewhat different keyboard. I don't say he's wrong, but it's going to be rough trying to make a standard here. Maybe the hp2621 keyboards we have at Wisconsin are just mush, that's all. The touch I like best is: Concept 108, uvaxII keyboard (dumb top-row layout, though). Given time, I can get used to anything, except a Visual 200. -- Dave Cohrs (608) 262-1204 ...!{harvard,ihnp4,seismo,topaz}!uwvax!dave dave@rsch.wisc.edu
jr@inset.UUCP (Jim R Oldroyd) (05/09/86)
Another point worth adding into this discussion is something
about Function and Arrow Keys and Application Keypads.
Arrow Keys are widely used, but their positioning is random
and there is another problem (below).
If you have a terminal with Application Mode or Function Key
facilities, you may have something like 20 Function Keys and 20
keys which can work in ``Application Mode''. Now, it is quite
likely that you don't use them. This is not surprising - there is
precious little software available which makes use of them. Why?
Well, all terminals which do have them have different numbers of
them and they work in different ways.
Now, I actually make use of them. All was well, when I was only
using my VT100 compatible terminal. CHAOS ensued when I had to
make use of several other terminals. Why? Because some
terminals haven't been designed with thought to the application.
Here are some problems encountered:
- There aren't enough Application Keys to be worth it.
- When in Application Mode, DEL (the erase character)
sends some bizarre escape sequence! There is NO
key which sends DEL!
- The Arrow Keys send simple control characters
as opposed to a special sequence. It is not
possible to determine if an Arrow Key was
pressed, or the actual control character.
- One terminal (which I used recently) has a
faulty emulation of a more useable terminal!
- Pressing some keys puts the terminal out of
Application Mode.
- The Application Keys are located in strange places.
I like the way my terminal does things. It's VT100 compatible,
with 16 programmable function keys.
The entire numeric keypad (on the right) transforms to Applications
Mode on receipt of some character sequence. It is very useful
on the right (I'm right handed) - I can do a lot without moving
my left hand.
The Function Keys are not so convieniently positioned;
they are above the $%^&*()_ keys, with F1-F8 unshifted and
F9-F16 shifted. It's really easy to hit the wrong one, and
two hands are required a lot of the time.
The Arrow keys are in a row: UDLR. Actually, I find that
I can cope with this; arrangement in a diamond is not
essential. However, horizontal arrangement in an order other
that UDLR is confusing.
I won't propose any standard here, yet. I'd like to hear other
comments first. But I would like to see see something about
this in a Keyboard Standard.
BTW: I know there's an popular existing standard for this sort of
thing, and I like it.
--
"A black sheep in the family is better than no sheep at all".
--
++ Jim R Oldroyd
++ jr@inset.co.uk
++ ..!mcvax!ukc!inset!jrkwh@bentley.UUCP (KW Heuer) (05/10/86)
In article <824@cyb-eng.UUCP> cyb-eng!lee (Bill Lee) writes: >Keyboards are seemingly being designed by people that haven't ever used a >real computer, and certainly not a Unix system. The keyboard should *not* be optimized specifically for UNIX*. ("DEL means kill process, and so does BREAK, so we'll put them both over here...") This sort of thing has led to problems in the past -- didn't DEC botch a keyboard because they were thinking in terms of VMS**? (I realize I'm responding to something you didn't say, but I thought the point should be made anyway.) Karl W. Z. Heuer (ihnp4!bentley!kwh), The Walking Lint *UNIX is a trademark of AT&T Bell Laboratories. **VMS is a trademark of Digital Equipment Corporation.
rb@ccird1.UUCP (Rex Ballard) (05/10/86)
> Mark Horton want's a keyboard standard, and wonders why every keyboard is so wierd. There's a funny little problem here Mark. The "sholes" keyboard is designed to be as inefficient as it can possibly be. Yes, DESIGNED to cramp your hands, fatigue you, and make you type slower. Of course, people invest a great deal of time and money LEARNING TO BEAT THE DESIGN, learning this "hand mangler", then learning do type faster on it. Of course, there have been many attempts to standardize the basic layout. People even keep statistics on how long it takes to type, avarage keying time, and all sorts of extra statistics. However, trying to determine the best layout of the control keys, tab key, return key, and other "minor" keys, is like trying to determine whether you run faster if you tie your shoes together with a square not, or an overhand :-). Way back in the 1930's DVORAK came up with a keyboard that was significantly faster. Of course he was somewhat restricted in what he could do with his keyboard, since he was basically using a modified sholes keyboard. In the 1970's, CCI took DVORAK's main ideas, and added a few of their own, like letting the thumb do more work. This isn't the exact layout, but, their specific needs were different. > ------------------------------------------------------------------ > | ! | @ | # | $ | % | ^ | & | * | ( | ) | _ | + | ~ | > | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 0 | - | = | ` | > ---------------------------------------------------------------------- > | TAB | : | < | > | P | Y | F | G | C | R | L | ] | | | > | | ; | , | . | | | | | | | | [ | \ | > -------------------------------------------------------------------- > |LOCK| A | O | E | U | I | D | H | T | N | S | " | } | > | | | | | | | | | | | | ' | { | > ------------------------------------------------------------------ > | ? | Q | J | K | X | B | M | W | V | Z | > | / | | | | | | | | | | > --------------------------------------------------- > | SHIFT | CTRL | SPACE | > | | | | > ------------------------------------ They found out that the thumb keys could be pressed reguardless of where the remaining fingers were. In fact, the Idea worked so well, they put two more keys down there (Personally, I think four would be the max). Even on their "powerterminals", there are two control keys which can be pressed by the little finger or the thumb. These keyboards were given to directory assistance clerks who had no typing experience (typists were paid better at the time). The result, within a matter of two or three weeks, the Dvorak clerks were running 20% faster than the Qwerty clerks, monitored by automatic statistics gathering built into the system. Stangely enough, since the DA pay is now higher and trained 80wpm typists are being hired to use the QWERTY models, the statistics have gone only slightly in favor of the QWERTY/experienced vs. the DVORAK/inexperienced. We had to make QWERTY boards so that "staff" could man the keyboards when the unions went on strike. Many Dvorak proponants claim anywhere from 50-100% gains for most usages. This doesn't appear to be quite true. However, the gain is significantly greater than the fraction of a percent gain of a PC over a SELECTRIC layout, most of which could be attributed to the typists familiarity with the faster model. If the keyboard is a QWERTY, there is very little that can be done about making them "faster". You will "like" what ever you become most familiar with. You will dislike anything different until you have become familiar with that. Ironically, full-travel, "heavy" keys, are less popular among DVORAK typists, who "sweep", rather than "pound" on the keyboard. Even those so-called "mushy" keyboards are a little firm for most DVORAK typists.
flaps@utcs.uucp (Alan J Rosenthal) (05/12/86)
Many people seem to be concluding that we can't come up with a standard
at all because there are so many different interests involved. I think that
we can probably come up with some standards, even if not an entire keyboard,
because of the simple fact that too many keyboard designers get too creative
and change a usual placement of a key in a way which gains negligible advantage
but disturbs many people accustomed to its old location.
Some example standards MIGHT be:
- no key between the left shift and Z keys
- control and caps-lock in the standard order
- two keys between the L and RETURN keys
- BREAK key away from main board
- arrows in meaningful physical order
and so on. I think it would be productive for people to propose these sorts of
individual, smaller things, rather than everything at once.
Alan J Rosenthal
{linus|decvax}!utzoo!utcs!flaps, {ihnp4|allegra}!cbosgd!utcs!flapslav@mtsbb.UUCP (L.A.VALLONE) (05/12/86)
> In article <163@mtsbb.UUCP>, lav@mtsbb.UUCP (L.A.VALLONE) writes: > > The streamlined feel of the "new" electronic > > keyboards just doesn't cut it. Of the above, I found the hp 2621 > > keyboard the closest to the Selectric and the least tiring. > > And herein lies the biggest problem. I *loath* the hp2621 keyboard. > I hate it so much I mutilated one last night (it was an accident, > really!). There are funny keys in funny places that I bump with my hands, > not my fingers. The CAPS and CTRL are reversed. The feel is terrible. > The timing for the BREAK key crashed my microvax (thus, the mutilation). You caught it, my mistake. When I wrote the above, I wanted to say that the Anne Arbor Ambassador was the keyboard I preferred and found the least tiring. [It seems that my fingers aren't the only thing confused :-( ]. Sorry about being too lazy to correct this earlier. > > Anyway, the problem is that we all like the keyboard we use the most. > Mark Horton and I use similar keyboards and like them. Lee Vallone > likes a somewhat different keyboard. I don't say he's wrong, but it's > going to be rough trying to make a standard here. > > Maybe the hp2621 keyboards we have at Wisconsin are just mush, that's all. > The touch I like best is: Concept 108, uvaxII keyboard (dumb top-row > layout, though). Given time, I can get used to anything, except a > Visual 200. > I agree. Up until last year we had quite a few 2621s here and almost without exception, they were all mush. It seems people in general can get used to anything. The problem is, once we get used to something (e.g. keyboard layout) we don't want to change (how many people are using Dvorak layouts?). > -- > Dave Cohrs > (608) 262-1204 > ...!{harvard,ihnp4,seismo,topaz}!uwvax!dave > dave@rsch.wisc.edu -- Lee Vallone AT&T Information Systems Merlin {... ihnp4, mtuxo}!mtsbb!lav
mark@cbosgd.UUCP (Mark Horton) (05/12/86)
Some of you misunderstand me. I am not trying to redesign the keyboard to be optimal for typing. If I wanted to do that I'd probably start with a Dvorak layout. What I'm trying to do is standardize the layout that already exists.
gene@cooper.UUCP (gene) (05/12/86)
This may seem totally trivial, but I'm very annoyed at the seemingly
stupid practice that has been followed on most IBM personal computers.
Namely, the practice of putting the <ALT> keys on either side of the
<SPACE> bar, and the <SHIFT>s fringing the last alphabetic row of the
keyboard.
Maybe it's just force of habit, but it seems much more logical to be
able to hit the <SHIFT> keys with your thumb when typing uppercase
letters, and to hit the less frequently used <ALT> keys with the last
finger of the active typing hand. (Try it.)
If _I_ had my personal choice, I would swap the keys and use them as
I just mentioned. I suppose, also, that practice at IBM keyboards
makes this inconvenience almost transparent to the typing "pro".
At any rate, I would like to get some response, positive and negative,
to my little proposal.
- Gene
Usenet (UUCP) Address:
harvard!cmcl2\
{ihnp4 | seismo}!allegra>!phri!cooper!gene
columbia/
philabs/pete@valid.UUCP (Pete Zakel) (05/13/86)
> In article <132@sphinx.UChicago.UUCP> sas1@sphinx.UUCP (Stuart Schmukler) writes: > >In article <1314@oddjob.UUCP> mrl@oddjob.UUCP (Scott R. Anderson) writes: > >>The vt200 has ",," and ".." keys, as on many typewriters, and > >>puts <> on a key between Z and SHIFT; as you might expect, this > >>is a real pain, and no one around here likes it. > > > >I have even seen some one from DEC strongly defend this positioning in a > >discussion of standard keyboards for the APL community. In general that > >keyboard seems to be motivated by an attempt to court the secretarial user and > >the europeans. > > It would seem so, except that if they were trying to exactly reproduce the > standard typewriter keyboard, they did a lousy job of key positioning, > what with the left shift key shifted left and the return key shifted right. > -- > Scott Anderson Also, there is no reason NOT to have a shifted '.' and ',' be something different on a terminal keyboard. The reason for '. .' and ', ,' was that you still wanted to be able to type those in "shift lock" mode, but since "caps lock" on a terminal only shifts the alphabetic keys, the reason for a '. .' and ', ,' go away. Some people just don't know how to think! -- -Pete Zakel (..!{hplabs,amd,pyramid,ihnp4}!pesnta!valid!pete)
larry@jc3b21.UUCP (Lawrence F. Strickland) (05/13/86)
In article <1884@mit-eddie.MIT.EDU>, barmar@mit-eddie.MIT.EDU (Barry Margolin) writes: > In article <824@cyb-eng.UUCP> lee@cyb-eng.UUCP (Bill Lee) writes: > >I just had the misfortune of ordering a Wyse 85 terminal (the > >one that does VT220 emulation). Since I was familiar with the WY-50 and > >the WY-75, I assumed I was safe ordering a WY-85 sight unseen. Big > >mistake. The keyboard is a complete mess. .... most of article(s) elided ... > Barry Margolin I just ordered 10 Wyse 85 terminals, but first had a chance to view them. We ordered them despite the mucked-up keyboard and something that no one else has mentioned, a character font that is on the sickly side. Despite the obvious problems (I feel the <> key is the worst, but there are other opinions), the WYSE 85 is still an excellent low-priced terminal! It has been my experience over the years that people get used to using a particular type of keyboard and that becomes THEIR standard (despite ANSI). The $40 programmer's manual is my biggest objection to the Wyse terminal. Not that it is not included, but rather that it is: (a) too expensive, (b) hard to obtain, (c) not terribly well written, and (d) a long, long, time in finally arriving. I feel that Wyse should have AT LEAST included a decent table of programming functions in their standard manual (the one that is in their is worse than the keyboard)! Despite all this, I would like to stand up for the Wyse 85 as being a useful tool! -----Lawrence F. Strickland (larry@jc3b21) --------------------------- Dept. of Engineering Technology + Cthulhu + St. Petersburg Jr. College + R`lyeh + P.O. Box 13489 + wgah`nagl + St. Petersburg, FL 33733 + fh`tagn + Phone: +1 813 341 4705 --------------------------- UUCP: ...akgua!akguc!codas!peora!ucf-cs!usfvax2!jc3b21!larry
henry@utzoo.UUCP (Henry Spencer) (05/13/86)
> I *LIKE* the 220, better than any other keyboard I've used. You > get used to the layout quickly, if you give it a chance... Sorry, I have better things to do with my time than getting used to some idiot's idea of good keyboard design. We will not even consider a terminal with a keyboard layout like the VT220. -- Join STRAW: the Society To Henry Spencer @ U of Toronto Zoology Revile Ada Wholeheartedly {allegra,ihnp4,decvax,pyramid}!utzoo!henry
henry@utzoo.UUCP (Henry Spencer) (05/13/86)
> (And while I'm thinking of it: why doesn't somebody make a keyboard without > explicit function keys, but where CTRL-1 thru CTRL-0 generate an appropriate > escape sequence?) Somebody did, for a while. The Ampex keyboard I use (peculiar in other ways and no longer in production) has no dedicated function keys, but has a pair of PROG keys (PROG A and PROG B) flanking the space bar. By themselves they do nothing (just as well, considering the location!), but they shift the numeric keys to function keys -- two sets, A0-A9 and B0-B9. The sequences to be sent are fully programmable. A fairly nice scheme. -- Join STRAW: the Society To Henry Spencer @ U of Toronto Zoology Revile Ada Wholeheartedly {allegra,ihnp4,decvax,pyramid}!utzoo!henry
henry@utzoo.UUCP (Henry Spencer) (05/13/86)
> There's a funny little problem here Mark. The "sholes" keyboard is > designed to be as inefficient as it can possibly be. Yes, DESIGNED > to cramp your hands, fatigue you, and make you type slower... This is a popular myth, and verifiably false. The claim is that the qwerty layout was designed to slow down the typist, because the early typewriters tended to jam when keys were hit in fast succession. WRONG. The problem was that early typewriters tended to jam when ADJACENT keys were hit in fast succession. The solution was to spread the frequently- used keys out over the keyboard. On average, this SPEEDS UP a touch typist slightly, by increasing the probability that successive keystrokes will be on alternate hands. > Many Dvorak proponants claim anywhere from 50-100% gains for most usages. > This doesn't appear to be quite true. However, the gain is [significant] The Dvorak claims have never been reproduced by unbiased experimenters. Note also that Dvorak was working with manual typewriters, where the penalty for putting high-workload keys on the little fingers was much higher. Both simulation and experiment indicate that the Dvorak layout is 5-10% faster than Qwerty, other things being equal. This is not seen as being worth the trouble of converting the keyboard industry. If one must tinker with alternate layouts, consider that *both* the Qwerty and Dvorak layouts are poorly laid out physically. A flat rectangular bank of keys does not match the human shape very well. Look at your hands on the keyboard -- they come in at close to a 45-degree angle from each side. There has been some work in England, which I'm not up-to-date on, with a radically different keyboard that puts a curved cluster of keys under each hand. Encouraging results are claimed, not so much in speed as in reduction in fatigue. Does anyone have current info (preferably including tests by unbiased third parties) on this? I believe it was called the "Maltron" keyboard. -- Join STRAW: the Society To Henry Spencer @ U of Toronto Zoology Revile Ada Wholeheartedly {allegra,ihnp4,decvax,pyramid}!utzoo!henry
henry@utzoo.UUCP (Henry Spencer) (05/13/86)
The Ampex D175 terminal I use (some quirks, and no longer in production, so I can't really recommend it in general) has an interesting keyboard layout that I like so much I would copy it if I built or designed one. Roughly like so: Spencer's Semi-Standard Keyboard ------------------------------------------------------------------------------ | ESC | ! | @ | # | $ | % | ^ | & | * | ( | ) | _ | + | ~ | | | | | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 0 | - | = | ` | \ | ------------------------------------------------------------------------------ | TAB | Q | W | E | R | T | Y | U | I | O | P | { | } | BACK | | | | | | | | | | | | | [ | ] | SPACE | ------------------------------------------------------------------------------ |CAPS|CTRL| A | S | D | F | G | H | J | K | L | : | " | RETURN | |LOCK| | | | | | | | | | | ; | ' | | ------------------------------------------------------------------------------ |BRK| SHIFT | Z | X | C | V | B | N | M | < | > | ? | SHIFT |DEL | | | | | | | | | | | , | . | / | | | ------------------------------------------------------------------------------ | PROG | SPACE | PROG | | A | | B | ---------------------------------------------------------- This is pretty close to the Horton keyboard, with a few important changes. BS is down where you can reach it, and sizeable, not a tiny key way up in the top row. DEL is easy to reach with a swipe of the right hand (admittedly this is less than ideal if DEL is a heavily-used key). While BRK (break) may appear dangerously exposed, in fact it is safe: to get break you hit CTRL-BRK. This is a better way of safeguarding dangerous functions than trying to put them in inaccessible places. The PROG keys are shift keys that turn the numerics into function keys. BRK and the PROG keys do nothing if hit by themselves. There is one omission: no LINEFEED. How often does one use both CR and LF as distinct keys? Actually, Ampex was not quite intelligent enough to do this right. The key I have labelled BACKSPACE actually sends a strange and seemingly-useless code, which we have simply taken over as our erase character. This is an implementation botch rather than a fundamental flaw in the scheme. -- Join STRAW: the Society To Henry Spencer @ U of Toronto Zoology Revile Ada Wholeheartedly {allegra,ihnp4,decvax,pyramid}!utzoo!henry
jbs@mit-eddie.MIT.EDU (Jeff Siegal) (05/13/86)
In article <273@valid.UUCP> pete@valid.UUCP (Pete Zakel) writes: >Also, there is no reason NOT to have a shifted '.' and ',' be something >different on a terminal keyboard. The reason for '. .' and ', ,' was that >you still wanted to be able to type those in "shift lock" mode, but since >"caps lock" on a terminal only shifts the alphabetic keys, the reason for >a '. .' and ', ,' go away. Some people just don't know how to think! Nope. On a VT220, the user can select CAPS-LOCK or SHIFT-LOCK in the setup mode, depending on his/her preference. Jeff Siegal
rcd@nbires.UUCP (05/15/86)
Keyboard layout and feel, as a matter of taste... > > The streamlined feel of the "new" electronic > > keyboards just doesn't cut it. Of the above, I found the hp 2621 > > keyboard the closest to the Selectric and the least tiring. ...but... > And herein lies the biggest problem. I *loath* the hp2621 keyboard. Sure, de gustibus non est disputandum, but there ought to be enough things that have gotten screwed up on existing keyboards and have been cursed by 90% of the people who have used them. Those are the things that could be straightened out by a reasonable "standard". For example, I have yet to meet or hear of anyone who likes the DEC VT2?0 handling of < and > (on one key instead of atop ,.) I have heard it cursed so many times I've stopped counting...so why did DEC do it in the first place? Because somebody had some bright idea and there was insufficient weight of opinion and/or pre- vailing standard to stop them, I suspect. If nothing else, the volume of the response to Mark's proposal and questions should indicate to anyone who's listening that there is interest. If you read the responses, you'll also find a lot of agreement. That's worth it. -- Dick Dunn {hao,ucbvax,allegra}!nbires!rcd (303)444-5710 x3086 ...Relax...don't worry...have a homebrew.
peter@baylor.UUCP (Peter da Silva) (05/15/86)
> Here's a thought. How about a keyboard that toggles CAPS LOCK when the user > hits both shift keys together? That sound interesting. I'm another devoted caps-lock non-fan... maybe I'll try it out in my terminal program & use "capslock" for a keypad lock or something. > (And while I'm thinking of it: why doesn't somebody make a keyboard without > explicit function keys, but where CTRL-1 thru CTRL-0 generate an appropriate > escape sequence?) Because then CTRL-^ (CTRL-6) gets left out in the rain. Let me think about it, though... oh, you get into that further down... I prefer just using ctrl-x without any further rigamarole, myself. How about doing something about cursor keys? I always have a soft spot in my heart for the ADM-3a layout: it makes some ASCII sense, and is right there on the keyboard... K H L J -- -- Peter da Silva -- UUCP: ...!shell!{baylor,graffiti}!peter; MCI: PDASILVA; CIS: 70216,1076
peter@baylor.UUCP (Peter da Silva) (05/16/86)
> - Pressing some keys puts the terminal out of > Application Mode. Actually useful if you want to use the non-application mode and some idiot program has left them in it. > The Arrow keys are in a row: UDLR. Actually, I find that > I can cope with this; arrangement in a diamond is not > essential. However, horizontal arrangement in an order other > that UDLR is confusing. I much prefer Left & Right to be at the extreme Left & Right sides of a cursor row. In fact my EDTINI.EDT file does this. -- -- Peter da Silva -- UUCP: ...!shell!{baylor,graffiti}!peter; MCI: PDASILVA; CIS: 70216,1076
rubin@topaz.RUTGERS.EDU (Mike Rubin) (05/16/86)
> I just ordered 10 Wyse 85 terminals, but first had a chance to view them. > We ordered them despite the mucked-up keyboard and something that no one > else has mentioned, a character font that is on the sickly side. Yup, Wyse characters are only 5 pixels wide instead of 9 on a VT220. > The $40 programmer's manual is my biggest objection to the Wyse terminal. Of course Wyse claims they're 100% VT220 compatible so you should be able to use a VT220 manual :-) It seems, however, that the Wyse 85 doesn't properly emulate the VT220's "download fonts" feature. We've got some canned applications from DEC that download fonts; they wedge the Wyse. Note this may have something to do with the fewer pixel problem (above). We are awaiting a visit from our Wyse salesman on the question.
dave@onfcanim.UUCP (Dave Martindale) (05/16/86)
In article <167@mtsbb.UUCP> lav@mtsbb.UUCP (L.A.VALLONE) writes: > >It seems people in general can get used to anything. The problem >is, once we get used to something (e.g. keyboard layout) we don't >want to change (how many people are using Dvorak layouts?). For me, this applies to key feel more than layout - I can always fumble for the special characters if necessary. But if the key feel is too far off from what I'm used to, I can't type *anything* without many errors, or going very slowly. This is true even for a single manufacturer. From 5 years ago until 1 year ago, I used an old-style Ann Arbor Ambassador, with long-travel keys with fairly even pressure during the whole stroke. A year ago, I switched to a new Ann Arbor Genie. The keyboard was redesigned, and the key pressure on these keys increases drastically at the point where the switch contacts close. So, over time, I got used to typing by "bouncing" my fingers off the keys, seldom getting full key travel. Now we've purchased some more and they have yet another keyswitch, which requires less pressure and has where the pressure momentarily decreases as the contacts close (sort of a "snap" action, like HP calculator keypads). I have some trouble with this new one. Has anyone ever tried to define a standard interface between keyboard and terminal, so I can have my very own keyboard that I can use on any terminal (or workstation)?
bobr@zeus.UUCP (Robert Reed) (05/17/86)
Several people have suggested moving the BREAK key away from the main keyboard. This is really unnecessary if you vary the spring stiffness of the different keys. The keyboard I regularly use has (I think) 15 lb. springs for most keys and a 40 lb. spring for the break key. With this approach, even if you accidentally hit the BREAK key, you are not likely to hit it hard enough to activate it.
edward@ukecc.UUCP (Edward C. Bennett) (05/19/86)
In article <359@jc3b21.UUCP> larry@jc3b21.UUCP (Lawrence F. Strickland) writes: >In article <1884@mit-eddie.MIT.EDU>, barmar@mit-eddie.MIT.EDU (Barry Margolin) writes: >> In article <824@cyb-eng.UUCP> lee@cyb-eng.UUCP (Bill Lee) writes: >> >I just had the misfortune of ordering a Wyse 85 terminal (the >> >one that does VT220 emulation). Since I was familiar with the WY-50 and >> >the WY-75, I assumed I was safe ordering a WY-85 sight unseen. Big >> >mistake. The keyboard is a complete mess. > >I just ordered 10 Wyse 85 terminals, but first had a chance to view them. >We ordered them despite the mucked-up keyboard and something that no one >else has mentioned, a character font that is on the sickly side. Despite >the obvious problems (I feel the <> key is the worst, but there are other >opinions), the WYSE 85 is still an excellent low-priced terminal! It has >been my experience over the years that people get used to using a particular >type of keyboard and that becomes THEIR standard (despite ANSI). > Someone told me the Tandy DT-100 is really a Wyse 85. Is this true? (I haven't actually seen a DT-100 yet, so I don't know what its keyboard looks like.) I had been considering buying at DT-100 but if it's keyboard is as bad as a Wyse 85 I guess I'll keep looking... -- Edward C. Bennett UUCP: ihnp4!cbosgd!ukma!ukecc!edward Kentucky: The state that needs Japan to bring it into the 20th century. "Goodnight M.A."
meissner@dg_rtp.UUCP (Michael Meissner) (05/20/86)
In terms of being able to touch type, I too have been frustrated by different keyboards (and these were different generations from one computer vendor). I recently ran into 2 different people at Data General who have customized their terminals to require touch typing. One rearranged all of his keys, and the other substituted blank keys. Fortunatly, I had spent 4-5 years on that particular make of terminal, and my fingers still remembered where the various non-alphabetic chars are, even though I have switched to the newer (and more comfortable) model.
dricej@drilex.UUCP (Craig Jackson) (05/20/86)
I suspect that the reason for DEC putting "<" and ">" on the same key on the
VT220 stems from that model for all brain-damage, the IBM 3278. Trying to
type an ALGOL-derived language (of which C is one) on a 3278 is extremely
painful.
It is true that the 3278 has a reconfigurable keyboard, by reprogramming the
3274. By why should one have to?
--
Craig Jackson
UUCP: {harvard,linus}!axiom!drilex!dricej
BIX: cjacksonkwh@bentley.UUCP (KW Heuer) (05/20/86)
In article <661@baylor.UUCP> baylor!peter (Peter da Silva) writes: [An unattributed quote]: >> The Arrow keys are in a row: UDLR. Actually, I find that >> I can cope with this; arrangement in a diamond is not >> essential. However, horizontal arrangement in an order other >> that UDLR is confusing. > > I much prefer Left & Right to be at the extreme Left & Right >sides of a cursor row. For each direction x in {down, left, right, up}, I want the x-arrow key to be at least as far in direction x as any of the other three arrow keys. For a horizontal arrangement, this means LUDR or LDUR. (The latter is ADM-3a style.) Similarly, there are two reasonable (to me) vertical arrangements. I presume the reason for UDLR is to keep L and R close to each other. That can be useful if you tend to use UD and LR separately, I guess. I suspect the only arrangement which has a chance of satisfying a large number of users is the diamond arrangement, or maybe the "collapsed diamond" below. U LDR Karl W. Z. Heuer (ihnp4!bentley!kwh), The Walking Lint
wmartin@brl-smoke.ARPA (Will Martin ) (05/21/86)
In article <226@zeus.UUCP> bobr@zeus.UUCP (Robert Reed) writes: >Several people have suggested moving the BREAK key away from the main >keyboard. This is really unnecessary if you vary the spring stiffness of >the different keys. The keyboard I regularly use has (I think) 15 lb. >springs for most keys and a 40 lb. spring for the break key. With this >approach, even if you accidentally hit the BREAK key, you are not likely to >hit it hard enough to activate it. This is a GREAT idea! I have long wished, for example, that the "}/]" key, which is next to the Return key on my ADM-42 terminal, would have a much heavier spring under it. I tend to not reach far enough and hit that key instead of the Return; if it had a heavier spring, at least it wouldn't DO anything when I hit it lightly! The times I really want to type } or ] I could hit it directly and overcome the heavier spring. I've thought about wedging some foam under the keycap to try to fake this; anybody think that would work? Will
mouse@mcgill-vision.UUCP (der Mouse) (05/24/86)
# = mark@cbosgd.UUCP (Mark Horton)
$ = henry@utzoo.UUCP (Henry Spencer)
% = peter@baylor.UUCP (Peter da Silva)
& = rcd@nbires.UUCP
# [asks what should be present in a standard keyboard]
# (1) Regular touch typing of letters, typewriter style, should work.
Of course.
# It's especially serious
# to have the left shift key broken, since some of us used keypunches a lot,
# and only the left shift key works on keypunches. I find myself overusing
# the left hand shift key to this day because of this.
Right shift key? You mean there's a right shift key? Lemme check...why
so there is. Gee, the things you learn on USEnet.... In case you need
telling, I never use the right shift key as a shift key (see below).
# (2) The CTRL, TAB, and ESC keys are very heavily used in computer work.
Right. ESC left of 1, TAB left of Q, CTRL left of A, I agree. An
intervening key between A and CTRL is quite acceptable.
# (3) The return key should be as large as possible, with exactly two
# keys between it and L. These keys are generally ;: and '". It should
# be at least two keys wide, and an L or 2x2 square are even better.
# In practice, however, I never seem to hit the top part of the L.
I find three keys between L and CR to be OK. I too never hit the top
part of the (invariably mirror-image) L.
# (4) The numbers 1-0 should be in the standard places. The shifted numbers
# should be !@#$%^&*(), typewriter style. (Bit paired keyboards,
Like the Hazeltine 1500
# where ( and ) are shift 8 and 9, are very difficult to use.)
Tell me about it. We have some h1500s.
# (5) The location of backspace doesn't matter much.
# (6) The strange ASCII punctuation ( [] {} \| `~ ) varies so much [...]
OK. But I would note that a keyboard that is slightly different from
what you are used to is worse than one that is very different.
$ -------------
$ | RETURN |
$ | |
$ ---------------
$ | SHIFT |DEL |
$ | | |
$ ---------------
I think this is the best place for DEL. My right little finger reaches
there automatically when I want to delete something....
$ There is one omission: no LINEFEED. How often does one use both CR
$ and LF as distinct keys?
Whenever I am writing program code. Linefeed is always bound to
newline-with-appropriate-indentation, whatever that means for the
language in question, while Return is just insert-a-newline. A
keyboard missing Linefeed (or effectively missing it, like the VT220),
is almost impossible for me to use.
% > Here's a thought. How about a keyboard that toggles CAPS LOCK when the user
% > hits both shift keys together?
I am typing this on a Sun-2 keyboard, but I fixed a few things. I run
my own terminal emulator (I have some flames for Sunwindows, if anyone
wants to email...), and I tell the kernel to give me the keyboard
completely raw and do my own mapping. For one thing, this makes the
keyboard completely soft (I can program any key with any string), and
for another it means I have *both* CAPS LOCK and SHIFT LOCK. The
latter is useful (for example) when I want to type long names in upper
case with underscores (_ is on a shifted key). The Sun-2 keyboard has
LEFT and RIGHT keys to the left and right of the space bar. I have the
following behaviour coded up:
LEFT is push-on, push-off (ie, toggle) CAPS LOCK.
Both SHIFTs at once is enable SHIFT LOCK.
Either SHIFT alone, is disable SHIFT LOCK.
Incidentally, RIGHT is a LOCAL key, things typed go directly to
the output code instead of being sent off to the shell process.
% > (And while I'm thinking of it: why doesn't somebody make a keyboard without
% > explicit function keys, but where CTRL-1 thru CTRL-0 generate an appropriate
% > escape sequence?)
As above, my emulator allows loading each button with up to four
different strings, for NORMAL, SHIFT/SHIFTLOCK, CAPSLOCK, and CONTROL.
But I don't like function keys unless I can load them, because most
programs can't use arbitrary escape/control sequences for function
keys.
By the way, anyone who wants the emulator is welcome to send mail.
% How about doing something about cursor keys? I always have a soft spot in my
% heart for the ADM-3a layout: it makes some ASCII sense, and is right there
% on the keyboard...
% K
% H L
% J
So THAT's where Rogue/Hack got their movement key layout! I am all in
favor of this one; so many of us Rogue/Hack players have trained this
layout into their fingers....
# [Mark Horton's suggested layout]
# [This keyboard] uses the popular convention of [ ] unshifted and
# side-by-side, with { } shifted, [...]
I hate this convention with a passion. When running the emulator I
mentioned above (which means always), I have loaded the keys so the
keyboard looks like
----------------- instead of -----------------
| ] | } | LF | | | { | } |DEL | |
| [ | { | | | | [ | ] | | |
-------------- | -------------- |
| " | | | CR | | " | | | CR |
| ' | \ | | | ' | \ | |
----------------- -----------------
? | SHIFT |DEL | ? | SHIFT | LF |
/ | | | / | | |
----------------- -----------------
If I had to use one keyboard the rest of my life, it would be a Lisp
Machine keyboard. Second choice is the TeleVideo 950. I want:
- Keyclick (I know, Lisp Machine doesn't have it. Lisp Machine
benefits outweigh this.) Keyclick should be electronic, though
mechanical is acceptable if it does not introduce tactile feedback.
- NO TACTILE FEEDBACK UNTIL THE KEY HITS THE BOTTOM OF ITS TRAVEL.
Just the spring resisting movement.
- [/] and {/} keys, not [/{ and ]/}
- DEL in the lower right
- ESC just left of 1
- TAB just left of Q
- CTRL (not necessarily immediately) left of A
- LF between [/] (right of P) and the top of CR
- The top row to read ... 0/) -/_ =/+ `/~
Otherwise, I think other people have covered it (says I, having just
posted 150 lines....).
--
der Mouse
USA: {ihnp4,decvax,akgua,utzoo,etc}!utcsri!mcgill-vision!mouse
philabs!micomvax!musocs!mcgill-vision!mouse
Europe: mcvax!decvax!utcsri!mcgill-vision!mouse
mcvax!seismo!cmcl2!philabs!micomvax!musocs!mcgill-vision!mouse
ARPAnet: utcsri!mcgill-vision!mouse@uw-beaver.arpa
"Come with me a few minutes, mortal, and we shall talk."crowl@rochester.ARPA (Lawrence Crowl) (05/26/86)
In article <429@mcgill-vision.UUCP> mouse@mcgill-vision.UUCP writes: >I want: >- Keyclick. Keyclick should be electronic, though > mechanical is acceptable if it does not introduce tactile feedback. >- NO TACTILE FEEDBACK UNTIL THE KEY HITS THE BOTTOM OF ITS TRAVEL. But I hate noisy keyboards. The IBM PC keyboard annoys me because of its noise. I much prefer tactile feedback. >- [/] and {/} keys, not [/{ and ]/} An I prefer just the opposite. I would also like to see () and _ and + as unshifted characters. I cannot stand QWERTY, and yet there seems to be little discussion of alternative layouts. Most of the discussion centers around keys that have extremely little use except in computer languages (like vi, Pascal, and others). What about non-ascii characters? The europeans have a whole mess of characters they have had to graft around ascii. Some of the ebcdic characters (corners, lines, and such) would be real nice to have. I would like printable arrows in all four directions. How about a real divide sign? How will we deal with these? Let's face it. No standard is going to please everyone. I would venture to guess that no standard will even come close unless people are forced to use it so much that they end up liking it. (Pavlov got his initial ideas after observing secretaries using QWERTY keyboards. Oh yeah, :-) We need to work on a keyboard interface standard where people can provide their own favorite keyboard. This way, they can have their own favorite feedback, keyboard layout, tilt angle, characters and designer monograms. Any suggestions? -- Lawrence Crowl 716-275-5766 University of Rochester Computer Science Department ...!{allegra,decvax,seismo}!rochester!crowl Rochester, New York, 14627