The Year Keyboards Changed

[quadibloc]

If you look at the keyboards of older devices used with computers (this applies to the United States and Canada only), you will find that the arrangement of characters on them is different from that on the IBM PC.

Some keyboards were "bit-pairing". Those had an arrangement based on the binary coding of ASCII. You could have an arrangement with upper case only that resembled the Model 33 or 35 Teletype, or one with lower case that resembled that of the Model 37 or 38 Teletype.

On these keyboards, ( and ) were over 8 and 9 respectively.

Those arrangements were completely different from that of the current IBM PC keyboard, although there is some resemblance between them and the arrangement of characters on a Japanese keyboard. However, one particular thing is the same: [ and ] are on separate keys, and { and } respectively are their shifts.

Other keyboards were "typewriter-pairing". That meant ( and ) were over 9 and 0 respectively, and the keyboard was laid out much the same as that on a standard electric typewriter. The cent sign, over the 6, was replaced by ^, and you didn't have keys with 1/4 and 1/2 on them.

That was pretty much the arrangement you see today on a personal computer. But with one difference. On typewriter-pairing keyboards, such as those of the LA36 Decwriter, or the Heathkit H8 terminal, ] was the shift of [, and } was the shift of {.

Borrowing the arrangement of those characters from the bit-pairing keyboard does seem to have been an improvement. (It creates a complication for terminals that support APL, though, because then the standard APL-ASCII encodings don't lead to a standard APL keyboard arrangement.)

When did this happen?

My research has led to the VT100 terminal from Digital Equipment Corporation, in 1978, as being at least among the first devices to embody this modified version of the typewriter-pairing ASCII keyboard. I think it, or some other device from DEC released that year, probably did originate the trend to this keyboard style.

It caught on quickly; as I also note on my web page, the HP 300 computer, that came out in 1979, the next year, did this also.

Since this is the place for the keyboard experts, I ask: did someone else inspire DEC, or was there some meeting where the official keyboard layout standard got revised, with DEC perhaps just being the first company to implement it?

[Lain1911]

Is this like the year cars changed?

[Parak]

I offer this:

https://www.flickr.com/photos/triplehaata/9764673284/in/set-72157635545864692

1973. Probably have to look even further back to try and find where this particular 'paradigm shift' was.

[quadibloc]

I offer this:

https://www.flickr.com/photos/triplehaata/9764673284/in/set-72157635545864692

1973. Probably have to look even further back to try and find where this particular 'paradigm shift' was.

I don't understand. The keyboard in that photo is a normal typewriter-pairing keyboard, with ] as the shift of [, and } as the shift of {, which is the older kind that preceded the VT-100 and the IBM PC.

So it is not an earlier example of the change I was talking about.

That's more clearly visible in this shot of the same keyboard:

https://www.flickr.com/photos/triplehaata/9764602755/in/set-72157635545864692/

Incidentally, as another data point, the Inside ASCII series of articles by Bob Bemer appeared in Interface Age magazine in May, June, and July 1978. Those articles included a picture of a bit-pairing keyboard, with a note that the typewriter-pairing keyboard was omitted because that standard was under review.

[E TwentyNine]

The standard seemed to come after the fact.  ANSI X4.23-1982  (can't see a copy of that anywhere without $$$, think that's correct)

I think the question might be what caused one layout to become the standard, and I'd speculate that would be the introduction of computers into office environments where typewriter layout was common, and simply the IBM juggernaut pushing through their layout as a standard, either by sheer numbers or industry pressure.

[HaaTa]

Neat.
As an extra datapoint. ~1969-1970.


Honeywell

[Parak]

I offer this:

https://www.flickr.com/photos/triplehaata/9764673284/in/set-72157635545864692

1973. Probably have to look even further back to try and find where this particular 'paradigm shift' was.

I don't understand. The keyboard in that photo is a normal typewriter-pairing keyboard, with ] as the shift of [, and } as the shift of {, which is the older kind that preceded the VT-100 and the IBM PC.

So it is not an earlier example of the change I was talking about.

Incidentally, as another data point, the Inside ASCII series of articles by Bob Bemer appeared in Interface Age magazine in May, June, and July 1978. Those articles included a picture of a bit-pairing keyboard, with a note that the typewriter-pairing keyboard was omitted because that standard was under review.

Ah, I misunderstood then. I thought that the change that you referred to was where caret was a shift of 6 and parentheses were shifts of 9 and 0.

[HaaTa]

Univac also was using separate keys for the [ ] around 1969 as well. Though they didn't seem to add the {}'s until at least 1974 (by the stuff I have).


Univac F-1355-00 1971


Uniscope 100 1974

[quadibloc]

Yes, but the Uniscope 100 is an example of a bit-pairing keyboard, and so is the Honeywell keyboard shown; the parentheses are over the 8 and the 9. On those keyboards, the braces and brackets always were this way. The change that happened in 1978 was that the VT-100, apparently for the first time, arranged them that way on a typewriter-pairing keyboard.

Incidentally, I see that the first keyboard mentioned, a Super Alps CB14182B/SCB1A163 keyboard, belongs to HaaTa, a later poster in this thread. As it has an interesting layout - it's an APL keyboard, which relegates the APL characters to the front of the key on the number and special character keys, but which has them on the top for the letter keys, because there's room on those.

As HaaTa is the one who bought it from eBay, would he know what computer terminal or other item that keyboard was originally made for?

[HaaTa]

Ahh, that makes more sense.

The earliest I have of this is some stuff manufactured in 1979.

https://plus.google.com/photos/113845661925823397356/albums/5681460096335006737
https://www.flickr.com/photos/triplehaata/sets/72157635937445325/
https://www.flickr.com/photos/triplehaata/sets/72157635829504416/

[HaaTa]

So the only thing I know about that keyboard is that is was in a box labelled "First computer of UCLA".

In terms of switch tech, I regard them as the holy grail of linear switches right now.

[dorkvader]

It's entirely possible that DEC was the first to set this trend. The late 1970's and early 1980's marked the beginning of the standardization era, where lots of companies widely copied keyboard layouts from others. How many keyboards in the 1980's are based on (either strictly or loosely) the IBM PC or the aforementioned DEC. Even the huge amount of apple ][ clones that appeared before this usually had a unique keyboard compared to the original.

Then of course, there are a lot of keyboards, even modern ones that don't follow the trend. Chilean keyboards, as well as some others have ( and ) over 8 and 9, which is similar to bit pairing. IBM 122-key keyboards usually have the old shifted } and ] keys well past the time when most manufactures stopped that trend. As always there is some overlap.

But for the first example to include both the typewriter paired () and "modern" [{ }] is likely the DEC. Even if there is an older example, I count it likely that DEC engineers were not inspired by it.

I will look through my own examples to see what I an see, but most of mine are post 1978.

[brhfl]

Not sure if it's the earliest, but the HP 9845A computer beats the DEC VT100 terminal. HP's terminals introduced around the same time as the VT100 (HP 2621A, specifically - terminals produced before the 9845A computer had the bit-pairing layouts) shared this section of layout. That said, the VT100 was a solid performer and an early adopter of ANSI escape codes at a time when terminals were a dime-a-dozen, and none of them played nice together. It was immensely popular at the time, and still plays a vital role as one of the safest fallbacks for terminal emulation. I think it's safe to say that the VT100 is the reason we still have those keys where we do today.

[IvanIvanovich]

To me it looks like the big shift to typewriter style happened around 83/84. Even the first at least 4 Cherry G80 models were bit shift which they referred to in the catalog as 'international' layout. Certainly by 85 typewriter style was the defacto standard. Before then more often than not each system had it's own unique keyboard and hardly anyone seemed to follow much of any real standards. I don't think it's a coincidence that the big PC compatible shift happened around the same time.
I've been toying with the idea of running a groupbuy set in bitshift... it's also convenient that Japanese Latin layout is still nearly the same.

[quadibloc]

Not sure if it's the earliest, but the HP 9845A computer beats the DEC VT100 terminal.

Thank you! That has answered my question, and I've corrected my web page. Yes, there may easily be something still earlier that neither of us had noticed; it's certainly hard to rule that kind of thing out.