Isn’t this what t1 and t5 are doing, complementing before and after the adder?
I’m not too familar with these low level aspects, but, maybe, there wasn’t any time left to add another stage, in order to complete in a single memory cycle. (The timing is rather complex, because there’s no central clock, just local timing networks, i.e., various delay lines. And this was designed from existing DEC modules, in just 3 1/2 months…)
Saving logic was important back then, that may have been more important than saving states.
Sorry for the nitpick, but LLNL usually means Lawrence Livermore National Laboratory. I have rarely seen Lincoln Laboratory shortened to anything, but the website does have LL: https://www.ll.mit.edu/
(Is it better to reply here, or on the GitHub issue page?)
Yes, my bad. Thanks for the correction.
I’m not sure what’s best, but I think I prefer a comment on the issue.
The TX-2’s keyboard is probably nearly the least important thing about the TX-2, but it is … probably one of the more interesting of computer keyboards. And it seems there’s almost nothing about it online (the biggest exception being some old contemporary documents from Lincoln Labs itself), so I think I should take the excuse to ramble about it here, even though I’m not the best qualified.
The Lincoln Keyboard was the keyboard of the Lincoln Writer, which was the model of terminal designed and built by Lincoln Laboratories specifically for the TX-2. Here’s a photo of the keyboard which appears as Figure 9 in “The Lincoln Writer” by J. T. Gilmore, Jr. and R. E. Savell, Lincoln Labs Group Report 51-8, DTIC AD0235247:
And here’s a diagram, from p. 24 of the June 1958 MIT Lincoln Laboratory Division 6 Quarterly Progress Report:
And a screenshot of the on-screen replica keyboard from the TX-2 Project’s in-browser TX-2 emulator, a much clearer image:
Yes, that’s two Soroban keyboard modules in one keyboard. For reasons that aren’t completely clear to me (see pp. 6-9 of “The Lincoln Writer”) the designers ended up both using two 44-key keyboard mechanisms and eschewing normal upper-case/lower-case shifting entirely. So the Latin capitals are on the near half of the keyboard in a normal QWERTY layout, while all lower-case Latin letters are on the far half; and similarly there isn’t an upper-case variant character on any of the number or symbol keys. But the designers also didn’t want to use up too many of the 88 keys listing the Latin alphabet twice, so not all the lower-case Latin letters are available: just the important ones that you really need for your mathematical work, clustered together in the centre.
The selection of Greek letters and (other) mathematical and logical symbols should probably look roughly familiar from the later LISP keyboards.[1] There’s also a SUPER key: on the Lincoln keyboard it’s meant to put you in superscript mode (“Lincoln Writer” p. 5) and SUB and NORMAL keys come along with it. Then there’s that right-pointing manicule (☛, or is it ☞?) on the top-right key. It has a semi-official name: p. 25 of the June 1958 progress report says
The character shown in Fig. 63-6(e) has been called the “Meta Hand” and is used to indicate an aside.
. The name is also mentioned in a note sent by Lincoln Labs’ Alexander Vanderburgh to the Communications of the ACM, which was published in vol. 1, issue 7, on p. 4:
The character ☛ will be used to indicate an aside. For example, it can be used to indicate special directions to the assembly program. (It has been nicknamed the “'Meta Hand.”)
So: the Lincoln Keyboard doesn’t have a key labelled ‘META’ or ‘Meta’, but it does have a Meta key.[2] Nor did the character set stay hidden inside Lincoln Labs, it seems. In the “New Flexowriter Type Face” memo (Lincoln Labs memo M-5001-11) for TX-0 users sent on 27 August 1959, about a year after TX-0 had been moved from Lincoln Labs and restarted on the MIT campus (see p. 15 of the 1974 MIT TX-0 history by John A. McKenzie), J.B. Dennis said
A new flexowriter has been obtained from Lincoln Laboratory this week. It is equipped with the set of characters used by Lincoln Laboratory for the operation of the TX-2 Computer. …
In view of the usefulness of the new symbols, and to insure compatibility with Lincoln equipment, the new format will probably become standard at TX-0.
(This memo doesn’t mention the “Meta Hand” label, though.) I’m not sure what actually came to pass: but presumably MIT’s TX-0 users did indeed get to see and use the new character set?
However this certainly wasn’t the first time that the TX-2 character set had been used with TX-0. It was first trialled on a TX-0 on-screen keyboard, the “Lincoln Scope Writer”, in 1957(!): see pp. 4-6 and Fig. 4 of “The Lincoln Writer”.[3]
A few other features of the Lincoln Keyboard: all or nearly all characters could be underlined, overlined or printed inside a box or circle (see “Lincoln Writer” p. 7 and figs. 5-6). Between that and the colour-change option and the superscript/subscript modes there seem to have been so many different modes or combining characters that it seems especially suprising that they didn’t just implement a normal case shift as well. The Lincoln Keyboard also used the Soroban mechanism’s mechanical key interlock to prevent more than one key from being pressed at once (“Lincoln Writer” p. 9). I’m not sure exactly how that system worked in the specific case of the Lincoln Keyboard and its double keyboard.
The Lincoln Writer is also like a number of LISP keyboards in having parentheses in an unshifted position. Now, all things considered it would be stranger if it had parentheses in a shifted position, true, but it also has them in a quite central place in the near half, just right of N and M. ↩︎
(Which is in fact quite don’t say it don’t say it) ↩︎
Or the “Lincoln Scopewriter”. See also p. 25 of the June 1958 progress report, and p. 23 of the December 1957 progress report. ↩︎
Thank you, that’s a very interesting post.
Being interested in the history of ITS, Emacs, and Lisp machines, I’m particularly curious about the Meta Hand. The LINC keyboard has a META key, use to enter commands. (Similar to Meta-x if you’re into Emacs.) “Indicate an aside” sounds like a similar concept.
The next keyboard to feature a META key was from the Stanford AI lab, per Les Earnest. From there it jumped right back to MIT and landed on the Knight keyboard, which was used with the AI lab PDP-10 and early Lisp machines.
From wikipedia:
The manicule , ☞, is a typographic mark with the appearance of a hand with its index finger extending in a pointing gesture. Originally used for handwritten marginal notes, it later came to be used in printed works to draw the reader’s attention to important text.
…
The oldest book known to contain a manicule is the 1086 land survey, Domesday Book, but the age of the annotation is unknown and may date to later than the 11th century
And ofc the LM-2 space-cadet keyboard also has keys with actual left-pointing and right-pointing outline manicules (☜ and ☞) on them, above thumbs-up and thumbs-down keys.
The “space-cadet” keyboard was initially designed by John Kulp for use with MIT’s CADR lisp machines, and the Plasma Physics TV system. The CADR machines previously used the same keyboard as the PDP-10 Knight TV system. The space-cadet keyboard was manufactured by both Symbolics and LMI. Symbolics later made a slimmed down version for newer computer models.
As an aside ☛ James Kulp, John’s brother, introduced Unix job control while at Berkeley. The functionality was mostly a copy of that on ITS.
As the Gilmore and Savell paper mentions in passing (on page 8 I think), the Lincoln Writer was created by connecting to the Soroban’s lower-case circuit board the keys with the upper-case (i.e. majuscule) letters, and to the Soroban’s upper-case circuit board the keys with the lower-case (i.e. minuscule) letters.
This caused me many months of confusion in trying to reconcile various pieces of TX-2 documentation until I came across that statement.
