The Qwerty Standard Letter Scheme
‘A tradition without intelligence is not worth having’.
Thomas Stearns Eliot
Telegraph Operators Guided the Design of Qwerty
In 1981 IBM held 62% of the mainframe computer market. On August 12th, the corporate giant introduced the IBM Personal Computer, legitimizing microcomputers in the office. On that day in 1981 the ubiquitous office typewriter became obsolescent, and skilled typists became keenly aware of that fact.
Because the new technology leveraged the existing skill set that all typists possessed, it was readily brought into prevalent use. That entrenched expertise was based upon an unwieldy keyboard with an inapt letter arrangement invented over a century earlier for the first successfully marketed typewriter: the 1874 "Sholes & Glidden."
The "Qwerty" typewriter keymap (Sholes - U.S. Pat. No. 207,559 - August 27, 1878) is the outcome of an inadequate design process. The inventors were focused on perfecting the mechanism, while the principal business partner was pushing for a payoff on his investment. A human-factored letter sequence was neither conceived nor developed.
Many believe that the Qwerty was contrived as a makeshift "work-around." The prototype machines suffered a high rate of jams due to sluggish mechanisms. The typebars had no return springs, relying upon gravity to return the typebars to battery. The type slugs recurringly became entangled when keys were struck in rapid succession.
The keyboard letter sequence was originally alphabetical. With no hope of perfecting the machine's mechanism in a timely fashion, the letters were reorganized to reduce the rate of jamming, making the "Type-Writer" tolerably serviceable. Home-row retains a long remnant of the alphabetical letter sequence: D, F, G, H, J, K and L.
Researchers have found evidence that Qwerty evolved to satisfy the incipient users' needs (Yasuoka and Yasuoka 2010). Professionals in various fields evaluated typewriter prototypes. Stenographers rejected the invention as a threat to their profession. Telegraphers found the crude machines to be almost useful for transcribing Morse code messages. By late 1873, after incorporating modifications to the letter scheme requested by a series of telegraphers, the layout had evolved into a "Morse code-centric" near-Qwerty scheme.
One example of the Morse code transcribers' influence is placement of the "Z," "S," and "E" keys. In Morse code "Z" can be confused with the bigram "SE" when either one occurs at the beginning of a word. Telegraphers would have to pause letter-by-letter transcribing until the letter or bigram was recognized within the context of a word. They wanted to hover over those three keys until they understood which letters to press, then they could promptly resume real-time transcribing.
Another example of the telegraphers' guidance is the placement of the letters "I" and "O," which stood-in for the numerals "1" and "0" that were missing on the archetype. The two letters were positioned just past the numeral "9" to spell out "IO," making the never-ending task of typing both dates and times easier, e.g., I879 or I2:O8.
In March of 1873, the chief inventor of the typewriter, Christopher Latham Sholes, took a lump sum of $12,000 for his interest in the machine. Sholes continued working to improve the typewriter until his death in 1890. His last patent was issued posthumously (Sholes - U.S. Pat. No. 568630 - September 29, 1896). It disclosed a human-factored letter scheme with all vowels set in the home-row under the right hand. Sholes did not want Qwerty to be his legacy!
The mechanical typewriter was an unreliable contrivance, nowhere near perfected, when investor James Densmore entered into negotiations with E. Remington & Sons. That gunmaker had the skills and resources to produce the typewriter on a commercial scale. Densmore secured exclusive marketing rights in lieu of royalties.
On July 1st, 1874, Remington began shipping the "Sholes & Glidden" typewriter, a machine that was barely usable. After the engineers at Remington had made final adjustments to the machine's keyboard letter sequence, all the letters needed to write the word "typewriter" resided in the top alpha-row. Some historians dismiss this apparent expedience as a fortunate happenstance.
That first Qwerty machine could only print capital letters. Customers complained, saying that recipients of their typed correspondence felt they were being shouted at... LOL. In 1878 an improved machine, the "Remington No. 2" was produced that would print either upper- or lower-case letters by moving the carriage with a "shift" lever. That innovation led to "Shift" keys, making touch-typing practical.
In 1880 subsequent manufacturers began entering a market where all typists possessed a Qwerty skill set. Although a few typewriter makers succeeded in marketing machines with disparate keyboard and letter arrangements, most chose to produce competing Qwerty models. The new technology was slowly adopted and through the turn of the century the industry had a little diversity; nevertheless, the Qwerty keyboard was ascendant.
The major Qwerty manufacturers formed the Union Typewriter Company trust in 1893, fixing the price of a standard machine at one hundred dollars. Moreover, the trust established overseas agencies for various member companies to "spread the Qwerty gospel" around the civilized world. By the early 20th century an international market for typewriters had developed in which diverse countries used the Qwerty keyboard scheme or a variant thereof.
The most notable alternative to Qwerty was the American-made Blickensderfer, introduced at the 1893 World's Columbian Exposition in Chicago. During the company's peak years of 1903 to 1907 over 10,000 typewriters a year were marketed to several continents in numerous languages. The typewriter keyboard's
letter scheme is human-factored. The home-row letters DHIATENSOR account for about 70% of keystrokes when typing English text.
Any typewriter manufacturers that enjoyed success marketing machines with unique keyboard layouts were dealt a mortal blow by the outbreak of war in 1914. Involved governments made the production of military equipment a national priority. Some typewriter makers were forced to retool their factories, switching production capacity over to war goods.
George C. Blickensderfer converted most of his factory over to war production. The inventor died in August 1917. The company he founded was sold off in 1919 and was sold off several more times. The Remington Typewriter Company acquired the company's assets in 1926. A Blickensderfer/Qwerty hybrid was last produced in 1928.
During the "Great War" much of the remaining typewriter production was commandeered to serve the needs of the participating governments. The Allied and Central Power governments exclusively awarded wartime contracts for typewriters to the producers and procurers of Qwerty variants. The value of used Qwerty machines and the variants skyrocketed. The die was cast...
After the massive international conflict ended, a young generation of experienced Qwerty typists flooded the workforce. The odd letter arrangement became the de facto standard; it is omnipresent and entrenched.
Ronald Earl Walker - Qwerty Typewriter Historian
The Qwerty Keyboard... Bane of My Existence
In 1970 I attempted to learn touch-typing using my mother's typewriter. After many days of dedicated study, I successfully typed "the quick brown fox jumps over the lazy dog" without looking.
Then my mind snapped. I picked the typewriter up over my head and almost threw it across the room while screaming, "'Who is the $#@% idiot that designed this $#@% letter arrangement."
That rage was brought out by realization that my intellect would suffer if I continued to touch-type using a Qwerty machine. At fifteen years of age, I rejected the idea of embracing stupidity.
Typing skills would have helped me raise my grades, now I was permanently stuck being a "B" student because of my illegible handwriting. There was no longer any hope for a scholarship.
Although I could not consistently make top grades because of my imperfect schoolwork, I did place in the top 5% of my class on the SAT. In public school I consistently “aced” multiple choice tests.
In 1973, during U.S. Army basic training, my CO offered an opportunity to attend the United States Military Academy Preparatory School. At that moment, in an odd twist of fate, I was suffering heat stroke... my brain was cooking. That offer was discussed and declined, then I reported to sick call.
I spent a very unpleasant week in the Moncrief Army Hospital with a severe URI. That ordeal made me certain that the Army was not my career. I was still coughing up blood when I returned to my unit.
After fulfilling my military obligation, I worked to advance my education, eventually flunking out of the UCF College of Engineering. Higher math was insuperable. I do appreciate the utility of calculus, and grok the basic concept, but I cannot practice it... I make too many little mistakes.
Over a decade later, I came to realize that I have "dyscalculia" and many of the multiple learning disabilities associated with it. I have an altered perception of time, a warped sense of spatial reality, difficulty telling left from right, meager ability to recall names/faces, dyslexia, et cetera.
My elementary arithmetic skills are glitchy. Results are unpredictable and erratic due to number substitutions, omissions, transpositions, additions, and reversals. A career in science or engineering was never possible. As an Army Infantry Officer, I would have caused a friendly fire incident.
Ronald Earl Walker - Dud Extraordinaire
Dholydai's Germane Information Bucket
Guidance for 3D makers:
The tolerances of the case top are tight around the switches. The 3D model's lattice structure shrinks upon cooling, so prints must be scaled up around 0.2-0.4 percent. The amount of scaling required varies by filament used. Make a partial print of a top section, stopping after the ribs begin to appear, then check the fit. There should be some wiggle room along the X axis. The top should not have to be forced over the switches.
The case bottom's deck is a solid structure. There is little shrinkage, so the models do not need scaling. Print the case bottoms with a 2-layer 15% fill between the top and bottom shells. Grid fill works fine on one of my printers, the other requires gyroid to support good bridging. The fill allows the parts to sit flat as they print and remain so after they cool. Discovering this approach stopped a constant string of warped bottoms.
Note: Printing a one-piece top on a large printer is a big investment of time and filament. My electric power provider is nicknamed Florida Flicker & Flash, so I prefer printing a sectioned multi-piece top. If you choose to print a one-piece top on a large format printer using an unfamiliar PETG batch, print a small top section the for the scaling proof. That takes a lot less time and filament.
Important information: Buy a UPS for each printer and computer on your farm. The UPS prevents damage to your printer and gives you time (2-3 minutes) to make an orderly shutdown. I use a 700VA UPS for my Prusa and a 1000VA for the Trident. Both printers have continued through short power glitches to complete their projects. A small UPS will not power a 3D printer through a long power failure.
Advice for Novices: Many STL files require repair before slicing, this is normal. Prusa Slicer and SuperSlicer have a manual repair function built into the software. After importing a badly meshed STL file into the splicer, a warning will appear. Right-click on the nearby repair icon, a triangle with an exclamation mark inside. Insanity will ensue if you try to slice without first repairing.
Getting consistent first layer adhesion with PETG can be a challenge. Adding a 3mm outer brim with a 0.2-0.3mm separation gap to G-code files will help ensure adhesion. Elmer's disappearing purple glue stick is a good aid to adhesion. Elmer's is also a releasing agent, preventing cooled parts from sticking too well to the build sheet. A thin slurry is all you need.
There is always a fan running in my print room, so the front and sides of my smaller unenclosed printer are temporarily bordered by cardboard walls when printing. That keeps drafts off the print bed, so the bed temperature stays consistent. Hot or cold spots will cause warping/lifting.
The internet is brimming with articles on 3D printing. Read more, learn more. Warning: Never use acetone to clean a PEI sheet.
Special Thanks:
Dr. August Dvorak: creator of the American Simplified Keyboard. He worked the system for many years trying to have his typewriter sanctioned. The man was ahead of his time. His life's work made it possible for me to write with a keyboard. My work extends his ideas into the 21st century.
PCBWay: my sponsor. Liam and Taylor gave me needed encouragement and helped me navigate the PCBWay+ side of their website. I have posted the Two-Hand Compact, Left-Hand Compact, Right-Hand Compact and Two-Hand TKL files on the PCBWay+ community project website.
Jeffrey Minton: creator of the keyboard_stl_generator. The software creates OpenSCAD models of keyboard cases. The case tops exhibit strong structural elements, requiring just a few tweaks. The case bottoms needed a lot of additional CADing. I named the resulting solid case 'The Slab'.
Ruiqi Mao: creator of the Keyboard Firmware Builder website. Without this tool I could not have produced the QMK firmware for my unique keyboards.
Ian Prest: creator of the Keyboard Layout Editor website. Without this tool I could not have used the Keyboard Firmware Builder or the keyboard_stl_generator.
Ronald Earl Walker - 3D Print Farmer