Talk:IBM Selectric

Unsorted text
Why does Wheelwriter redirect here? Shouldn't it have its own page? Apparently no one care about the Wheelwriter, which was in production for roughly... a decade at least(not sure on when they ended production exactly) 76.184.5.83 (talk) 07:32, 15 November 2013 (UTC)

Historical tidbit: When IBM first released the Selectrics, and for a long time after, you had to purchase them directly from IBM as they didn't allow authorized dealership status to independent stores. Miconis (talk) 19:26, 9 August 2011 (UTC) Miconis

^ This was true until at least 1980. Some time between 1978 and 1980 I bought a correcting Selectric III and it was a bit of a problem just to buy the thing. As I remember we had to find a local IBM rep and ask him to make a special dispensation for a non-IBM customer. The typewriter cost $1500 at that time, but it was worth it. Lovely machine. — Preceding unsigned comment added by 210.22.142.82 (talk) 04:58, 14 May 2013 (UTC)

Does anyone remember whether Selectric I allowed half-spacing verticially? My Selectric II does, and I think my Selectric I didn't, but it's been so long I can't remember for sure. -- isis 06:33 Oct 30, 2002 (UTC)

I have rewritten this paragraph to remove the reference to WYSIWYG because I don't see how it's accurate in this context -- typewriting was always that, almost by definition. -- isis 20:53 Oct 30, 2002 (UTC)
 * The ability to change fonts, combined with the neat regular appearance of the typed page, gave many users their first taste of "what you see is what you get" text processing. Later models with selectable pitch and built-in correcting tape carried the trend even further.  Selectric models were also widely used as terminals for time-sharing computers, replacing Teletypes.


 * Yeah, but it meant a lot to me, two-finger typist that I was, to have a real nice manuscript. The auto-correct was almost as important as any other feature.  What you say is okay, but I feel more enthusiastic than that about it.  Ortolan88

Give it up -- there's no way we're ever going to make them understand what a delight typing on one of them was compared to ordinary typewriters. You had to be there. I just barely passed typing (on Remington manual machines) in high school (thanks to extra-credit projects), but the feel of those Selectric keys was a joy I sigh for with every stroke on my PC keyboard now. -- isis 07:04 Oct 31, 2002 (UTC)


 * No kidding. I just added a description of the "breakover" feeling of the keyboard, but you really had to have used one to understand it.  On a Selectric I used to be able to type sequences like "th" and even whole words like "this" by positioning the various fingers at different heights and coming down on the keyboard with one smooth movement of both hands, relying on those little spheres in the tube to sort things out! All subconscious... This also worked on the 029/129 keypunch. As nice as the IBM Model M computer keyboards and the Northgate Omnikeys are, they can somewhat simulate the breakover but they can't simulate the rollover. Jeh 20:40, 4 June 2006 (UTC) Jeh 20:40, 4 June 2006 (UTC)

I very much doubt that the Selectric had any "typeahead buffer", it used a fully mechanical linkage and encoding mechanism between the keyboard and the typeball. I have seen old repair manuals on them explaining this mechanism and how to clean and repair it. This mechanism DID support some level of "key rollover", but if it was only 2-key rollover or n-key rollover I do not recall. This "key rollover" might have given an operator an impression that there was a buffer though. -- RTC 01:03 25 Jul 2003 (UTC)


 * As nobody has commented on it I changed the item from "typeahead buffer" to "n-key rollover feature". If you have documentation either way or on exact details, please provide them. -- RTC 20:59, 8 Aug 2003 (UTC)

"the electric typewriter design that brought the typewriter into the electronic age." Is there any electronic component in the Selectric ? Ericd 21:06, 8 Aug 2003 (UTC)


 * From the manuals I have seen, no. However the mechanical linkage and encoding mechanism used was VERY easy to interface to an electronic system. Much easier than a standard electric typewriter (e.g., the IBM Executive series typewriter which had a mechanical linkage but NO internal encoding mechanism at all). Of course I have only seen manuals on the original versions, electronics may have been added later. -- RTC 16:39, 11 Aug 2003 (UTC)


 * The orginal versions up to the II's had no electronics. The only electical parts were the motor and switch. The design feature that made them easy to use for automatic operation was a set of latches that were connected to an interconnected lever system ("balance beams"?). The latches were pulled under a "bail", or bar that moved down with each keystroke cycle. The rotation (and tilt) of the ball depended on which latches were pulled under the bail. The choice of latches was made by the typebars under the keys. I never saw an electronic version, but these latches easily been actuated by solenoids place of the keybars. I used to be an IBM OPCE (Office Products Customer Engineer) in the late '60s, so I've seen a lot of the insides of these beauties. Wake 01:56, 24 Feb 2004 (UTC)

The Tricks section has a negative IBM POV without citation. I plan to remove it soon. Wake 02:27, 19 October 2006 (UTC)
 * Done Wake 01:14, 26 October 2006 (UTC)

Re the 'typehead buffer' - the IBM term for this was "Stroke Storage" meaning that as soon as the machine had started printing the current character the next character could be latched ready to go and the main cycle clutch would not pause between characters.Cookee nz (talk) 05:54, 18 December 2011 (UTC)


 * Yes, this is pretty well described in the "Features" section. The term "typeahead" was excised from the article a long time ago. Jeh (talk) 06:03, 18 December 2011 (UTC)

Selectric II vs. Correcting Selectric II
FRom my experiences with Selectrics, I'm pretty sure that the early Selectric II's didn't have correcting tape, but this article kinda blends the versions together. Anyone have a good reference to doublecheck this? Kaszeta 12:41, 10 Sep 2004 (UTC)

My high school bought a few Selectric II 's with correcting tape. They're still in use. (12/26/10) —Preceding unsigned comment added by 68.255.100.12 (talk) 00:46, 27 December 2010 (UTC)

What superscripting? I see no superscripting here
The ability of typewriters in the 70s to insert superscripts has become an issue in the presidential campaign. Somebody (anonymous, first and only wiki-contribution) whipped in here last night with a can of white out and removed the statement that the platen-adjustment on Selectric II's was for inserting superscripts. I put it back in and now add a note here just to make sure no ax-grinder takes it back out. I also web logged on this, if anyone's interested. Ortolan88 15:20, 10 Sep 2004 (UTC)

For his second edit of his Wikipedia career, the same address trashed the External links, but someone else caught it and put them back. Ortolan88 20:30, 10 Sep 2004 (UTC)

Some info from the Musuem
Vertical line spacing: A quick check of the stack of Selectrics shows that the half spacing appears to have been introduced with the Selectric II, but as an option. My original Selectrics and some of my Selectric IIs have only full spacing. The 1973 II Operating Instructions note that some models allow single or double line spacing (on carraige return), while others allow single, space-and-a-half, or double. The mechanical differences include twice as many teeth on the detent wheel on the platen.

Selectric II models from 1971-1973 did not have the correcting feature. Starting in 1973, with the introduction of the Correcting Selectric II, the correcting feature was available. However, non-correcting models were still available, even in the time of the Selectric III (I have some cloth ribbon non-correcting Selectric III typewriters in my collection).

The Selectric does not appear to be designed to use superscripting per se, but the Symbol type balls have a set of numbers (0-9) of reduced point size, in the superscript position. Per a mid 1970s IBM Selectric Type Styles brochure.

RS-232 Link: I would like to see some documentation on this! I know that the first use of the Selectric mechanism was as a terminal for mainframe computers, and that many machines were modified in the early days of the microcomputer revolution to be used as terminals. But I am not aware of the widespread use of factory built Selectrics as RS-232 interface terminals.

Rollover is two-key, using a tube filled with balls that will let only one key be pressed at a time. It's just a mechanical lockout system, but in use it feels like a buffer.

The Operating Instructions call the platen detent release lever the "line finder".

by jforb, 13 sep 04, the stamping doesn't seem to be working?

Okay, infrogmation, why did you delete the images?

Atlant 23:17, 12 Dec 2004 (UTC)

What is it with people and images in this article? Why were all images removed?

Atlant 02:18, 21 Dec 2004 (UTC) "I know that the first use of the Selectric mechanism was as a terminal for mainframe computers" Nope the first Selectrics were typewriters, terminals like the 1052, for example, came later. The Selectric based IBM 2741 Terminal had an RS232 interface. Terry (IBM Hursley Museum) — Preceding unsigned comment added by 86.134.52.145 (talk) 09:44, 19 October 2011 (UTC)

I specifically remember learning to type on an IBM Selectric typewriter in high school in 1957. Why is the date this typewriter was introduced given as 1961? Did IBM introduce them as test machines before actually putting them on the market in typing classes? 75.79.140.185 (talk) 03:17, 30 September 2014 (UTC)


 * Because there are good reliable sources, like this one, that say 1961. Per WP:RS, your personal recollections cannot be used as sources on Wikipedia. Jeh (talk) 05:18, 30 September 2014 (UTC)


 * It wouldn't have been a Selectric, but possibly one of the earlier IBM electric models you used in 1957, such as the Model B. These had conventional type hammers and a moving carriage, not the golfball mechanism of the Selectrics.  97.86.156.248 (talk) 19:12, 25 December 2017 (UTC)

Golf balls and the Wikipedia logo
Just a bit of idle speculation here, but does anyone else notice that the Selectric strike balls look a lot like the Wikipedia logo? -Litefantastic 14:59, 27 January 2006 (UTC)

Dvorak Simplified Keyboard
The Dvorak Simplified Keyboard article says that "the original ANSI Dvorak layout was available as a factory-supplied option on the original IBM Selectric typewriter". One slight problem is that the ANSI standard for Dvorak dates from 1983... So did ANSI just take the Selectric's Dvorak layout, or is the Dvorak article wrong? (After this has been cleared up, I think that a short sentence in this article mentioning Dvorak might be a good idea.) --StuartBrady (Talk) 02:42, 19 July 2006 (UTC)

Carriage length
The lengths of the carriages were 11", 13" and 15 inches w/the 13" and 15" carriages commonly refered to as 'letter' and 'legal' respectivly as that was the size of paper you could put in sideways.

Only the selectric 1 came in all three sizes. The Selectric 2 & 3 only came in 13" and 15".--Miconis 22:38, 25 April 2007 (UTC)


 * For Google's sake, let's call it "width"! Unfree (talk) 16:45, 14 December 2009 (UTC)

IBM Personal Selectric
This was a limited edition Selectric model called the Personal Selectric which was a hybrid of the Selectric 2 & 3. It only came in a 11" carriage and was also only single pitch. It used the 72 charecter ball from the Selectric 2 but the "bicycle" ribbon from the Selectric 3. --Miconis 22:38, 25 April 2007 (UTC)

Some more about the elecromechanical interface
Many years ago I bought a surplus Selectric terminal with bold plans to hook it up to my TRS-80 Model 2 to use as a printer. It never happened, of course, but I learned a lot from playing with it and doing what research I could. Some things I remember: Someone who knew the details of the device called it "A triumph of implementation over design."

As noted elsewhere, the terminal was created by taking the standard correspondence typewriter and adding a bunch of switches and solenoids. The results were fed to a big socket on the back of the deepened case. Here is an example of what this augmentation consisted of: the print ball was designed with the upper and lower cases 180 degrees apart. "M" and "m", for instance, had exactly the same tilt and rotate codes, but with a different half ball facing the paper. The Shift key or the Caps Lock key would spin the print ball around by throwing an arm far to the right. To allow the programmer to test if the device was in Shift mode, there was a switch in the frame of the typewriter that was held down if this arm was deflected to the right. To do the job right, the programmer would test the data pin coming from this switch before sending the Shift command from the computer. When he sent the Shift command, it activated a solenoid that somehow threw this shift arm to the right. The Caps Lock did the same thing, but then it locked it in place until the Unshift command was sent.

There was a story that when the System/360 was first demo'ed, IBM did not want to call attention to the system console because they had not released the Selectric yet. So they showed the computer with a piece of cardboard wedged in the open bail of the system console. This was to hide the fact that there was a spinning ball doing the typing, not a bunch of type bars. From what I know of the release dates of the two machines, the Selectric was at least five years old and a common sight in offices all over the country when the 360 was released, so I doubt this story very much.

I remember that the most important individual component in the 360 was the "5" key. When the Shift-5 was pressed, this sent the EOB (End Of Block) code to the machine, which signified the end of the operator's system command. Since it was used so much more than any other key on the console, it failed before any other key and its fault brought all new processing to a halt. Anything running would complete; anything new waited until the Customer Engineer was rousted out and came in to replace the mashed switch. I spent a lot of my young life waiting for this repair so that I could get on with system testing. Charley6alphacharley 20:49, 28 July 2007 (UTC)


 * As has been said, the story about S/360 and the Selectic can't be true because S/360 was announced (1964) AFTER the Selectric (1961). Terry

only way?
Nevertheless, between 1968 and about 1980, a Selectric was about the only way to get high-quality output from a computer.

This implies that something happened in 1968 that made it harder to get high quality output. What was that? Also, it's just plain false: it was possible to put a film ribbon on a 1403 and get really nice looking output. Some books were set this way; the Griswold SNOBOL book and Gries' Compiler Construction for Digital Computers come to mind. K6rfm 05:23, 6 August 2007 (UTC)


 * No, I wouldn't call it "really nice looking." While far better than most drum printers at character registration the 1403 was nowhere near good enough to be called "letter quality" if a Selectric was your standard for that. Jeh 01:10, 7 August 2007 (UTC)


 * Did you ever see a 1403 with a film ribbon? (Also had to have the hammer flight times adjusted recently, and a clean chain.) K6rfm 15:44, 8 August 2007 (UTC)


 * I never actually used one, but I certainly read a number of books that were "mastered" that way, not excepting some of the IBM Systems Reference Library (I'm pretty sure the PL/I Level F Language Reference Manual was one). I'd assume that IBM did whatever they could to get the output as clean as possible. And others, like the APL/360 manuals, that were prepared on a 2741, presumably with film ribbon. There was just no comparison. But the 1403 was certainly better than any other line printer of its day.


 * Meanwhile, though, the 2741 and similar machines were never the only way to get high quality output from a computer. All you had to do was punch 6-level TTS code on paper tape and feed the tape to a Linotype or phototypesetter with a tape reader. The Linotypes existed long before anybody thought of driving them with computer-punched tape. So I'd say your original objection ("what happened in 1968?") was spot on. Jeh 20:49, 9 August 2007 (UTC)

Keyboard layout
what we know today as the selectric keyboard layout: did it in fact originate with the selectric? it should probably be mentioned in any case —Preceding unsigned comment added by A plague of rainbows (talk • contribs) 01:04, 3 January 2008 (UTC)


 * There's something to be said here. 2 used to shift to ", not to @, and many typewriters didn't have a "1" key at all; the typist had to use a lowercase L. However the IBM Executive Model D would be something to look at here. Jeh (talk) 13:12, 9 November 2008 (UTC)


 * I heard you could order it in either a QWERTY ("Sholes", that is, the "ordinary" one) or Dvorak layout at a store in midtown Manhattan. I'm sure there were many other layouts for languages other than English.  Dvorak wasn't very popular, probably because Dvorak typists were inhibited by the cost of ownership, no longer an issue on PC's for personal use.  Being one myself, it's a real drag using the QWERTY's in public places, especially when switching to Dvorak isn't enabled!  I rarely look at the keys anyway, so that's no problem, and the digit keys are the same.  Unfree (talk) 16:36, 14 December 2009 (UTC)


 * What is the "Selectric keyboard layout"? "Layout" doesn't appear in the article.  Unfree (talk) 16:44, 14 December 2009 (UTC)


 * Electric typewriters in general, before the Selectric, used the modified keyboard layout where " and ' were on the same key, instead of " being over the 2 and ' being over the 8 - with _ over the - instead of over the 6. That was so that the small characters, ' " - and _ which needed to hit the paper with a lighter force would be on the same key, so that the mechanism for controlling the impression strength of different characters wouldn't have to take the shift into account. Quadibloc (talk) 18:48, 13 October 2012 (UTC)


 * At least some of the IBM Executive series (the later ones) did that, and they also added "one" and "zero" keys to the top row (previously we had to use lower case ell and upper case oh)... which also made room for two more special characters (we used to have to overstrike the single quote and the period to get an exclamation point!). However some of the more basic/cheaper electrics like the Smith Corona portable electrics retained the old "two shifts to double quote", etc., arrangement. They were relatively inexpensive and  I suppose SCM couldn't be bothered with changing the impression strength.


 * Just FYI: The old "two shifts to double quote", etc., arrangement was enshrined in the ASCII code set, in that the change from the code for a 2 to a code for a double quote is a change of just one bit, and same for all other keys on the keyboard ("shift" changes different bits for different keys, though). So a keyboard laid out this way is referred to as "bit-paired" because the difference between the unshifted key and the shifted key is always just one bit changing state. This layout was adopted by 7-level Teletype machines like the famous ASR-33 because it was easy for the partly-mechanical encoders and decoders to implement, and many other early computer terminals followed suit. Later computer keyboards and terminals used a Selectric-like layout, which eventually became the standard, but it made the keyboard encoding slightly more difficult. (These days on a Mac or PC the keys don't directly emit ASCII codes anyway, so it doesn't matter.) This was referred to as "typewriter-paired", as it followed the "pairings" on the Selectric and other better electric typewriters... but, confusingly, not all typewriters! Jeh (talk) 01:11, 14 October 2012 (UTC)

Triple Pitch
I would swear that I've typed on a triple-pitch Selectric (8/10/12). Am I nuts?

I changed the word "selective" on this page to "selectable". They don't mean the same thing.

WilliamSommerwerck (talk) 11:55, 9 November 2008 (UTC)


 * As there were no type elements made for 8 pitch I'd say yes, you're nuts. ;)


 * Seriously, both the Selectric II and III had dual pitch (10/12) but not triple. Some of the daisy-wheel electronic typewriters of later years could handle 10/12/15 pitch and they had a typewheel (called "Micron" iirc) for 15 pitch. Jeh (talk) 13:09, 9 November 2008 (UTC)

You may be thinking about the IBM "Composer" Typesetting machine, available as "Selectric Composer", "Selectric Composer System" and "Electronic Selectric Composer". These indeed had three pitch settings but were not really 'pitch' as such as the spacing was in typesetting measurements.Cookee nz (talk) 05:43, 18 December 2011 (UTC)

Influential design
To say that the Selectric "was" an "influential design" (as opposed to "influential typewriter" or "model") in the opening sentence sounds off-topic, if not judgmental, something you might read in a trade magazine. Its design might have influenced typewriter engineering, but who cares about that? Its design does matter; its influence matters; but would you start an article "The Stradivarius violin was an influential violin design"? It's of interest to violin makers, those who design, but to the vast majority of people, who listen to or play violins, it's the instrument itself, not its design, which is of interest. A violin made by Stradivarius sounds good. (I know that's judgmental, but it's not the writer's personal judgment.) In today's world, new models of things come out every few minutes; you can't get the model you like, because it's yesterday's.  But the Steinway model B piano won't change, and can't be improved upon. (The model C can be improved upon; that's what a model D is: an "upgraded" model C. Or so I've heard.)  What about that classic jazzy electric organ (I forget the name) with the slidebars? Who wants something "like" it? You want the original. Its design is trivia; it's the thing itself that matters. The Selectric was more than a "design". It was a phenomenon. Unfree (talk) 14:51, 14 December 2009 (UTC)
 * Fixed by substituting "model line" for "design". Unfree (talk) 16:17, 14 December 2009 (UTC)

A minor note about the physical design aesthetics: In addition to being squarer, the keys on the Selectric 3 were rougher then previous models which made the feel very different. Miconis (talk) 19:28, 9 August 2011 (UTC) Miconis

13
Neither "thirt" nor "13" show up in the article, except in "IBM 1130". Here in the discussion, 13 has come up in discussing platen widths. But I recall hearing somewhere that the Selectric had a top speed of 13.5 characters/second. True? Unfree (talk) 16:17, 14 December 2009 (UTC)


 * Hard to say. The IBM 2741 version could and did achieve 14.1 char/sec (this is calculated from the serial bit rate and character format) but that was a ruggedized mechanism; it may have been slowed down a bit for reliability and that still might not be the mechanism's top speed. This person seems to be pretty knowledgeable about them and says "over 15 characters per second." I believe I read 15.5 somewhere, way back when. Whatever the limit is, it's enforced by the basic rotation rate of the motor and hence the main drive shaft after the clutch - the machine completes one print cycle per rotation of the main shaft and however fast the shaft turns, that's the limit. I do know that if you type fast enough, the clutch doesn't declutch between cycles, so the main shaft turns continuously, but this doesn't require typing at exactly the maximum speed - there's a small window. This was a critical factor in adapting an office Selectric for computer output, as constantly declutching and clutching would wear out the clutch very quickly, so you wanted to keep the machine busy. Jeh (talk) 01:46, 6 March 2010 (UTC)

Golfball
"(occasionally known as the IBM Golfball typewriter)" That's pure hogwash, isn't it? Is it a verifiable registered IBM trademark? What's it doing with a capital G and a small t? What was the nature of the "occasion"? Even as a colloquialism, I doubt it's worth mentioning very prominently. How often, and why, would one refer to a "golfball typewriter", and why mention IBM? Can you imagine being in a room full of various typewriters and asking for the "IBM Golfball" one? I bet "golfball" was never used except in reference to or describing the typeball, or to tell somebody to look for a machine with a black golfball in it. Unfree (talk) 17:16, 14 December 2009 (UTC)


 * I worked in many places with Selectrics and heard the term only very very rarely and absolutely never from anyone from IBM. (IBM was a very no-nonsense organization.) Jeh (talk) 01:49, 6 March 2010 (UTC)

Selectric terminals - mechanical vs. electronic connection between keyboard and printer (1050 vs. 2741)?
The article says


 * The keyboard and printing mechanism were mechanically separated (so that keystrokes do not necessarily result in immediate printing)

but I seem to remember that, in the IBM 2741, the connection was mechanical. The online 2741 manual says on page 9, in the section "IBM 2741 Keyboard" of the chapter "IBM 2741 COMMUNICATIONS TERMINAL DESCRIPTION", that "the 2741 keyboard is physically identcal to the standard IBM Selectric® typewriter". There was a "Print Inhibit" feature that allowed the host to put the 2741 into a mode where "all printable characters received or keyed at the 2741 are recognized and cause the print element to function, but no printing takes place", but, as I remember from that feature's use in Multics to suppress the echo of passwords, it worked by the high-tech mechanism of raising a small bar that prevented the golfball from striking the ribbon or the platen.

I didn't use an IBM 1050 as much, but I have the impression that, in the 1050, the keyboard and print mechanism weren't mechanically connected, and echo suppression could be done purely electronically. Guy Harris (talk) 06:40, 28 June 2010 (UTC)


 * From my further recollection I believe you are correct on all points for the 2741, and related machines like the Mag Card Selectric. On the 1050 series the keyboard and printer were manufactured, packaged and, in some cases, offered as separate units in the first place (in the "remote terminal" version of the 1050 you could configure a system with printer only). But that sentence fragment is about converting a Selectric, so the 1050 example doesn't apply. I'll remove that bit. IIRC I wrote it in the first place, I don't have a problem with pulling it. Jeh (talk) 20:37, 3 September 2010 (UTC)

Correcting feature
I picked up a Selectric II at a garage sale that reads "IBM Selectric II" on the label without the word "Correcting" but it does have the correcting feature. Does anyone have an explanation for this? —Preceding unsigned comment added by 97.126.174.176 (talk) 04:53, 7 August 2010 (UTC)


 * Either - Perhaps in later years IBM only offered the "Correcting" version and so dropped that designation from the machine's name. Or - It's possible you got a "refurbished" machine that has a cover from a non-correcting model. Both of the above are pure speculation. Jeh (talk) 20:39, 3 September 2010 (UTC)

Overseas or so-called "World Trade" Models often omitted model descriptions like this. In New Zealand for example, the Selectric badges were almost always just "IBM" with no mention of 'Selectric' at all.Cookee nz (talk) 05:47, 18 December 2011 (UTC)

Video in wiki commons showing Selectric Type Element in Motion
I just put a video in wiki commons that shows the IBM Selectric type element in motion - using a high speed camera - and which discusses the whiffletree mechanical digital to analog converter used to control the ball. I am the author of this video and so cannot attach it to this wiki page without a conflict of interest ... I leave up to those editing this page. If anyone has comments about what would make it more appropriate for wikipedia let me know ... might be able to revision the vids currently being released to meet wikipedia needs. These are, of course, licensed under creative commons ShareAlike 3.0. —Preceding unsigned comment added by Hammack (talk • contribs) 22:08, 16 November 2010 (UTC)

End Date
Could someone please provide the date production stopped. —Preceding unsigned comment added by 70.51.120.145 (talk) 15:24, 27 February 2011 (UTC)
 * I added info on the Wheelwriter's introduction in 1984. Selectric production presumably stopped around then, but I haven't found any source that gives a specific date.--agr (talk) 20:31, 27 February 2011 (UTC)

Wheelwriters were introduced in 1984, but Selectric III's were still available until 1986. IBM states that the "Selectric" brand was retired in 1986. Clark Hinson (talk) 13:27, 12 September 2014 (UTC)

Noise
Maybe this is too trivial to include, but how about a comment on the noise the machines made? It was loud and quite different from conventional electrics and manuals. Kdammers (talk) 10:31, 19 June 2011 (UTC)

"It was loud" ? I can't comment too much on the I and II but the II was definitely not "loud." If anything, a conventional mechanical typewriter is noticeably louder than a Selectric (of any vintage.)  — Preceding unsigned comment added by 210.22.142.82 (talk) 06:04, 14 May 2013 (UTC)


 * I too remember the noise, not of the machine running, but the tremendous pounding of each ball-strike. The Selectric was heavy, and the pounding vibrations could damage the common wooden desks and their spring-loaded, swing-down typing platforms. Do I remember the pressure being settable for multiple-carbons? This is 1964–67 high-school I'm trying to recall, so Selectric I era. A current neighbor who did compositor work at home in '69, reports no noise issue. I would have thought noise was noteworthy in the article; maybe not, though. John Sinclair (talk) 22:21, 18 September 2013 (UTC)


 * Yes, there was a little lever with a small round red knob with five possible positions for the impact force. However the notion of damaging a wooden desk's swing-down typing platform seems to me to be preposterous. I could see that happening on a manual typer, where you're pushing down to make the typebars move, but on the Selectric (or any other electric for that matter) there is no such downward impact at all. It is all typebar or element against platen, and this is resisted by the frame of the machine, but it is in the fore-to-aft direction, not downward. Also, consider that the impact force, even set to maximum, was not even as great as on a conventional typewriter; the Selectric was known for not being able to cut carbon copies (or mimeo stencils!) all that well.... though this could be improved by picking a fine-lined font rather than the heavier Courier. I also agree with .82 above that it was quieter than a conventional typewriter too. Probably those desks had already been damaged by years of use with manuals. Jeh (talk) 00:11, 19 September 2013 (UTC)

An optional feature on the Selectric 2 & 3 was a "security motor". This was a bulbous extension attached to the regular motor added as an extra security precaution due to belief that someone could tell what was being typed just by the sound of the ball hitting the platen. Miconis (talk) 19:27, 9 August 2011 (UTC) Miconis


 * Another feature in this area - the Selectric II was available with a transparent "noise shield" that covered most of the carriage area. However I distinctly remember a Consumer Reports article on the machine commenting that they didn't find it particularly noisy without the shield, nor much less so with it.Jeh (talk) 21:43, 11 September 2011 (UTC)

There was indeed a security motor for the Selectric and it was used by all US Government agencies in London e.f Embassy, Coastguard, Navy, etc. but its uses was nothing to do with sound. Because each letter used a unique "latch code", it was shown that each code could cause a different current drain, thus allowing a spy to see what was being typed by simply monitoring the electricity mains. The flywheel attached the motor smoothed the current take. Terry — Preceding unsigned comment added by 86.134.52.145 (talk) 10:02, 19 October 2011 (UTC)

"Selectric as computer terminal" section
It's arguable that much of this is more about the IBM 2741 than about the Selectric, and maybe should be moved to that article. Jeh (talk) 21:41, 11 September 2011 (UTC)

The 2741 does not use EBCDIC, but a special code related to the tilt and rotate needed. The conversion from either EBCDIC or ASCII is done on the host. Also, no 2741 that I used had the ability to not print. Passwords were blocked by printing and backspacing combinations of M, X, W, and maybe some other letter. Gah4 (talk) 21:02, 26 February 2015 (UTC)


 * As is described at IBM 2741 there were actually two different line codes available, "correspondence" (elements were interchangeable with ordinary office Selectrics) and "PTT/BCD" (elements were interchangeable with the ones used on e.g. a S/360's 1050 console typewriter). A 2741 had to be ordered for one or the other. 2741's certainly did implement "print inhibit". It's described in the IBM manual, referenced from this article. The control codes to do it are called "bypass" and "restore". Perhaps the OS you were using wasn't written to take advantage of it. Jeh (talk) 21:18, 26 February 2015 (UTC)

Yes, but neither correspondence code nor PTT/BCD are EBCDIC. For one, the 2741 uses shift and unshift to shift the characters printed by the rest of the codes (and rotate the ball one half turn). If the OS (WYLBUR) knew that the 2741s could do "print inhibit", I don't know why they wouldn't have used it. Maybe it cost more and they didn't want to pay for it. If only some could do it, then the system would have to assume that they didn't, unless it could detect it. Gah4 (talk) 01:37, 27 February 2015 (UTC)


 * Who is claiming that anything is EBCDIC? The article says ... oh, I see. The second mention of EBCDIC. WP:SOFIXIT :) Jeh (talk) 02:59, 27 February 2015 (UTC)


 * Re the "bypass" command, I didn't write that portion. But again, it's documented in the 2741 manual (which is linked from the mention) and not mentioned as an option. And the character codes for "bypass" and "restore" are certainly in the codeset (also in the manual, and in the IBM 2741 article). Why is this such a point of contention? Jeh (talk) 03:08, 27 February 2015 (UTC)

In Popular Culture
In Rėgis Roinsard's film ''Populaire (2012), character Louis Échard shares his idea for a 'golfball' with friend Bob Taylor, who offers it to an American manufacturer at the World Typing Championship. When asked why the French would offer this idea to Americans Bob Taylor, (American born, naturalised French), offers "France for love, America for business!". — Preceding unsigned comment added by 195.137.58.249 (talk) 12:00, 6 June 2013 (UTC)

In the Columbo section, a Selectric is also featured in the opening scene of the first episode, s01e01 "Murder by the Book", the victim of the first episode is an author seen writing at his Selectric. I dont think just using a typewriter is generally sufficient for mention, however since the section for Columbo already exists and it is so prominent being the opening scene of the opening episode it seems worthy of inclusion. full episode w/ subscription and still from episode — Preceding unsigned comment added by Doc hammerfish (talk • contribs) 03:15, 2 April 2023 (UTC)

lines per inch
This article mentions a pitch lever that selected either 10 and 12 characters per inch horizontally.

How many rows of text could be typed per inch vertically with "normal" single-spaced text? I see some non-reliable sources (Talk: Courier (typeface)) that seem to indicate this was 6 rows per inch for "12 point" 10 pitch "pica" text. Was the leading exactly the same 6 rows per inch for the smaller "10 point" 12 pitch "elite" text? Is there a reliable source that mentions the leading? --DavidCary (talk) 02:36, 8 June 2015 (UTC)


 * It was exactly the same both pitches (3, if you include the composer's 15 per inch) deisenbe (talk) 09:35, 7 May 2016 (UTC)


 * Yes, changing the character pitch did not change the vertical leading.
 * 6 lpi was also very commonly used - given the "forms layout templates" and "printer spacing charts" we used to use, I would say almost universal - by the computer line printers of the day.
 * However, at least one other line pitch was available. A place I once worked for used many Selectrics with OCR elements for text entry (at the time they were about a fourth the cost of a computer terminal). The OCR scanners were more reliable at 5 lpi than 6, so the Selectrics had a special toothed wheel that made them work at that pitch. This was a special order from IBM. Jeh (talk) 18:35, 7 May 2016 (UTC)

IBM Wheelwriter article needed
There is a Wikipedia entry for "Wheelwriter", but it is just a redirect to this article, and there is scant information about the Wheelwriter here. I don't have much direct experience with the Wheelwriter, because I moved on to a different job, with mostly non-IBM equipment. But it seems to me that there is enough historical information for a separate (probably much smaller) article about IBM's successor to the Selectric. Is anybody up to starting such an article? Reify-tech (talk) 19:30, 27 November 2016 (UTC)


 * I paid little attention to the Wheelwriter series, but it seems to me it was not particularly significant. Certainly it did not not revolutionize the typewriter business the way the Selectric did. It was also not a new mechanical concept as the Selectric was; it was a late entry to the category of typewriters built on a daisy-wheel mechanism, a trend that was already well started (by Xerox, Brother, and others) while IBM was still selling Selectric III's and Electronic Typewriters to a shrinking market. Maybe we need an article on "Daisy wheel typewriter" to cover them, or maybe these machines should be covered in a section of the Typewriter article, or Daisy wheel printer. Jeh (talk) 22:56, 27 November 2016 (UTC)


 * Done: IBM Wheelwriter. DigitalIceAge (talk) 02:32, 8 September 2023 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on IBM Selectric typewriter. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20040403105833/http://www.ibmcomposer.org/ to http://ibmcomposer.org/

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 06:33, 23 March 2017 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on IBM Selectric typewriter. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20111002034133/http://www.krsaborio.net/research/1970s/770214.htm to http://www.krsaborio.net/research/1970s/770214.htm

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 01:04, 10 November 2017 (UTC)

Link 25 is dead
Reference #25 now goes to a placeholder website. — Preceding unsigned comment added by 24.130.68.55 (talk) 00:01, 18 September 2018 (UTC)
 * Thanks for letting us know. I've replaced the dead link with a link to an archived copy. 28bytes (talk) 01:40, 22 September 2018 (UTC)

On upgrading from pole vault to Howitzer
My mother had one of these in the early 1980s (with the correction ribbon). Not mentioned in this article (that I could find quickly), these things were real brutes. If a poltergeist was levitating your office desk, one of these babies put an end to that in short order.

The internal mechanism had quite a bit of recoil, but due to the incredible mass and stiffness of the steel enclosure, it didn't actually rock your desk very much. There was very much a Howitzer-style recoil issue that the design had to somehow resolve.

The Selectric was also unkind to "burst mode" touch typing, where familiar words burst out of your fingers almost like a complete pattern. I've been touch typing almost my entire life: 30 wpm on an Underwood at age 14, a manual typewriter with a shift key so heavy and with such a deep throw I used to call it the pinky-finger pole vault (poor technique buckled the last knuckle on my youthful pinky finger more than once). At this point in life, especially during email (idiomatic), full words pour out in rapid pulses. But you can't do this on the Selectric as I recall it due to the mechanical interlock. Far better to pace yourself to a uniform typing speed. You had to be pretty metronomic to achieve typing speeds anywhere near 90 wpm as I recall it. I doubt I ever got over 60 wpm on this beast, though I now burst easily up to 120 wpm on an electronic keyboard with a completely different typing cadence.

I found the Selectric much better suited to "production" typing at the pace of known material than composition at the pace of your imagination over your desk.

Much of that is mere opinion, of course, but the industrial "heft" of this machine could certainly be mentioned in the article.

Psychologically, for me at least, the correction function was a disaster. Instead of worrying about getting you words straight, one became far too fixated on obtaining "clean" copy in the first pass. It would have been far better to leave mistyped words behind—starting each mistyped word over again—and then to go back later with a pen an cross out the corpses. The worst pages could have been retyped later. But as an adolescent, I lacked the judgement to figure this out. I ended up doing my essays with a pen, which I found unbearably slow. Seven times out of ten, when handwriting a long word with barely any stem (such as "unforgivably"), one or more letters would go missing en route as my brain raced ahead to finish the sentence. And this would often happen again if I crossed the mess out to rewrite it properly (though if I slowed myself down enough to voice each letter out loud, I could finally defeat this: you en ef oh are gee eye vee eh bee el why). My "missing letters" handicap goes away above 60 wpm, which is not achievable with a pen. Once I got my first computer with a printer, I've never used my cursive handwriting since for anything other than signing my name. &mdash; MaxEnt 16:16, 10 April 2020 (UTC)

Why not straight to the daisy wheel?
In view of the complexity of the ball control mechanism, and how much simpler and more straightforward daisy wheel printing works, it would be interesting to know why the ball element design came first.Cancun (talk) 15:29, 23 May 2021 (UTC)
 * I believe the cylinder for the Teletype ASR-33 came earlier, and maybe other teletype models. The Selectric is electromechanical. (Hopefully the article says that.) Daisywheel didn't come until computers were good enough to time the wheel motion. Though it might also be that belief that plastic was rugged enough took time, too. I will guess that an electromechanical daisywheel would be more complicated than the sphere. (That is, mechanical powered by a rotating shaft and solenoid clutches.) It is too easy now to forget how things were done before digital electronics was as fast and cheap as today. I believe that the Intel 8008 was used in some daisywheel terminals. Gah4 (talk) 23:44, 23 May 2021 (UTC)
 * I believe the cylinder for the Teletype ASR-33 came earlier, and maybe other teletype models. The Selectric is electromechanical. (Hopefully the article says that.) Daisywheel didn't come until computers were good enough to time the wheel motion. Though it might also be that belief that plastic was rugged enough took time, too. I will guess that an electromechanical daisywheel would be more complicated than the sphere. (That is, mechanical powered by a rotating shaft and solenoid clutches.) It is too easy now to forget how things were done before digital electronics was as fast and cheap as today. I believe that the Intel 8008 was used in some daisywheel terminals. Gah4 (talk) 23:44, 23 May 2021 (UTC)

Olivetti LEXIKON 83 D.L. and other generica
Even though this lemma includes IBM as trademark, we should give reference to succeeding products that occured when patents ended. 77.1.52.57 (talk) 09:14, 15 February 2023 (UTC)
 * (Presumably that's "lemma" in the lemma (morphology) sense.)
 * If somebody wants an article to discuss more than is currently mentioned in the article, one very good way for them to do that is for them to add it themselves. (It's not as if the word "IBM" in the title is preventing anybody from mentioning non-IBM products.) Guy Harris (talk) 21:24, 15 February 2023 (UTC)