-*-Text-*-  File: TERMS, Node: Top, Up: (DIR), Next: Buyer's Guide This file attempts to describe a few of the numerous display terminals on the market. If a terminal is supported by either ITS directly, or by CRTSTY indirectly, then it should be described herein. We may choose to describe other terminals as well. This file does not deal with all aspects of display terminals. We are speicifically concerned with those features that affect the terminals usability with display editors such as Emacs or Unix's ex/vi. Many of the terminals we will describe have features which would be of note if we were considering other applications, such as filling out forms. However these features do not generally affect the value of the terminal for use with a display editor; if anything, they tend to degrade it. This file is also distributed with Twenex EMACS as a terminal buying guide. A terminal can be supported by Twenex EMACS if it has cursor positioning, CLEOL and CLEOS; insertion and deletion of lines and characters are recommended. * Menu: * Buyer's Guide:: What to look for when buying a display terminal * Quick Summary:: A quick look at the current market leaders. * Sample:: Sample terminal entry; Read this First! * ACT-IV:: MicroTerm ACT-IV A & B * ACT-V:: MicroTerm ACT-V * ADM2:: Lear Siegler ADM-2 * ADM3A:: Lear Siegler ADM-3A * ADS980:: ADDS Consul 980 * Ambassador:: Ann Arbor Ambassador * Bantam:: Perkin Elmer Bantam 500 * CDC713:: Control Data 713 * C100:: Human Designed Systems Concept 100 & Concept APL * DD4000:: Delta Data 4000 * DD5200:: Delta Data 5200 * DM1520:: Datamedia Elite 1520 * DM2500:: Datamedia Elite 2500 * DM3025:: Datamedia 3025 * DM3052:: Datamedia 3052 * DM4000:: Datamedia 4000 * DM-DT/80:: Datamedia DT-80 * FOX:: Perkin-Elmer FOX-1100 * Freedom-100:: Liberty Systems Freedom 100 * H19:: Heathkit H19 * H80:: Hazeltine Executive 80 * H1500:: Hazeltine 1500 * H2000:: Hazeltine 2000 * HP2621:: Hewlett-Packard 2621 * HP2640:: Hewlett-Packard 2640 * HP2645:: Hewlett-Packard 2645 * IBM3101:: IBM 3101 * IMLAC:: Imlac PDS-1 and PDS-4 * I100:: Infoton 100 * I400:: Infoton 400 * LP:: NLS Lineprocessor * MB:: Beehive Mini-Bee * MIME:: Microterm MIME * MOD1:: Hazeltine "Modular One" * OMRON:: Omron something * OWL:: Perkin-Elmer OWL-1200 * SB:: Beehive Super-Bee * SWTPC:: SWTPC CT-82 * TAB132:: Tab Products 132 * Tek4020:: Tektronix 4025 series * T1061:: Teleray 1061 * TVI912:: Televideo 912 and 920 * TVI950:: Televideo 950 * VIS200:: Visual 200 * VT:: DEC Video Terminal * VT50:: DEC VT-50 * VT50H:: DEC VT-50H * VT52:: DEC VT-52 * VT55:: DEC VT-55 * VT61:: DEC VT-61 * VT100:: DEC VT-100 * VT132:: DEC VT-132 * VT4800:: VTT VT-4800  File: TERMS, Node: Buyer's Guide, Up: Top, Previous: Top, Next: Quick Summary Most of this file and this section in its entirety is maintained by Charles Frankston . The opinions expressed herein are wholly his own. This section describes the criteria by which most of the terminals in this file are judged. The user is specifically presumed to be interested in a terminal for use with a display editor such as Emacs. The terminal reviews mention the state of software support of a particular terminal on the ITS systems. However, in reviewing a terminals features we try to point out whether a terminal has features, that though not required by ITS currently, might someday be useful, or ought to be useful by other editors of simlar philosphy. Ie. ITS has no use for extra pages of memory in a terminal, but Berkely Unix's ex/vi editor can make use of the feature. To begin with any terminals that do not use the ASCII character code, any that do not display all printable ASCII graphics (ie. upper and lower case) will not be given serious consideration in this review. In addition it is essential that there be a way of generating all 128 possible character codes from the keyboard. Sometimes, an otherwise decent terminal seems to miss a code or two, in which case it will usually be mentioned in the review. To be used for display editing purposes a terminal must have certain capabilities of cursor movement and screen modification. In addition to the required capabilities there are some desirable capabilities which generally serve to improve the bandwidth of an editing session. Different system software may be prepared to compensate for various missing capabilities. For example on ITS the CRTSTY program is capable of using a terminal with the most minimal capabilities, whereas direct ITS operating system support (sometimes called :TCTYP) support, demands certain minimum functions be available. Multics Emacs and Twenex Teco terminal will both attempt to support nearly as minimal a terminal as ITS CRTSTY, however, Twenex Teco does not try nearly as hard to send the absolute minimum number of characters to acomplish the job. The most minimal capabilites necessary for support by ITS CRTSTY are simply the ability to randomly position the terminals cursor anywhere on the screen and the ability to use the cursor to erase or replace characters on the screen. Note that random cursor motion does not mean the terminal needs a code to specifically position to some absolute place on the screen; it is sufficient if there are codes to move the cursor in any of the 4 directions. The system can figure out how to get anywhere it needs using this. Of course "absolute positioning" capability is a desired feature, since it often takes fewer characters to get someplace that way (ie. improves "bandwidth"). Similarly, there needn't be any commands to specifically erase portions of the screen. CRTSTY knows how to use spaces to erase characters if the terminal has no erasure commands. However this is extremely poor in terms of bandwidth and is only for the most patient people, preferably connected at high speed. For support by the system, the terminal should have minimum of the ability to clear from the cursor to the end of a line (called "CLEOL"), and the ability to clear from the cursor to the end of the screen (called "CLEOS"). CRTSTY usage puts an extra load on the system and slows down user typeout and response in general [*Note CRTSTY: (INFO;CRTSTY)Top.]. In addition to these rudimentary features, there are 4 features commonly found for shifting text on the screen which are greatly desired for improving bandwidth. They should be considered essential for anyone operating 2400 baud or less, especially 300 baud! These are the ability to insert a line in the middle of a screen and have the rest of the lines on the screen "ripple" downward (the bottom line disappears), the ability to delete a line and have all lines beneath move up one line (the bottom line should preferably be blank then) and the analagous operation for characters, inserting a character and having all characters to the right of the cursor shift rightward one character (and hopefully disappear if the right edge overflows, not wrap to the next line!) and the ability to delete a character in a line and have the rest of the characters in the line shift left 1 column (but not the next line!). These are usually called line insert, line delete, character insert and character delete (I may often just say line/char ins/del). It used to be that these functions only appeared in the more expensive terminals, however lately terminals costing apx. $1000 have begun to have these features. Similarly, some of the cheapest terminals on the market (for example, the ADM-3A) lack some of the features mentioned, such as CLEOL. DON'T BUY ONE. There are terminals as cheap or cheaper that do offer such features, and they will make life much more comfortable. With this introduction, I am going to list, in order of increasing price, ONLY THOSE TERMINALS WHICH I BELIEVE ARE WORTHWHILE LOOKING INTO. I will list few details here, you should peruse the individual entry for each terminal that interests you. There is a second section on terminals I do not recommend buying, but which are listed because many people have asked me about them and I want to explain the reasons for a non-recommendation. At the end of each entry I state whether a terminal is supported by ITS or not. Unless it is stated otherwise, one can assume that a particular terminal is either supported, or presumed to be easily supportable by CRTSTY. If there is a particular problem CRTSTY has in supporting a terminal though, I will mention it, because only a gross misfeature could make CRTSTY give up on something. There is a problem insofar as many of the terminals I have listed as being most cost-effective are fully supported by ITS right now. This will clearly have to change in the future. Clearly whichever of these terminals becomes popular will become candidates for full support. If the terminal you choose does not become popular it will likely not be supported because each supported terminal takes up a considerable amount of space in the resident system. Note that some of the newer editing/terminal support systems, such as Multics Emacs and Berkeley ex/vi seem to have the same requirements/performance tradeoffs as CRTSTY. Lastly, a note on prices. Things are not as firm as they used to be. Dealers are marking many of these terminals down. If you want to get more up to date info than I give here, you might be well advised to thumb through the ads in the latest issue of Byte magazine. Remember though, it might be worth a few extra dollars to buy from a local dealer who'll be more responsive to your problems than a mail order house.  File: TERMS, Node: Quick Summary, Up: Top, Previous: Buyer's Guide, Next: Sample This page is just meant to give people a quick idea of the best terminals to look for in each price range. You should definitely read the full entry for any terminal that strikes your eye. I don't have full entries for many of the cheap terminals however with these is you can usually get to try one at your local computer store. The date next to each entry is the last time I remember rereading it for accuracy. I have not been paying very much attention to the under $100 part of the market recently. There are about 5 manufacturers seriously fighting in the $600 - $1000 arena. A few of these even have line/char insert/delete. In general though, the features are not high, the keyboard quality low etc. $645 Lear Siegler ADM-5 (Sept 81) Finally the successor to the ADM-3A. They added a couple of keys, numeric keypard, a couple of standout modes, clear to end of line and clear to end of screen. $650 ADDS Viewpoint (Sept 81) Has random cursor positioning, clear to end of line and screen, and detachable keyboard or fair quality with N-key rollover. $695 Televideo 910 (Sept 81) I frankly don't know the features of this terminal. Higher priced Televideos have had rather poor keyboards without N-key rollover. $750 Hazeltine Esprit (Sept 81) This is supposed to have char/line insert/delete, but its a bit hard to tell at this point. $750-$900 Microterm ACT-V, MIME (1979) Microterm has several new models now. I just don't know much about them. All entries after this definitely have line/char insert/delete. $695-$995 Zenith Z-19 (1980) The $695 is for the kit, $995 assembled, but is heavily discounted by dealers. Has fairly respectable keyboard (not detachable) with 2+ key rollover. Sensible mechanical construction. This seems to be the best bet in this price range nowadays. $1000-$1095 Teleray 1061 (1978) This terminal was fairly populer for a while. Has detachable keyboard and possibly a nicer looking screen than Z-19. Can be gotten with Meta key. $1170-$1300 Ann Arbor Ambassador (Sept 81) This terminal is currently my leading choice for price performance. One of the first "real" (reasonably-sized screen of 48-60 lines, very little padding requirements, etc.) terminals to appear on the market. This terminal does many things right. Latest models have a Meta key. $1350 Human Designed Systems Concept-100 One of the most featureful terminals in terms of power of commands for manipulating screen communications lines etc. The packaging recently improved somewhat. Overall it probvably does more things right than the Ambassodr, but still only has 24x80 screen, with four screens of memory. Meta key is coming as an expensive option ($200?). $1200-2000 DEC VT-100 & imitators There can be a wide variance in price depending on whether you are buying the real DEC model and through whom, or an imitator. Big feature is 132 column capability (only 14x132 unless you buy "Advanced Video Option" $270). Standrad VT00s have no insert/delete capability, but the "scroll region" capability is actually more general and can be used to simulate; but no char insert/delete. VT132 adds these features but somehow makes them and the whole terminal so slow you don't want to bother. Some of the imitators that have the VT132 type capabilities may have done this better though. The imitators (Datamedia, Plessey, Microterm, Datagraphix, C. Itoh, Direct, Cado, HDS, Televideo, Teleray, Tab Products, TI, is probably only a partial list). generally give you the advanced video option standard, and sometimes one or two other features (most interesting perhaps is Datagraphix which has a very sharp 132 columns). If you want a VT100 I advise looking up some of these companies and compare the emulators. You might save some money or get some more features by reaping the benefits of free competition.  File: TERMS, Node: Sample, Up: Top, Previous: Quick Summary, Next: ACT-IV SAMPLE Sample Terminal to show format of terminal documentation. Status: Says whether terminal is supported by system, CRTSTY or just here for informational purposes. Users: Those people who actively use the terminal and can probably supply more information about it. Price: The best guess as to the current purchase price. Screen: size of screen, color of phospher if known etc. Chars: size of dot matrix used for characters Keybrd: Evaluation of keyboard feel. Should say if detachable, how much rollover, and if it is typeamatic (repeats when a key is held down). Might also mention whether keyboard layout corresponds to either bit paired (model 33 Teletype, ADM-3A) or typewriter paired (VT-52, HP2621, C100, Office Selectric Typewriters) layouts. Most keyboards use one of these two layouts for the keys closest to home position. There is unfortunately far less standardization on the placement of keys like Return, curly brackets, rubout, tab, or the control key. Speeds: Communications speeds in baud; mention if split. Has: Those features that the terminal supports with hardware operations; can be: Random Random cursor movement (really a necessity) Addr Direct Cursor Addressing Tabs Fixed or settable 8 column tabs CLEOL Clear to End of Line CLEOS Clear to End of Screen InsChar Insert Character DelChar Delete Character InsLine Insert Line DelLine Delete Line Standout Some form of highlighting text on screen. (Eg. reverse video, blinking, differing intenstity) Overstrike Actual overstriking of characters. Lacks: Presumably whatever from the list doesn't appear under Has: Misfeatures: "Extra" goodies that manage to get in the way. There may follow a few lines of comment, followed by perhaps the culprit(s) most responsible for the entry and the date on which is was last updated.  File: TERMS, Node: ACT-IV, Up: Top, Previous: Sample, Next: ACT-V Microterm ACT-IV A & B Status: Supported by ":CRTSTY ACT-IV" Users: RAM Price: ACT-IV A: $550 (A is without monitor) ACT-IV B: $880 (B is with monitor) Screen: 24 x 80 Chars: 5 x 9 (in a field of 6 x 10) Keybrd: fair Speeds: ...,19200 Has: Random, Addr, Tabs, CLEOL, CLEOS, Standout Lacks: InsChar, DelChar, InsLine, DelLine, Overstrike Misfeatures: no insert/delete line/character commands in standard unit. a) For $50, they are willing to supply a new ROM with the control- codes reassigned. Note that they ARE willing to include remote functions for insert/delete char and line, and for tab (it's a fixed 8 column tab ... but still better than naught.) b) The engineer (Tom) warned me, and I saw it for myself, that the insert/delete line functions are rather slow. Depending on the amount of data on the screen, you can sometimes actually see the data "scroll" down or up over a finite interval (100 ms or so?) The explanation for this is lengthy, but technically acceptable as a questionable design decision. He admits that there are othe ways of doing it, and understands the problem with the current method. He implys that in a month or so he'd be willing to supply with a new control ROM (they apparently use 2708s throughout) which would implement a different system as a test of feasibility. (The basic problem is that they didn't want to blank the display, so they use the current method to avoid the screen from "flashing" with the change. Since the ins/del operation would be about 100 times faster under the easiest alternative, it is unclear whether or not the overall effect would be disconcerting or not. They never tried it. However, as I say, he seems pretty reasonable about the whole thing. c) The unit uses an F8 microprocessor. The code is in F8 assem. For 25 bucks, they'll sell a complete technical manual with all the schematics, timing diagrams, logic flows, program listing with comments, etc. Right now the program is in 1K with about 20 bytes free, though supposedly in 3 months or so they will come out with a new board which will have 2K capacity. (Obviously, the dedicated hacker could add another K now if he really wanted to.) The willingness to give out the program listing and discuss modifications is really pretty rare in my experience. The 2K version will also have protected fields (useless to me.) d) The unit appears to have virtually no serious speed problems in its basic state (not counting the ins/del stuff mentioned above, which is not basic state.) I believe that one or two operations require one null of padding at 19,200 baud or some such. The processor apparently uses a UART for character disassembly instead of doing it in code (a definite win, as the superbee experience proves) and simply ignores data from the UART while working on internal operations. Thusly, padding causes no problems (and shouldn't be too bad for ins/del either ... though it would have to be calculated if you wanted the minimum delay.) e) Glitch. The unit I saw had a typical cursor glitch. Whenever the cursor is in the lefthand column (col. 0) a small, rather faint image of the cursor appears over on the right (about two dots wide, not full cursor width.) Apparently they all do it. When I asked about it, Tom offered two explanations for me to choose from: the sales explanation and the truth. The sales explanation is that it provides verification that the cursor is all the way at the lefthand column. (If anyone wants to buy that, I have some swampland in Jersey I'd like to talk to them about.) The truth: it's a glitch! It's a hardware problem, a bear to solve (I know the feeling, I've fought them myself) and not worth it for a terminal which is, after all pretty damn inexpensive. Basically it was not particularly annoying, but I did see it. The preceeding evaluation is from Lauren@UCLA-Security. I've used one at home for 3 months with surplus monitor and speed aside I'm content.But negotiations for quantity 3-10 purchase leave me with impression that they're still in the little leagues. .... RAM 2/13/78  File: TERMS, Node: ACT-V, Up: Top, Previous: ACT-IV, Next: ADM1 ACT-V Micro-term ACT-V Status: not yet supported. Users: ?? Based on preliminary information - no users yet. Price: $865 quantity 1, $650 quantity 25 - add $30 for non-glare screen. Screen: 24 x 80 or switch selectable to two columns 24 x 39. Chars: 5 x 9 matrix, true descenders. Keybrd: see comments. Speeds: 110...19200 Has: Random, Addr, Tabs, CLEOL, CLEOS, Standout (intensity), Underline. InsChar, DelChar, InsLine, DelLine are available from the keyboard. It is not clear yet, if they are available from the host computer. (see the ACT-IV on this same issue). Lacks: Overstrike Misfeatures: see comments. Available from: Micro-Term, Inc.; P.O. Box 9387; St. Louis, MO.; 63117 Comments: Thus far, the reports have been somewhat conflicting: RPOOR 10 Nov 1978 It appears that MicroTerm is breaking out of the hobbyist market. This terminal comes in a reasonable looking case. It has lots of features: printer interface, transmit line, transmit screen, numeric keypad, cursor control keys, request character at cursor, display control codes. Some folks at SLAC have been in posession of MicroTerm's MIME-I, and were impressed with the feel of the keyboard and the clarity of the display. MicroTerm keyboards seem pretty good. This one has auto repeat on the space, period, and cursor keys, and a separate repeat key for the others. For IBM keypunch fans, has 'NUM' key which puts the numerics above the j,k,l,u,i,o keys (of dubious value). If that's not to your liking, it does have a separate keypad. It can produce all 128 ASCII codes (I approve). Haven't seen a ACT-V yet (07-Nov-78), but a demo is coming soon and I will report further then. ------- Date: 13 Dec 1978 1310-PST From: Mo at SRI-KL Subject: Micro Term Act V These are some of the obvious problems we saw with the current design: 1) All features which work in local mode only should work in remote mode. 2) Underlined lowercase letters with descenders look horrible -- j,g, q They didnt do it quite right. Either take the feature out or do it right- doing things half assed is silly. 3) Num key is the most incredibly stupid idea I have ever seen in 8 years of computer work, especially with the separate pad right there. Gas. 4) The graphics on the keytop for unshifted = looks like an underline when in fact it is a minus (-) sign. 5) The keypad should either have a comma or a space key nearby, or both. How are you going to separate numbers without using the other hand? 6) The screen wavers with the insert/delete features. A little is acceptable, but the amount these things wiggle is very distracting. 7) Bell should not end character insert mode. 8) The lowercase character set is a bad choice. Now, things that should be done to make it acceptable as a Datamedia replacement include: An Edit key. Prefer two of them, near the space bar, double length. Pay the extra dollars and use a better keyboard. Cherry or Micro switch keyboards feel much better and last longer. The internal layout is the worst. Why did they put the board right next to the crt yoke when it looks like to mount the thing on the side would have required no additional anything-- just a remount. Use a better lowercase -- Datamedia's is the nicest I've seen.  File: TERMS, Node: ADM1, Up: Top, Previous: ACT-V, Next: ADM2 Lear Siegler ADM-1 Status: Supported by ":CRTSTY ADM2". Users: ETC Price: about $1500 Screen: 24 x 80, P4 (white) phosphor, apparently old ones available 12 x 24 Chars: 5 x 7, lower case optional, may be standard nowadays. Keybrd: ? Speeds: 110,150,300,600,1200,2400,4800,9600 Has: Cursor, Addr, Tabs, "Edit option": CLEOL, CLEOS, InsChar, DelChar, InsLine, DelLine, Lacks: Standout?, Overstrike Misfeatures: Escape key does not transmit until 2nd character is pressed. 2 Escapes however will transmit as one. Vendor: Lear Siegler Inc;Data Equipment;Aneheim, CA 92803;(714)774-1010 Escape misfeature is found annoying, otherwise users seem to like it. Is priced competitively for features. CBF 5 March 1978  File: TERMS, Node: ADM2, Previous: ADM1, Up: Top, Next: ADM3A ADM2 Lear Siegler ADM-2 Status: Supported by ":CRTSTY ADM2". Users: LRH, CFFK Price: somewhere in the $1600 range, you should probably negotiate. Screen: 24 x 80, P4 (white) phosphor Chars: 5 x 9 (in a 7 x 11 field) Keybrd: Apparently good. Speeds: 110,150,300,600,1200,2400,4800,9600 Has: Cursor, Addr, Tabs, CLEOL, CLEOS, InsChar, DelChar, InsLine, DelLine, Standout (Reverse; blinking & half intensity available but not used) Lacks: Overstrike Misfeatures: Escape key does not transmit until 2nd character is pressed. 2 Escapes however will transmit as one. There seems to be an option available from the manufacturer that might fix this, but it is reported the fix has side effects worse than the bug. Vendor: See ADM1 Escape misfeature is found annoying, otherwise users seem to like it. It used to be priced competitively for features. CBF 5 March 1978  File: TERMS, Node: ADM3A, Previous: ADM2, Up: Top, Next: ADS980 ADM3A Lear Siegler ADM-3A Status: :CRTSTY ADM3A tries real hard, but at slow speeds you might be happier with :TCTYP GLASS Users: Lots of people. Its cheap. Price: about $800, look for ads in Byte etc. Screen: 24 x 80 Chars: 5 x 7 without descenders. Keybrd: actually sort of usable Speeds: ...,19200 Has: Cursor, Addr Lacks: CLEOL, CLEOS and everything else. Misfeatures: interprets some control code to lock keyboard. (Diasablable by a physical switch on the terminal). Is missing enough features to qualify as misfeatured. Vendor: See ADM1, also every computer store, Byte magazine. Two virtues are being cheap and fast. On the scale of things its not as cheap as it used to seem. However, there still aren't many terminals that can stand up to a continues 19,200 data rate. Also, being made out of a lot of discrete parts improves chance for doing ones own maintainance or extension. However, useing as an ITS display at less than 4800 baud is an excercise for the infinitely patient. CBF 5 March 1978 CBF 8 March 1980  File: TERMS, Node: ADS980, Previous: ADM3A, Up: Top, Next: Ambassador Status: Supported by ":CRTSTY ADS980". Users: BYRON Price: About to be discontinued and replaced by new "Regent" series. Screen: 24 x 80 Chars: ? Keybrd: Poor-to-fair. TTY-style keyboard but advantageously has all controllifiable characters on separate keys so that - is never necessary. Speeds: ? Has: Random, Addr (not supported), Tabs (5 columns!), CLEOL (CR clears to end of line and does carriage return), r InsLine, DelLine, Standout. Lacks: DelChar, InsChar, CLEOS, altmode key (!) Misfeatures: DEL and underscore exchanged The terminal also has "graphics" capability with 160 by 72 resolution, and two local "editing" modes which are useless on ITS, but are useful on line-at-a-time systems. The number pad may be selected to send escape sequences rather than digits: - is sent instead of . This is used by CRTSTY to simulate Altmode, Meta-, Ctrl, Top-, and Call, as well as to toggle the graphics and standout features.  File: TERMS, Node: Ambassador, Previous: ADS980, Up: Top, Next: Bantam Ann Arbor "Ambassador" Status: Supported by Twenex EMACS, ":CRTSTY AAA". Users: CPR@SRI-KL, CBF@MIT-MC Price: Compact box, $1300 single unit; bigger box, $1430 (no difference inside); compact box is designed to be carried. Bigger box is designed to be bigger. 7 terminals get you a 5% discount, 16 gets 10%, 31 gets 15%, etc. That makes 25 terminals cost $1170 each, for example (I'm referring to the cheaper box version only; it's more compact and less Captain Marvel-looking). MIT price: $1170 Screen: 60 lines max by 80 characters; P39 high-retention green phosphor; large 15" diagonal screen; very readable even at 60 lines. Has "zoom" feature, allowing 60x80, 48x80, 40x80, 36x80, 30x80, 28x80, 24x80, 20x80, 18x80 (maybe one or two more) format (it changes from non-interlaced to interlaced at some level like 36x80). At 60 and 48 lines, I think there's only one scan line between lines, which makes it seem a bit cramped, but it's still quite readable. Inverse video mode makes it more readable at high densities. Chars: 7 x 7 when more than 40 lines on screen (not sure of boundary), 9 x 9 otherwise. Keybrd: Simulated N-key rollover, good tactile feel (highly subjective, of course). (Early versions have 2-key rollover.) Detached keyboard, uses funny ribbon cable; typeamatic (*), keyclick (*) (*) = can be disabled. Typewriter-paired layout. Non-typewriter keys well-laid-out, looks somewhat like: -- function keys, local editing keys -- ESC - numerics, etc. - BREAK plus TAB -standard qertyuiop- LF BS your CTL CL - etc - RET RUB basic PAU SHFT - - SHFT UP DN keypad where CL is Caps Lock, PAU is Pause (sort of XOFF/XON toggle), and UP, DN are explained below, I think. Has 12 down-loadable shiftable function keys (for a total of 24 if I might be so bold to calculate) above the major grouping, with another 3 settable on the keypad; total function key storage: 256 bytes (to be arbitrarily distributed among the keys). The rest of the local editing keys aren't down-loadable. Speeds: all standard to 19.2KB, not splittable Has: everything; ANSI standard with extensions. Standout modes: reverse video, underlining, blinking, bold, blank, in all combinations (this is implemented by storing 12 bits for each character, so, e.g., you can't complement the screen easily---this feature should be trivial to add in hardware by XORing the high order bits during display). However, any erasing operations (such as kill to end of line) set the erased region to spaces in the current standout mode, so just setting it to reverse video and typing ^L to your editor flips the whole screen (unlike Heath's, which always kill to blanks). A crude, probably useless, form of windowing (see below). No padding requirements at 9600 baud for all functions we tried, except for screen clear. Flow control: XON/XOFF, settable, but, interestingly enough, you can download the characters that get sent for XON and XOFF (up to four for each function). Lacks: True overstriking. Misfeatures: The keyboard is not a true N-key rollover. Due to the hardware scanning technique, there are certain simultaneous combinations which are ambiguous; however they seem extremely obscure. P39 phosphor is VERY high-retention and thus scrolling is too streaky for words. Some have also complained that the characters are too "fuzzy". Vendor: Ann Arbor Terminals, Inc., 6175 Jackson Road, Ann Arbor, MI 48103. 313-663-8000. 10 sales regions, Boston area rep is Digital Sales Assoc's, 9 Spring Street, Waltham, 617-899-4300 (ask for Art Dempsey). See below for service information. It's a full ANSI-standard terminal, driven by a 4MHz Z80, with a battery- backed-up CMOS memory inside to store settings; it has a VT100-like "setup" mode, with 4 separate mode groups to deal with. The 15-inch screen is truly nice; even at 48 lines, it doesn't look too cramped; it does look extremely wide. At 60 lines, it's a matter of taste whether it's readable or not. [A feature on the new models is the ability to store a powerup sequence of some 200 chars in the non-volatile memory. This can be used to load the function keys and the terminal answerback.] It has full print and copy features through an auxiliary port, if you're interested in that kind of thing. It has block mode sending, etc. They actually seem to have done it right, e.g., you can have nearly all flavors of printing and copying happening through the printer port, which operates asynchronously at any speed. The screen format is pretty interesting: it's always 80 characters wide, but as you step down from 60 to 48 to 40 to 28 to 24 to 18 (or some such sequence, I forget), the lines get farther apart and the characters get slightly bigger (they use two dot matrices: a 9x9 and a 7x7). The UP and DOWN keys next to the right SHIFT key are used in conjunction with CTRL to shift the window (locally) up and down a line or to zoom out and in when CTRLified (the host can also zoom). Don't know how useful this is or how confusing it might get if you somehow zoom around by mistake. It has some notion of 3 "screen areas" which are totally settable; they refer to them as the "upper host area", "middle operator area" and "lower host area"; I suppose they're useful in some kind of special application mode. It does appear that you can position the "operator" (middle) area and use insert/delete line to do pretty much arbitrary region scrolling like the VT100. However, since it has insert/delete line, it is probably not worth doing region scrolling that way. Nearly everything is settable from either the keyboard (in setup mode) or from the other side of the communications line (e.g., a host machine). It has a few interesting features, such as Push and Pop Line, which let you shove any number of lines down to the bottom of the screen (remember, all 60 lines are considered part of the screen), or pull them back up to any point; I suppose this can be used as a way of moving lines around on the screen arbitrarily by pushing them to the bottom from a cursor position, moving somewhere else and popping them back. It has an erase-character-under-cursor feature just like ITS loves. Seems to have a "justifying tab" (?) You can push and pop the cursor position, like on the H19, which is pretty useful if you don't want to remember where you were but want to get back there later (e.g., mode line hacking). At 9600 baud, it seemed to need NO PADDING for even the hairiest "insert n lines" functions, which makes it another first; even the hardwired DM2500 requires some padding for insert/delete line. [Later, when adding Twenex EMACS support, we found it needed some 2-400 ms of padding or waiting for a full screen clear.] It can go at up to 19.2Kb; it'd be interesting to see what it looks like at that speed. (One interesting psychological fact: at 9600 baud, 60 lines on the screen, full-screen refresh looks a lot slower than on standard terminals, for obvious reasons.) [For outputting simple text, including letting the terminal scroll, I could not get it to lose even at 19200 baud. -CBF] Ann Arbor has always had a reputation of building sturdy products. They've gotten a somewhat black name from places that bought custom terminals, such as BBN, and modified them even further; AA could never keep up the documentation for over 50 different versions of TTL logic boards; they claim it'll be different this time, since they aren't customizing the Ambassador. The hardware "theory of operations" manual looks reasonable. The Ambassador in the "E" has only one CPU board (piggybacked by a smaller board) and one power supply/CRT electronics board), and their early-life failure rate is claimed to be very low (they burn them in at 50 degrees Celsius for a week). They only have self-maintenance, which helps explain why their overhead is so low and they can keep their prices down. The current service theory is that you return a bad card (or whole unit if you're stumped) and they will return it within 5 days, fixed, for "usually less than $100 per card or $200 per terminal". You can also get 24-hour turnaround if you're willing to pay a $200 premium. If you buy more than a few terminals (10, say) they will give you a free, full-day maintenance training session (you get to fly to Ann Arbor, of course). They have an extended year's warrantee which costs $130 per terminal and which covers any repair costs for the first year of life. The keyboard feels quite sturdy and everything is housed in metal casing; the smaller-box version is rather compact and both the terminal and the keyboard are housed in lightweight, very sturdy plastic cases. I don't know how the keyboards themselves they'd hold up under heavy abuse, but I will wager they're not as fragile as VT100s. (The keyboard cable is a flat ribbon, fairly durable-looking frob, but it's raised some eyebrows of people who've seen it.) I pinned them down on the question of keyboard scanning, seeing the way that Teleray lost initially with the T10's, and they said the keyboard was scanned 100 times a second, with interrupts off, so there was no possibility of losing typed characters due to high screen activity. They said they hadn't heard of n-key rollover (sigh) until recently, and that it wouldn't be possible until the second version of the Ambassador (but at least they admitted it was a good thing). Still, with 2-key rollover, it isn't awful. [I had no complaints about the keyboard, but I don't touch type - EBM.] As far as ucode fix distribution goes, they're willing to send new PROMs if you find gross bugs that prevent you from doing your job with the terminal. It's not clear what happens with bugs of the form "when you print 8 asterisks, backspace 3 times, sneeze, and then go into inverse video mode, the screen glitches"; they will probably not send out PROMs for such obscure problems. They only come with a P39 phosphor, which is a very green, high-retention phosphor; this means it blurs and streaks quite a bit when scrolling; at MIT-EECS, we run them in wrap mode so they don't scroll, but this is a local Twenex feature only. Rand Corporation, which evaluated them, has complained about a particularly bad 15KHz whine which bothered their secretaries (one of the students at MIT complained about it, too); Ann Arbor claims they've fixed that. [Latest update: the whine is still noticeable.] All in all, looks like a nearly-complete win if you want the screen size, balanced by the blurring and keyboard lossages. --CPR, EBM, Saturday, 13 September 1980 Wednesday, 1 October 1980 Thursday, 11 December 1980 (after three weeks using them) The marketing manager (Sarah Freeman of Ann Arbor) is now quoting 12 week delivery on a version including N-key rollover and an option to have the PAUSE key act as a Meta key (ie. 8th bit shift). --CBF 8 April 1981 I got it and it works. -CBF 9 September 1981  File: TERMS, Node: Bantam, Previous: Ambassador, Up: Top, Next: CDC713 Perkin Elmer Bantam Status: not supported yet. Users: MIT Architecture Machine Group, LSP@ML, RK@MC Price: You should be able to negotiate $600 to $800 depending on quantity. Screen: 24 x 80; 12" diagonal; P4 phosphor Chars: 5 x 9 in a 7 x 10 field. Keybrd: Has at least 2 key rollover, fair feel, non-detachable Speeds: 110, 200, 300, 600, 1200, 1800, 2400, 3600, 4800, 7200, 9600 Has: Addr, Random, Tabs (fixed every 8), CLEOL Lacks: CLEOS, InsChar, DelChar, InsLine, DelLine, Overstrike Misfeatres: ? Basically a Fox with a somewhat better keyboard reengineered for cheaper mass production. Physically much smaller.  File: TERMS, Node: CDC713, Previous: Bantam, Up: Top, Next: C100 Control Data Corporation 713 Status: Supported by ":CRTSTY CDC713". Users: DPR Price: forgotten in the dawn of time Screen: 16 x 80 Chars: 7 x 9 with descenders Keybrd: rotten Speeds: 110,300 Has: Random, CLEOL, CLEOS, Standout Lacks: Addr, Tabs, InsChar, DelChar, InsLine, DelLine, Overstrike Misfeatures: 300 baud maximum speed! CBF 8 March 1978  File: TERMS, Node: C100, Previous: CDC713, Up: Top, Next: DD4000 Status: :TCTYP C100 Users: RJL@AI, GSB@ML, BYRON@ML, CEH@MC Price: 1-9 10-24 25-49 50-99 Educational discount $1575 $1418 $1339 $1261 equivalent to 1 quantity level. APL: $1750 $1575 $1488 $1400 APL really means overstrike Most MIT people have been quoted $1400 for the non-APL one right now. Screen: 24 x 80, P4 (white) phosphor, 12" diag 9"x6" tilt and swivel adjust. Chars: 8 x 12 in 10 x 12. The 7th and 8th columns of each character repeat what is in the 6th column. This can somewhat inhibit doing graphics. Keybrd: Detachable, has a clicker. Layout is close to VT52 with a big return key, Control where VT52 has Caps Lock. Most found feel acceptable, several thought it was very good. Speeds: 50,75,110,134.5,150,300,600,1200,1800,2000,2400,3600,4800,7200,9600 Has: Random, Addr, Tabs, CLEOL, CLEOS, InsLine, DelLine, Standout, 1 level Overstrike (only with APL option) InsChar, DelChar Misfeatures: Cannot keep up at 9600 baud. This is due to marginality of Z-80 DMA chip and may be fixed as of March 1979 or so. I believe HDS's stated policy is to replace it free upon request at that time. The list of misfeatures on this terminal has become a moving target, due to the policy of this company to try fixing them and often distributing new firmware releases free (they have claimed they will start charging a fee soon though). It actually has InsChar and DelChar, but they're very diffcult to use due to undesirable side effects. (BBN has been promised a fix for this, which presumably will find its way into the standard firmware releases). Operations such as InsLine need moderate (ie. 1/25 second) padding. Also it automaticaly CRLFs when you type in the last column (BBN fix may also cure this). It should be noted that CEH@MC has decoded their program and made his own mods. [The fixes I mention above about InsChar and last line autoCRLF have been fixed in current versions. The speed problem as far as I know has not been fixed as of Oct 1980. The terminal can in general keep up with 9600 baud text in ordinary output. If you have a window on the screen then it cannot.] Vendor: Human Designed Systems;3700 Market St;Phila, PA 19104;(215)382-5000 As of very late 1978 they were experiencing delivery problems due to non-availibility of 16K EPROMS and talking of going to ROM. [These have of course cleared up]. Must choose overstrike character set, so cannot overstrike APL and Ascii in the same terminal. However it works to overstrike underscores even on terminals without overstrike character set! Character set is still in PROM and they seem willing to assist in making own PROMs (SAIL character set for example). Up to 4 character sets including 1 overstrike character set can be accomodated. All terminal states and commands can be set under computer control or keyboard control. (ie. including things like baud rate). There is a key to show a terminal mode status display on the bottom line. There are no toggling states and there is a Reset All command. The normal commands are two character sequences starting with Escape, and do not seem to resemble any other terminal. There will apparently also be another model however which shares a common subset of commands with the VT-52, but they are not quoting any dates for availability of this. Among those not to be included is the hold-screen (mis)feature. Windowing: What you have is the ability to preface terminal commands with a window description (row, column, length, width). This defines the "active" window for all following terminal data. Apparently almost all operations are then local to that window (insert and delete char, insert and delete line, and scrolling and screen clear. Cursor addressing is also local to a given window.) But note, it does not remember any windows. To go back to a window you select it fully (with row, column, lenghth, width) again. Overstriking: There are 16 bits for each screen position. Normally this is divided into 7 bits to hold the "foreground" character, then 1 bit which says whether there is a "background" (overstruck) character or not. If not, then the other 8 bits are used for "display enhancement", which I believe are blinking, inverse video, intensity, underlineing, protected, and I guess a few others. If there is a "background" character, then then 7 bits hold the background character, the last bit is still available for inverse video however. Overstriking takes place when there is already a non blank or null character in a character position that is being inserted into. If there is already a background character, then it is replaced. (Ie. multiple overstrikes will end up with the first and last characters of a position overstruck.) Insertion of a space or null into a character position will erase it, whether there is a backgroup character there or not. Note that it knows enough to turn on the Underline bit upon an attempt to store an Underscore in an occupied character position, even if there are no overstrike character ROMs present. Function keys: Although you get 8 (shiftable) function keys (extendable to 14 for $50) you only get a total of 34 character total to assign to these keys. This can be improved by paying $140 for an extra "page" of memory which gives you an additional 128 characters in addition to another page of 24 lines of 80 chars. You can do this twice by getting two additional pages (thus a total of 256 characters more) beyond this you can get more pages of memory but no extra characters. [The multipage versions also provide a 256 character input FIFO which can be useful if your system doesn't have the padding forumulas exactly right (or doesn't try to pad at all). The single page versions only have 15 character FIFOs]. CBF 31 December 1978 CBF 29 October 1980  File: TERMS, Node: DD4000, Previous: C100, Up: Top, Next: DD5200 Delta Data 4000 Status: Supported by ":TCTYP SB" and ":CRTSTY DD4000". Users: BSG Price: dunno, probably too much. Screen: 25 x 80, green phosphor Chars: 5 x 7 with descenders. Keybrd: lots of keys on front, I think there is keyboard there someplace. Speeds: ...,9600 Has: Everything except: Lacks: Overstrike, kitchen sink. Misfeatures: Has about 40,000 switches and modes. If any are in the wrong position, things won't work right or won't work at all. Curiously almost identical in specs to Beehive SuperBee. Padding requirements are arcane and not fully documented, e.g. Clear to end of line seems to need some padding prior to the operation. Sometimes inserting characters has undesirable random side effect of sliding lines over further further down the screen. These lossages are also similar to Beehive SuperBee. CBF 8 March 1978  File: TERMS, Node: DD5200, Previous: DD4000, Up: Top, Next: DM1520 Delta Data 5200 Status: Supported by ":CRTSTY DD5200" (or NIH52 for the NIH version) Users: RICART Price: dunno, probably too much. Screen: 27 x 80 Chars: ? Keybrd: ? Speeds: ? Has: Everything except: Lacks: true CLEOL and CLEOS maybe. Misfeatures: Stores chars in an abominable way which requires it to generate spaces to fill up a line if you write something into an "empty" line. See CRTSTY source to see how misfeature is thwarted.  File: TERMS, Node: DM1520, Previous: DD5200, Up: Top, Next: DM2500 Datamedia 1520 Status: Supported by ":CRTSTY DM1520". Users: FJW ? Price: a little over $1000 Screen: 24 x 80 Chars: ? Keybrd: mediocre Speeds: ...,9600 Has: Random, Addr, Tabs CLEOL, CLEOS Lacks: InsChar, DelChar, InsLine, DelLine, Overstrike Misfeatures: There is an internal strap for having CR do a CR-LF (ie. a NewLine operation). CRTSTY assume it does this and compensates. Fairly reasonable vannila terminal. Has an APL model available for whatever that's worth. CBF 5 March 1978  File: TERMS, Node: DM2500, Previous: DM1520, Up: Top, Next: DM3025 Datamedia 2500 Status: Supported by ":TCTYP DM2500" and ":CRTSTY DM2500". Users: Very popular on West Coast; SRI, SAIL etc. Info here is by KLH. Particularly for SAIL support options contact ME@SAIL or TED@SAIL Price: The 2500 may be discontinued as such, but you can buy a 4000 series with a 2500 compatibility ROM. Also should look at 3025 and 3052 which are not program compatible with the 2500. Screen: 24 x 80 Chars: ? Keybrd: There are 2 generations. The later one is only poor, not terrible. Speeds: ...,9600 Has: Cursor, Addr, Tabs, CLEOL, CLEOS, InsChar, DelChar, InsLine, DelLine, Standout. See below for details. Lacks: Overstrike Misfeatures: Difficult to compute padding, hairy auto-newline. There is supposed to be a ROM available for the 3025 which effectively reduces padding time to nil. Talk to DANG@MIT-DM. See comments below. -- Detailed DM2500 doc -- The DM2500 is available with an Edit key which can be used as the Meta key with EMACS. You can't get this from Datamedia; you have to get it from Martin Hardy Product Associates 465 Convention Way, Suite 1 Redwood City, CA 94063 (415) 368-0640 The DM2500's main misfeature is the way it hacks "Roll mode" and automatic newline. Automatic newline is a feature whereby, if a move forward command (^\) or character is sent and the cursor is in the last column of a line, an automatic CR-LF is done (henceforth known as "auto-nl"). There are two interdependencies to watch for: [1] a CR (which does a newline) will not function immediately after an auto-nl! [2] a LF (move down) will not function immediately after an auto-nl OR a CR! Now, what happens when one of auto-nl, CR, or LF is done on the bottom line of the screen? That depends on whether or not the terminal is in Roll mode: a) NOT in Roll mode: Auto-nl or CR moves cursor to beginning of top line (home position), LF likewise wraps to top line but stays in the same column. b) IN Roll mode: screen scrolls upwards; top line is lost, and cursor either moves to beginning of new line (auto-nl, CR) or stays where it is (LF). Roll mode can be set and cleared either locally or remotely. DM2500 Control Codes: ====== ======= ===== ^Z = Up-line (if in first row, doesn't move) ^J = Down-Line (if in last row, goes to first row; but if Roll mode is on, screen scrolls up instead. LF will not function immediately after a CR or auto-NL) ^\ = Forward (if in last col, does an auto-NL with resulting side effects) ^H = Backspace (if in first col, doesn't move) ^I = Tab (this is used to move to next "variable field". no tab stops.) ^M = Newline (Not CR! Moves to beg of next line (see misfeature description)) ^B = Home ^L = Cursor positioning. Followed by X, Y coordinates (column, line) of the form #140, i.e. XOR the value with 140. ^W = Erase to End-of-line ^_ = Erase Screen - clears all "variable"-field data. If no protected fields are used, this is equivalent to Master Clear (but slower). On some DM's, this code is strapped to mean Clear-to-EOS, but not all. ^^ = Master clear - zap entire memory and home cursor. ^N = Blink On - all chars after this one will blink (see note below) ^O = Form On - all chars after this will be in a "protected field". Note that ^N and ^O do NOT affect characters already on the screen, rather they affect characters received by terminal chronologically after the mode is enabled. Thus to change an already-existing line to blink or form or back to normal the host must retransmit the entire line. (This makes editing at SU-AI, where line-editor is bold, slow!) Most SU-AI et al DM2500s are modified to have ^N and ^O swapped. ^] = Roll On - turn on Roll mode. ^P = Insert/Delete On - see below. ^X = CAN - turn off all of Blink, Form, Roll, and I/D. Insert/Delete Codes: ============= ===== ^P = Enter Insert/Delete Mode ^\ = Insert Character - create a space at current cursor position, moving following chars on the line one space to the right. chars moved out of the last column are lost. Cursor remains where it is, i.e. now pointing at the space. ^H = Delete Character - delete char at cursor position, moving following chars left by one. a space is shifted into the last column. Cursor does not move, i.e. now points at the char previously just to the right of the deceased char. ^J = Insert Line - insert line at cursor position, by moving current line and all succeeding lines down by one, creating a fresh, cmpletely blank line. The line shifted out at the bottom is lost. Cursor does not move, i.e. is now sitting in the blank line at same column position. ^Z = Delete Line - delete line at cursor position. All following lines move up by one, and a blank line is shifted in at the bottom of the screen. Cursor does not move. ^X = CAN - leave I/D mode (as well as any other) Some words about padding: A normal DM2500 only requires padding for the following: 1) Erase (^_, not ^^) - at 9600 baud, this code should be sent twice in succession; otherwise just once is sufficient. This is sort of a pseudo-pad. 2) Insert Char - At 9600 baud, one rubout should be sent after each ^\. At lower speeds none are needed. 3) Delete Char - At 9600 baud, one rubout should be sent after each ^H. At lower speeds none are needed. 4) Delete Line - At 9600 baud, one rubout after each ^Z. At lower speeds, none needed. 5) Insert Line - the slowest. Here is an empirical table of # rubouts needed after each ^J: Speed # pads 9600 15 4800 7 2400 3 1200 1 6) Tab - If you really want to use this, you should get a Datamedia Technical Manual and learn about the protected vs. variable field hackery. NOTE: on some older model DM2500's, Insert Line needs one more rubout than those given in the table (possibly two), and Insert Char is unreliable in the first 11 or so lines of the screen if running at 9600 baud, no matter how many pads are given. This is a design flaw with respect to the memory board that was corrected on later DM's by redesigning it to halve the access time for I/D mode. FURTHER NOTE: on these losing DM2500's a frob that lets you win with Insert Char is to send a Space (40) instead of a FS (^\). This shifts the line properly, but has the side effect of moving the cursor and can leave garbage in the position spaced over, so after leaving I/D mode one must backspace over and write spaces or something there. The number of rubouts needed for Insert Line is really dependant on the current line position, here is a formula that seems to work: ((30.-VPOS)*OSPEED-2400.)/14400.  File: TERMS, Node: DM3025, Previous: DM2500, Up: Top, Next: DM3052 Status: VT52 subset compatible, support with ins/del stuff via :CRTSTY DM3025 Users: Furst Screen: 24 x 80 Chars: Price: $1750 range. Keybrd: Seems to have N-key rollover. Some feel, but they squirm a lot. (Is like the newer DM2500 keyboards). Speeds: ...,9600 Has: Cursor, Addr, Tabs, CLEOL, CLEOS, InsChar, DelChar, InsLine, DelLine, Standout (inverse, half intensity, blinking & all combinations). Lacks: Overstrike Misfeatures: Line/Ins delete can take up to 130 MS. This is the part of the new Datamedia line. CBF 6 July 1978 It may be possible to get this terminal with an Edit key which can be used as the Meta key with EMACS, if you ask Martin Hardy Product Associates 465 Convention Way, Suite 1 Redwood City, CA 94063 (415) 368-0640  File: TERMS, Node: DM3052, Previous: DM3025, Up: Top, Next: DM4000 Status: VT52 subset compatible. Users: Dang@DM had a demo. Screen: 24 x 80 Chars: Price: $1745, seems hard to get a discount. Keybrd: Seems to have N-key rollover. Some feel, but they squirm a lot. (Is like the newer DM2500 keyboards). Speeds: ...,9600 Has: Cursor, Addr, Tabs, CLEOL, CLEOS, InsChar, DelChar, InsLine, DelLine, Standout (inverse, half intensity, blinking & all combinations). Lacks: Overstrike Misfeatures: Line insert/delete is fairly slow (130 ms), and the terminal will spew out ^S^Q, even if not in hold screen mode when it gets behind. If padded properly this lossage should not occur. This is the part of the new Datamedia line. CBF 6 July 1978  File: TERMS, Node: DM4000, Previous: DM3052, Up: Top, Next: DM-DT/80 Status: Supported as DM2500 if using 2500 PROM. Otherwise random, since terminal is user programmable. Users: KLH (sort of) Screen: 24 x 80, but can be 48 x 80 Chars: same as DM2500 Price: $2500, the Dm2500 simulating PROM is $135 and 48 line option is $250. An additional 8K of user RAM is $425. Keybrd: Like Dm2500, adequate but no n-key rollover. Can be obtained with SUMEX keyboard, which has an EDIT shift-like key to set parity bit. Comments: Will stick in more stuff if I manage to use one regularly. KLH 3/20/78 To get the Edit key, which can be used as the Meta key with EMACS, contact Martin Hardy Product Associates 465 Convention Way, Suite 1 Redwood City, CA 94063 (415) 368-0640  File: TERMS, Node: DM-DT/80, Previous: DM4000, Up: Top, Next: FOX Datamedia DT/80 Status: Complete VT-100 clone, down to the bugs Users: ? (see below) Has: Advanced Video Option is standard, auxiliary printer port standard Price: Computrend, Boston area vendor, quotes $1899 ($100 less than DEC's 'standard' price) (one site was offered $1672 in 25 quantities); note that the DT-80 includes the Advanced Video Option, which normally costs another $270, so the cost (list) comparison should be $1899 to $2170. (Whoops, Computrend is not going to offer them anymore.) Screen: larger than VT-100, characters more readable (probably just size made the difference) Keybrd: Typical Datamedia, physically tough with 3-4-key rollover; reviewer thought it actually was nicer than VT-100 but others have expressed the opposite opinion. Keyboard seemed to generate spurious chars once in a while and be over-sensitive (ie, just resting fingers on some keys would send them once in a while). Enclosure: Typical Datamedia: metal (much more sturdy than VT-100) -- see VT-100 node for rest of technical details -- I had one of these for a day (though it would flake out totally after being on for an hour, sigh, but this was a demo model so it's allowed to have some marginality). The above remarks are from that day, so take them with a sack of salt. No one really has any experience with them over any period of time, as far as I know; when someone gets some, please replace this summary. CPR, Tuesday, 8 July 1980  File: TERMS, Node: FOX, Previous: DM-DT/80, Up: Top, Next: Freedom-100 Perkin-Elmer FOX-1100 Status: Supported by ":CRTSTY FOX". Users: MIT Architecture Machine has several. Try ARQMAQ@MIT-ML . Screen: 24 x 80, P4 (white) phosphor Chars: 7 x 11 (in a 9 x 12 field) Price: Probably around $1200. The P-E Bantam effectively replaces the FOX. Keybrd: 1 key rollover sucks. Speeds: 75,110,200,300,600,1200,1800,2400,7200,9600 Has: Cursor, Addr, Tabs, CLEOL, CLEOS. Lacks: InsChar, DelChar, InsLine, DelLine, Standout, Overstrike. Misfeatures: Seems unreliable? Upper case DEL key. Vendor: Perkin Elmer Data Systems;Randolph NJ 07801;the cretins don't believe in phone numbers; They do have a Boston office. Comes within 2 inches of being defacto VT-52, HP compatible. Uses ESC K to mean clear entire screen, clear all tabs and home, instead of Erase to End of Line, and uses Esc I to mean CLEOL instead. See also big-brother OWL-1200, also see its replacement: the Bantam. CBF 5 March 1978 CBF 23 Sept 1979  File: TERMS, Node: Freedom-100, Up: Top, Previous: Fox, Next: H19 Freedom-100 Freedom 100 Status: Informational. Users: Faletti@Berkeley (ucbvax!faletti) Price: List: $595.00, cheapest seen so far: $550. Comes with 6 month warranty and additional 1 or 2 yrs for $60/$90. Screen: 12", green phosphor, 24x80 plus status line, tilt from 5 to 15 degrees. Chars: 7x9 in 9x12 with fairly large "open" and solid look. Cursor may be a slow-blinking or solid block, or a fast-blinking underscore. Keybrd: Detachable, cord is like curled phone cord with modular attachment at terminal. Good feel -- some would say "too springy". Typeamatic (auto-repeat) keys built with sealed plunger type (like dm1500) so less likely to get dusy. 2-char rollover. Typewriter paired. Keyboard map below. Key unusual differences: large return key right next to quotes; large shift-lock which locks down; arrow keys in a line above -/_ =/+ `/~ instead of on numeric keypad. Speeds: 110 - 19200, (same for printer port). Has: Random direct or Relative cursor addressing, settable tabs with bactab, CLEOL, CLEOS, InsChar, DelChar, InsLine, DelLine. Any combinations of reverse video, blinking, half-(actually- quarter)-intenstity, underline, (blank, write-protected). 10 shiftable function keys sending ^A X for X ranging through 20 characters. Also has reverse/normal-video for whole screen allowing a visible bell without destroying other features. Lacks: Overstrike Misfeatures: Status line is "dumb", not allowing standout or random addressing or even bells or tabs! Escape sequences are not the ANSI standard but rather imitate the Televideo/ADM style, but DO NOT have the "magic cookie bug" i.e., no screen space is needed to switch display attributes. The bell is too loud but can be turned off by the host. Padding requirements for Unix-VAX-Emacs (Gosling's) at 9600 baud are at least: insert-line: 8.5 ms per line. delete-line: 11.5 ms per line. insert-mode: 6 ms per char. This still fails sometimes so it needs work somewhere. No padding needed with VI at 1200 baud but 6.5, 11.5, 0 needed at 9600 baud. Keyboard Map: ---------------------------------------------------- ------------------ off clr eop del del line f1 f2 f3 f4 f5 f6 f7 f8 f9 f10 ^ v < > home clr line char print eol ins ins ---------------------------------------------------- ------------------ ! @ # $ % ^ & * ( ) _ + ~ esc 1 2 3 4 5 6 7 8 9 0 - = ` bs brk 7 8 9 - ---------------------------------------------------- ------------------ bcktb { } RET -tab- q w e r t y u i o p [ ] RET rub 4 5 6 , -------------------------------------------- RET ---- ------------------ SHIFT : " RETURN | EEE CTL -LCK- a s d f g h j k l ; ' RETURN \ 1 2 3 NNN ----------------------------------------------------- -------------- TTT -SHIFT- < > ? -SHIFT- ---0--- EEE -SHIFT- z x c v b n m , . / -SHIFT- LF ---0--- . RRR ------------------------------------------------ ------------------ | SPACE BAR | ------------------------ Terminal just became available in late 1982. This description was written in December 1982 based on 3 weeks' experience with serial number 480 (mostly at 1200 baud) so no information is available on reliability yet. Joe Faletti (Faletti@Berkeley, ucbvax!faletti)  File: TERMS, Node: H19, Previous: Freedom-100, Up: Top, Next: H80 Heath H19 Status: Supported by :TCTYP H19 or :CRTSTY H19 TCTYP support does not know use Ins/Del Char. Users: Many. H19-People at MC is a mailing list of users. Price: $995 assembled and test. $695 kit. 10% off quan 10 from Heath. Screen: 24 x 80;25th "status" line. P4 white phospher 6.5"x8.5" (12" diag) Chars: 5x7 Upper case, 5x9 lower case with descenders, 8x10 for 33 graphics Keybrd: Non-detachable, VT52 layout, unremarkable tactile feel. Clicker is more of a sharp beep than a click. Speeds: "110 to 9600" Has: Random, Addr, Tabs (fixed every 8), CLEOL, CLEOF, Ins Line, Del Line, Ins Char, Del Char, Standout (Reverse video) Lacks: Overstrike Misfeatures: Included in description below. The following material is supplied by Winston Edmond : * It has a 128 character FIFO receive buffer. If this buffer gets more than 96 characters full, the terminal sends XOFF, and sends XON when the buffer drops below 96 characters again. I did not see a way of disabling this "feature". (Of course it is a kit, and you have the schematics...) [On most terminals this would not be so bad as long as one was careful to properly pad the terminal. However, the H19 inserts DEL codes into the buffer. Thus the the character that many systems use for padding whill only serve to exasperate the problem of triggering an XOFF since the FIFO will fill with DEL's. This can be cured by having the host pad with NULs which are not inserted into the FIFO. Unfortunately not all system allow the user to set the pad character. -CBF 20 Dec 79] * The following terminal properties can be changed: Key clicks/no key clicks. The click is medium loud for my taste. I would bet that there is some easy way to adjust the loudness. Cursor: underscore/box/none at all. Cursor always XORs with character cell. Underscore is at typewriter level and intersects with lower case descending letters. * 25th line can be used as mode line. Scrolling, insert/delete line, erasing, etc. do not affect this line. One uses this line by using cursor positioning to get to it, whence operations apply only to the 25th line. Use cursor positioning to return. When disabled, it is always blank. When enabled, it is always initially blank. * Keyboard has a acceptable feel to it. It is lighter than a VT52 but heavier than a Concept-100. For comparison, I am a touch typist and like the Concept-100 keyboard. * Alternate keypad mode (same as VT52 and sends same character strings.) * Scrolling * Reverse scrolling * Hardware tabs every eight positions, except from 72 to 80 where TAB is treated like space. * Communications parity: Even, odd, always 1, always 0. * Control-space does not produce NUL. The Control key only applies to "strictly legal" keys. Otherwise, you get the normal character. * There are no programmable function keys, and none of the function keys happen to transmit NUL. You do have ESC followed by A, B, C, D, E, H, J, L, M, N, P, Q, R, S, T, U, V, W, or @ available from function keys. If you use the alternate keypad mode (a la VT52) you get ESC ? for = {M,n,p,q,r,s,t,u,v,w,x,y}. If you leave the terminal in ANSI mode, you usually get ESC O instead, but you may get ESC [ for four of the function keys. * Although the terminal has both "Heath" mode and ANSI mode, sending an ANSI insert-N-lines causes the terminal to insert one line N times. (Much slower, less aesthetic.) * The terminal is Z80 based and uses two 2716 EPROM and two 2316 (8316) 2K*8 ROMs. It also uses a 5740 keyboard encoder and a 6845 CRT controller. There was apparently no truth to the rumor that microcode listings were available from Heath, but the question is still being pursued. * Some people at Tech Square say that an order of 10 or more gets them a 10% discount on the price. * There is no fan. (However, like all TVs, there is a very high-pitched note, no worse than my Concept-100.) This terminal is also available as part of the H-89 "All-In-One" Computer. This adds a second Z-80 processor and 16K RAM (expandable to 48K) and sells for $1250.00, kit form. With a minifloppy, 1650.00, or around $2250 assembled. The initial announcement of the H19 claimed it would be field upgradable to an H-89, however this upgrade does not seem to have been marketed. I would appreciate hearing from anyone who becomes aware of such a kit being sold. One gets an assembler for the Z-80 as well as BASIC and whatever other programming languages they offer. CBF 11 June 1979 CBF 10 August 1979 (mostly integrating Winston Edmond's comments)  File: TERMS, Node: H80, Previous: H19, Up: Top, Next: H1500 Hazeltine Executive 80 models 20 and 30 Status: new product Users: ? Price: $1295 (model 20 12" screen 80 columns non-detached keyboard) $1695 (model 30 15" screen 132/80 columns detached keyboard) Screen: 24x80; 25th status line. 12" diagonal (9"x6") Enhanced video option: (24x80|24x132)+25th status 15" diag (10.5"x8") Green or yellow(?!) phospher, 60Hz non-interlaced refresh (18MHz) Chars: 7x8+descenders in a 9x12 block or 5x7 in 132 column mode .082"(2mm) wide x .160(4mm) high on 12" 15": .102"(2.6mm)x.160"(4mm) or .057"(1.4mm)x.186"(4.7mm) in 132 mode Keybrd: detached on model 30, option on model 20; typewriter paried, selectable clicker, decent feel; numeric pad, cursro keys etc. 8 more function keys on model 30 Speeds: 110,300,600,1200,2400,4800,9600,19,200 optional RS449, optional 20ma Has: Random, Addr, Tabs, CLEOL, CLEOS, InsChar, DelChar, InsLine, DelLine, Standout (inverse, blink, underline, dim) Lacks: Overstrike Misfeatures: Unknown, no sample available. Aux Port: optional RS232 printer on mod 20, optional buffered RS232 on mod 30 Selectable Hazeltine standard (like H1500??) or ANSI 3.64 control sequences. It looked ok, at the NCC, I'd be interested in hearing from anyone who gets one. Tilt display standard on model 30 optional on model 20. The big (15") screen gives good size characters for those tired of squinting; Enhanced video option also offers double height/width chars. Both screens appeared to have good video quality. There also seems to be a "horizontal split screen" feature that apparentely lets you scroll the right and left siedes separately. Its not clear is this stops a clear to end of line or even an insert delte line for that matter. Selectable smooth scroll. No flow control option apparent, except perhaps Xon/Xoff on model 30. 21 June 1980 CBF  File: TERMS, Node: H1500, Previous: H80, Up: Top, Next: H2000 Hazeltine 1500 Status: Supported by ":CRTSTY H1500". Users: RP Price: $1100 single Screen: 24 x 80;P4 white phospher;10 inch diag, 6 x 9 inch display area Chars: 7 x 10; .204" high, .088 wide Keybrd: Unusable, funny layout, mediocre feel, supposedly detoriates with time. Speeds: 110,300,1200,1800,2400,4800,9600,19200 Has: Random, Addr, CLEOL, CLEOF, Ins Line, Del Line, Standout(hi/lo intensity) Lacks: Tabs, Ins Char, Del Char, Overstrike. Misfeatures: No possible way to display tilde "~", as this prefixes control codes. Note: Avoid the H1400, it has no features. The Stanford LOTS people have rewritten the firmware of the H1500 to emulate a Datamedia 2500. This doesn't fix the keyboard, but can get you Char insert and delete and the abilitiy to display a tilde. Contact J. Q. Johnson (Admin.JQJ at SU-Score) for more info. CBF 5 March 1978  File: TERMS, Node: H2000, Previous: H1500, Up: Top, Next: HP2621 Hazeltine 2000 Status: Supported by ":CRTSTY H2000" or by ":CRTSTY H2LOSE". Users: ? Price: Varies, used to be expensive (ie. $3K), not really currently marketed. Screen: ? Chars: ? Keybrd: Sucks by all accounts. Speeds: ? Has: Random, Addr, Ins Char, Ins Line, Del Char, Del Line, Standout? Lacks: CLEOL, CLEOS, Overstrike. Misfeatures: Galore. Keyboard is supposedly arcane in that it can generate lower case, but not by any simple means like a shift key. has strange padding requirments, but they appear to differ from instantiation to instatiation from the terminal, so CRTSTY may not work for yours. Documentation probably corresponds to no real instance of the terminal. CBF 5 March 1978  File: TERMS, Node: HP2621, Up: Top, Previous: H2000, Next: HP2640 Hewlett-Packard 2621 Status: Supported by :TCTYP HP and :CRTSTY HP2621 Users: WRS@OFFICE-2, you may wish to read his 9 page evaluation. Price: $1450 single quantity, $2550 with a 120cps thermal printer. Screen: 6"x8.5"; 24x80; P4 (white) phosphor;60 Hz refresh;non-interlaced Chars: 7 x 9 in a 9 x 15 cell by use of descender and half shift indicators. Keybrd: Typewriter paired (but rubout is shift-escape), detached, N-key rollover. Feel is not impressive, but much better than HP 2640 series. Caps lock key in usual place for shift lock toggles caps lock mode, but gives absolutely no indication of state, such as latching or setting an LED as one might expect. Speeds: 110,150,200,300,600,1200,1800,2400,3600,4800,9600 (cannot split speed) Has: Random, Addr, Tabs, CLEOL, CLEOS, InsChar, DelChar, InsLine, DelLine, Standout Lacks: Overstrike Misfeatures: RUBOUT is SHIFT-ESC, caps lock key lossage. The terminal powers up to settings taken from a settable non-volatile memory. Said to have nice setup mode. Tabs are not in non-volatile memory. There are no straps or switches (except the power switch of course). Optional thermal printer is integral to case. Terminal sits on pedastal, but takes up considerably more desk space than a Concept 100 or VT 100. EAK 28 November 1978 CBF 20 March 1979 CBF 11 June 1979 CBF 3 February 1980  File: TERMS, Node: HP2640, Previous: HP2621, Up: Top, Next: HP2645 Hewlett-Packard 2640A, 2640B Status: Supported by ":TCTYP HP" (padding will not be sufficient at 240cps) and ":CRTSTY HP2640". Users: 2640A: DPR Price: $2640 Screen: 24 x 80, white on grey. Chars: 7 x 9 in a 9 x 15 cell by use of descender and half shift indicators. Keybrd: plastic on plastic no tactile feel, poor layout 2640A only. Detachable. Speeds: 110,150,300,1200,2400 Has: Random, Addr, Tabs, CLEOL, CLEOS, InsChar, DelChar, InsLine, DelLine, Standout Lacks: Overstrike Misfeatures: Maximum speed of 2400 baud, and needs a fair amount of padding even at that speed. 2640A often does not manage to pay attention to its keyboard when it gets busy. Lowercase ROM is $100. Some degree of user programmability. They can all load a pong game. Other members of family include the 2648 graphics terminal (360x720 raster scan), models designed for much more user programmability, 2641A model for APL. CBF 5 March 1978  File: TERMS, Node: HP2645, Up: Top, Previous: HP2640, Next: IBM3101 Hewlett-Packard 2645A Status: Supported by ":TCTYP HP" and ":CRTSTY HP2645". (CRTSTY will use direct cursor addressing) Users: MSB, CBF, EAK Price: $3500 Screen: 24 x 80, white on grey. Chars: 7 x 9 in a 9 x 15 cell by use of descender and half shift indicators. Keybrd: plastic on plastic no tactile feel. Detachable. Speeds: 110,150,300,1200,2400,4800,9600 (options for split & more speeds) Has: Random, Addr, Tabs, CLEOL, CLEOS, InsChar, DelChar, InsLine, DelLine, Standout Lacks: Overstrike Misfeatures: Upper case Delete key (reversed by CRTSTY) Lowercase ROM is $100. Some degree of user programmability. They can all load a pong game. Other members of family include the 2648 graphics terminal (360x720 raster scan), models designed for much more user programmability, 2641A model for APL ($4100 !) CBF 5 March 1978  File: TERMS, Node: IBM3101, Previous: HP2645, Up: Top, Next: IMLAC International Business Machines 3101 Status: New, supported by :CRTSTY IBM3101 Users: None Price: Model 10: $1295, Model 20 (block mode): $1495 Discounts: 10% Educational; quantity: 25-49 7.5%, 50-99 15%, 100+ 20% Screen: 24x80 with 25th status line, P39 (Green) phospher, 15 inch diagonal Chars: 7x14 in a 9x16 field. Very well formed. Keybrd: Detachable, good tactile field with breakthrough and clicker (can be disabled). Alphabetics are in usual office Selectric positions, but other keys are where they have never been before. Control is labelled ALT and sits to the right of the space bar. Speeds: 110,150,200,300,600,1200,1800,2400,4800,9600. Some combinations of split speeds seem possible; Current Loop and RS-422 options available Has: Random, Addr, Tabs, CLEOL, CLEOS, [Standout Model 20 and above only.] Lacks: Overstrike, [Ins Line, Del Line, Ins Char, Del Char availalble on Model 20, but they only work in block mode!] Misfeatures: Unusual Keyboard (placement of Control key makes it almost useless) The report of the Insert/Delete functions only working in block mode comes from J.Q. Johnson (Admin.JQJ at SU-Score). He also claims they may fix this in later releases. This is IBM's first ASCII terminal. Control codes are similar to the VT52, but the terminal cannot succesfully masquarade as a VT52 (row and column reversed in absolute cursor position, Escape I for clear to end of line instead of Escape K, etc). Insert and Delete Line operations require 50-100 milliseconds of padding on those models which have them. Mechanical packaging is very nice. The unit comes in 3 pieces; the keyboard, the electronics box and the monitor, connected by cables. The monitor can loosely sit in any of 6 positions on the electronics box (facing left, center or right while pointed level or upward). The electronics box and the keyboard are roughly similar in shape, each being almost 2 feet wide (this is from memory) 3 to 4 inches high and about 10 inches deep. The monitor is about a foot by a foot by a foot. One of their brochures claimed the monitor refreshed at 70 Hz, interlaced. This could be a typo. However, in spite of it being P39 Green phospher, no one could really detect any smearing when it scrolled. I suspect they may be doing interesting things with the monitor. The salesman claims he was told the monitor took a standard composite video signal. It does not use a standard video cable for connection, though. The Models 10 and 20 are RS-232 only. Models 12 and 22 ($25 extra) have both RS-232 and 20MA current loop. Models 13 and 23 ($25 extra) have both RS-232 and RS-422. RS-422 is the EIA standard intended to replace RS-232. IBM claims you can drive RS-422 4000 feet (1219 meters) without modem at 9600 baud. There is evidence other companies, such as DEC, intend to support RS-422. Field upgrades are possible throughout the models by purchasing appropriate replacement assemblies. Deliveries of Model 10 and 12 start November 1979. You're too late for first day orders already. Model 13 January 1980, Models 20, 22, and 23 March 1980. They do not seem interested in on-site maintainance. Return defective subassembly to factory contracts are available at a cost of $70 per year for the models 10, 12, and 13 and $80 per year for models 20, 22, 23. If not on warranty for a minimum charge of $64 per machine to cover handling, inspection, cleaning, adjustments, testing, and return shipping charge will be applied. In addition all parts will be billed at IBM's prevailing parts prices and additional time required for repairs will be billed at IBM's applicable service rates. I have not yet discovered what the initial warranty period is. The customer is expected to install the machine himself. Component prices are as follows: Monitor assembly: $315 Keyboard: $320 Logic (Models 10,12,13) $660, $685, $685 Logic (Models 20,22,23) $860, $885, $885 Some personal impressions: For IBM the terminal is aggressivley priced. The packaging, monitor and keyboard are clearly all more expensive that a VT-100 for a example, but the terminal is cheaper. Emacs users, and just about any one who uses the Control key extensively is going to find it difficult to adjust to the keyboard. This is not the only strange thing about the layout. Escape is in a separate group to the left, delete in a separate group to the right. Note that most previous IBM displays (like the 3270) are connected to the main CPU by coax cable. They have never really produced a display terminal intended for dialup use the way the 2741 was intended for dialup use. In this light, it is more understandable to view the 3101 as an attempt by IBM to sell dialup display terminals (which many of their customers were buying elsewhere anyway, hence the capitulation to ASCII?) than an attempt to attack a traditional DEC market, for example. CBF 13 October 1979 CBF 20 January 1981  File: TERMS, Node: IMLAC, Previous: IBM3101, Up: Top, Next: I100 Imlac PDS-1, PDS-4 Status: Supported by ":TCTYP IMLAC" if you're running SSV or SST. Users: MIT-DM, Plasma Physics (PLASMA-GROUP@MC) Price: PDS-1G is minimum $9K all the way up to PDS-4 with options >$20K Screen: PDS-1 gets 44 lines by 88 chars with standard font. PDS-4 gets 50 lines by 89, but can hack smaller fonts yet. Chars: Stroke drawn, you need the long vector option! Keybrd: no great feel, but stands up to an amazing amount of abuse. It takes a lot of multiplexing to generate all 128 Ascii codes. Speeds: I have heard of Imlacs going 80Kbaud and better. Has: Everything. Lacks: If it doesn't have it, you can always write your own program. Misfeatures: Expensive. It might be handy to attach your own keyboard. Features are based on use of the SSV or SST program of the MIT-DM group. Since the display processor is a full minicomputer, you can of course run any sort of program you wish in the terminal. In particular there is an excellent set of games available. CBF 8 March 1978  File: TERMS, Node: I100, Up: Top, Previous: IMLAC, Next: I400 Infoton 100 Status: :TCTYP VT52,+%TOLID,PADTAB 1 Users: Price: $925 single?, $625 quantity 100. $50 for char ins/del. Screen: 24 x 80 with 25th status line. 12" diagonal with P4 white phosphor. Chars: 5 x 7 upper case, 5 x 9 for lower case. Keybrd: Has N-key rollover. Keys have rather stiff springs, poor feel. Speeds: 110, 200, 300, 1200, 2400, 4800, 9600, 19200 Has: Random,Addr,Tabs,CLEOL,CLEOS,InsLine,DelLine,Standout (reverse and low intensity). InsChar,DelChar $50 options. Lacks: Overstrike Misfeatures: Lower case "l" and the number "1" have graphics easily mistaken for each other. Has a logical not symbol instead of underscore!  22 February 1979, CBF  File: TERMS, Node: I400, Up: Top, Previous: I100, Next: LP Infoton 400 Status: ? Users: Radia@bbn, MILES.BAREL@cmu Price: 400/4 $1330 (quantity 1) 400/5 $1460 (quantity 1) Screen: 25 x 80 Chars: 9x9 (over I believe 10x10 area) Keybrd: Reasonable (Keytronics - same as superbee) Speeds: ...,19200 Has: Random,Addr,Tabs,InsChar,DelChar,InsLine,DelLine,Standout Lacks: CLEOL,CLEOS,Overstrike Misfeatures: The following is the text of a letter sent to Infoton: Here is a description of the problems we encountered with the Infoton 400. The first set of problems are general ones which are application independent. 1. When you power up the terminal, it comes up in block mode with caps only set. This may be intentional but I would think it would be preferred that neither of these be set. 2. When entering text onto the screen, either from the host computer link or from the keyboard, there is noticeable flicker of the text already on the screen. 3. The underline and the cursor obscure the descenders, making these characters hard to read, and sometimes impossible to differentiate. For instance, you cannot tell the difference between a lower case Q and a lower case G if the cursor is beneath the character. 4. The insert and delete character functions shift all characters on following lines along with the line being modified, making these functions useless. 5. The terminal appears to have either a parity or overrun problem. When used with several of our computers, many characters echo as question marks, instead of the character sent by the host, irrespective of the parity settings in the rear of the terminal. The remaining problems are specific to our application - a screen editing terminal. 6. The special graphics functions of underline, blink and reverse video cannot be used in arbitrary combinations. 7. Insert and delete character functions move characters over the graphics formatting. This may be fine for data entry applications where the formatting indicates different fields, but for our application, where the graphics are used to indicate different character fonts, the graphics must remain with the text which is in the alternate font. 8. The use of the graphics functions require insertion of a blank character on the screen at the position that the mode was changed. We often use alternate character fonts within one word to emphasize portions of that word. For example, the first several letters of a command are often an ACROnym for its use. Miles Barel@CMUA 23 August 1978 CBF transcribing Miles' message.  File: TERMS, Node: LP, Previous: I400, Up: Top, Next: MB NLS Lineprocessor This refers to the so-called Lineprocessor that is used with NLS workstations and will allow use of ITS without having to unplug the cretinous little black box. If anyone is actually interested in the specs, I can set them straight. In general a terminal being used as a LP will run much slower than it otherwise would be capable of. KLH 3/20/78  File: TERMS, Node: MB, Previous: LP, Up: Top, Next: MIME Beehive Mini-Bee 0 and 4, probably also applicable to B100. Status: Supported by ":TCTYP MINIBEE" fairly well, ":CRTSTY MB0" or ":CRTSTY MB4" supports more features. Users: ? Price: ? Screen: 25 x 80 Chars: ? Keybrd: Reasonable. Speeds: ...,9600 Has: MB4: Random, Addr, Tabs, CLEOL, CLEOS MB0: Random, Addr. Lacks: MB4: Ins Char, Del Char, Ins Line, Del Line, Standout, Overstrike. MB0: Everything. Misfeatures: Upper case shift. Probably many more misfeatures hiding in non-obvious places waiting to be discovered. CBF 5 March 1978  File: TERMS, Node: MIME, Up: Top, Previous: MB, Next: MOD1 Micro Term MIME Status: :TCTYP VT52 in VT52 mode or :CRTSTY MIME in ACT mode uses line ins/del Users: Emery@MC, CSD.Halbert@Berkely Price: $795 single, $825 with non-glare screen (recommended) quantity discounts: 10-24: 10%, 25-99: 25%, >100: 27% Screen: 24 x 80, 12" diagonal white phospher (presumably P4) Chars: 5 x 7 plus descenders Keybrd: Claim they use same kbd as VT52. No clicker though. Speeds: 110, 150, 300, 600, 1200, 2400, 4800, 9600 Has: Random, Addr, Tabs, CLEOL, CLEOS, InsLine, DelLine, Standout (half intensity) Lacks: InsChar, DelChar, Overstrike This terminal has internal switches that can be set to have it "emulate" either an ADM3A, a Hazeltine 1500, a VT52, or an ACT-IV. EMERY@MIT-MC 02/23/79 01:47:48 Re: Mime-I It also supports a very nice little switch inside so that when emulating a terminal that doesn't have a given feature, it adds it. For example, when in VT52 mode, it will normally not understand a ins/del line. But, if this switch is turned on (appropriatly enough, it is named Enhance), it will. Of course, the assignment for these extra features [in these modes corresponds to the assignment of them on the ACT series terminals]. It has the same quark as the ACT-V does about underlining a character that has descenders. It has the characters transposed on the keytops for underline and minus, and the underline is actually a left arrow (presumably since it supports underline as a special feature anyways). All in all, it is a really good terminal. Having an F-8 in it, and 3 EPROMS (2708's) to choose between the different terminals (ACT-IV is hardwired) makes it very versatile for the hobbyist. For some strange reason, you have to open the terminal in order to switch from one terminal setting to another. This is presumably since it isn't changed so often. However, it would be much more useful if it was accessible from either the keyboard or behind the terminal. Emery Galambos 25 February 1979 The follow report came from DCH@MIT-ML on 27 February 1979: The main idea of this terminal is that it will emulate an ADM-3A, a VT-52, a Hazeltine 1500, or a Microterm ACT-IV, all at the flip of a switch. It has remote ins/del line, but not ins/del char. However, we have been told that for free they will substitute a ROM so that it will no longer have the emulation feature, but will do ins/del char. When not emulating something, all function codes are single control characters, not escape sequences. This is a minor problem on UNIX, which runs full-duplex, but stupidly echoes everything anyway, unless one is in raw mode. but on any reasonable full-duplex system, this should not be a problem. Ins/del line causes the screen to glitch. It needs about 80ms padding. This means that several in a row causea lot of action on the screen. But it works. The cursor is a double underline, non-blinking, and sometimes hard to find. When the cursor is on a character with a descender, the screen goes double- bright where the cursor and descender overlap. This looks ugly, and means they did it wrong. The terminal has half-bright and underline. To underline a char, one positions the cursor and then sends a ^U. The underline appears in exactly the same form as the cursor, with the same double-bright problem. but various operations, such as linefeeding over a line with underlines (I'm not sure about this) cause the underlines to disappear. So this is bad. The keyboard is ADM-3Aish, but slightly softer, and with a different layout (but RUBOUT is still shift-_, albeit now near the upper RH corner. It does not have n-key rolloover. There is an embeedded numeric keypad. This is the MIME-I. The MIME-II is the same except it comes in a larger and slightly nicer-looking box, and has a separate numeric pad. It also costs more. Underline is backarrow, but uparrow is caret! This is the terminal we are most seriously considering, because it has ins/del char/line (we will get the substitute ROM), and because it is cheap, has an OK keyboard, and its problems (e.g. double-bright, funny underline) do not make it unusable at all, but are just annyoing. If one of us gets up the energy, we may buy the tech manual ($25), and do some reprograming. But I doubt it.  File: TERMS, Node: MOD1, Up: Top, Previous: MIME, Next: OMRON Hazeltine "Modular One" Status: Supported by ":CRTSTY MOD1". Lacks: Clear-to-EOL, hence line erasure is slow. Comments: CRTSTY works all right with it - I had access to one for about a month, and have never seen any since. As far as buying one goes, don't. KLH 3/20/78  File: TERMS, Node: OMRON, Up: Top, Previous: MOD1, Next: OWL Omron 8025. Status: Supported by ":TCTYP OMRON" and ":CRTSTY OMRON". CRTSTY supports more features, but neither works quite right and wont until we get our hands on one. The current stuff seems to be called 8030, and seems to be different and thus not supported.  File: TERMS, Node: OWL, Up: Top, Previous: OMRON, Next: SB Perkin-Elmer OWL-1200 Status: Supported by ":CRTSTY OWL". Users: See the FOX Screen: 24 x 80, P4 (white) phosphor Chars: 7 x 11 (in a 9 x 12 field) Price: $1500 through Educom (MIT included), about $2000 otherwise. Keybrd: 1 key rollover sucks, assuming it is like the FOX keyboard. Speeds: 75,110,200,300,600,1200,1800,2400,7200,9600 Has: Cursor, Addr, Tabs, CLEOL, CLEOS, InsChar, DelChar, InsLine, DelLine, Standout Lacks: Overstrike Misfeatures: Upper case RUBOUT key. Vendor: See the FOX Comes within 2 inches of being defacto VT-52, HP compatible. Uses ESC K to mean clear entire screen, clear all tabs and home, instead of Erase to End of Line, and uses Esc I to mean CLEOL instead. See also little-brother FOX-1100. EAK 11 March 1978  File: TERMS, Node: SB, Up: Top, Previous: OWL, Next: SWTPC Beehive Super-Bee There are two models, the SB1 and the SB2. The SB2 has some a new PROM to fix certain bugs. Status: Supported by ":TCTYP SB" and ":CRTSTY SB1" or ":CRTSTY SB2". SB1 should be used via CRTSTY. System support for SB2 might work. Users: CBF, EAK Price: $3K, not actively marketed I hope. Screen: 25 x 80, white phosphor, black background Chars: 5 x 7 (in a 7 x 10 field), descenders Keybrd: Reasonable. Speeds: ...,9600 Has: Random, Addr, Tabs, CLEOL, CLEOS, Ins Line, Del Line, Standout. Lacks: At least on the SB1 Ins char & Del Char do not work. Overstrike. Misfeatures: Peruse the following from Mark Horton: Date: 7 Jun 1980 19:59:14-PDT From: CSVAX.mark at Berkeley Subject: superbee The terminal is unable to scroll at 9600 baud, and can't always keep up at 4800 baud either. 2400 baud is safe. This is not a problem on ITS because most people don't use scrolling, but it is serious on other systems. Even though it has insert line, the feature is unusable. This is because the terminal keeps a fancy data structure only using up the amount of memory needed by each line instead of a full 80 characters. Insert line inserts a zero length line, and when you try to put something on that line it either (1) ignores it if the line begins with a tab, or (2) takes characters away from the bottom of memory to put them in the line, causing decay of the screen if the bottom of memory is on the screen. Insert character might be useable with lots of special code but it is unlike any IC I have ever seen before. When the line grows past 80 characters, it inserts a blank space and starts shoving the next line to the right. When this second line hits column 80 it pulls the same trick on the third line, and it can eventually mess up the entire lower part of the screen. Delete line and delete character are quite ordinary and usable. Since there are only 2048 chars of memory, often the terminal will do the wrong thing and fail to display all 25 lines. Sometimes the top or bottom line will be blank until you position the cursor there. This doesn't seem to have a serious effect on a screen editor, but can be annoying. The standout modes (inverse and blinking) have the same misfeature as the Teleray and TVI: they leave a blank space between modes. The keyboard has an amazing number of keys crowded around it, including separate keys for newline (^_), return, linefeed, _, rubout, tab, backtab, and several special functions such as scroll up and down, but amazingly enough there is no backspace key! Shift lock is where control should be, control is to the left of the left shift key. Another botch is the SPOW "feature". This is supposed to make spaces be nondestructive on the screen when a latch is set. (Set by return, cleared by tab, newline, linefeed or home.) But what they don't tell you is that when the latch is set and you are entering a line of text, when you hit the space bar it actually TRANSMITS ESCAPE C (the nondestructive space sequence) instead of space! CRTSTY compensates for this by turning escape C back into space (I wonder if escape C does something useful in EMACS?) CBF 5 March 1978  File: TERMS, Node: SWTPC, Up: Top, Previous: SB, Next: TAB132 SWTPC CT-82 This terminal is very bizarre. It has an incredible number of useless features. The keyboard feels good, but it does not have N-key rollover. It has ins/del char/line. There are modes for whether ins-char pushes characters off the end of the line, or causes the line to wrap. The character can be inserted so things get pushed to the right or to the left. Similarly, inserting lines can push lines up or down. All the ins/del stuff, and all the other assorted features (such as slide the screen left or right, or scroll the quadrant that is north-east, NW, SE, or SW or the current cursor position) seems to happen very fast without glitches. The screen is small (9" diagonal??), but the characters are very sharp. The nominal layout is 16 lines by 82 characters. 20 lines by 82 characters-mode can be set, but then the descenders on one line overlap the tall letters on the next line (by one or two scan lines). There is a 22 by 92 format using uppercase only, but the sample we saw could not do it. All escape sequences begin with ^[, ^\, or ^], followed usually by one or more control characters. For some reason they did not catch on to the fact that one can follow an escape with printing characters as well as control characters, except for sending coordinate positions or something like that. What I mean is that all the escape sequences that don't take variable arguments consist entirely of control characters. The most commonly used (by their definition) functions are done by a single control character (so the first sentence of this paragrph is not entirely correct). However, it is possible to demand that these single control characters be preceded by some arbitrary characgter. Unfortunately, in the current version of the software, this means that CR and LF must also be escaped when this mode is requested! This is a known bug to them. It is possible to reprogram any sequence to be some arbitrary sequence. Thus it would be very easy to make this terminal emulate many others. I'm not usre what kind of length restriction there is on an reprogramed escape sequence. The escape sequence tables are loaded into RAM from ROM every time the terminal is powered up. There seem to be other software bugs as of this version of the terminal (release 1A). We were unable to get the terinal to work over the phone to a UNIX. It kept on going into some kind of control-char transparent mode. but we may have had it configured wrong in some way. The terminal has some kind of graphics. It will put a 2x3 pixel in a char position, using the char position to encode the pixel. But in version 1A software, the graphics stuff does not work. (This is another known bug to them, listed in the front of the manual). Price is $795/ qty. 1. The ByteShop owner where we saw the thing said SWTPC is shipping 40% of their production to Europe (I have no idea why). This terminal is too bizarre, too flaky, and too buggy , at least now, to warrant consideration. My impression is that to pad on most terminals, one sends nulls, which get thrown away by the terminal. However, the CT-82 manual says this won't work, because they use a UART, and so the nulls will get stuck in the UART FIFO buffer. So they say send nothing while padding (which is necessary sometimes, though I think not very often on the terminal). So, either they are wedged in thinking about UART ubffers, or they forgot to throw away nulls in the right way. Anyway, I am confused by that section of their manual. -DCH@MIT-ML (Considering the quality of Dan's report and the bizarre nature of the terminal, I have decided to dispense with the normal format; especially considering that this terminal does not seem to have enough going for it to bother comparing it around anyway. -CBF 11 March 1979)  File: TERMS, Node: TAB132, Previous: SWTPC, Up: Top, Next: Tek4020 TAB132 TAB model 132/15 Status: Upward compatible with vt100 (well, almost) and vt132 Users: ? Price: $1500 (this may be a university price) Screen: 24x80 or 24x132, memory is 96x132, green phosphor Chars: 9 x 14 (9 x 16 at 50 hz) Keybrd: Detachable. Cherry keyboard, good feel. However, the locations of some of the keys is just too weird. (Big) caps lock key to left of A, break is to left of caps lock (!), ctrl is above break. Large, tall tab key to left of Q and 1 makes escape (left of tab, above ctrl) hard to hit, similarly, large, tall return at right makes backspace and del hard to hit. ~`, \|, del, and linefeed are to the right of the return key. You have to type ctrl ~ to get ^^, ctrl ? to get ^_. All 8 home row keys are recessed. Numeric keypad. Typewriter layout. 8 "soft keys" at top driving menus in an incredibly bushy tree - very hard to find things, often takes 5 soft keystrokes to get something simple set if you know where to find it. Choice between n key rollover & 2 key lockout. Speeds: All up to 19200, send/recv separately set. Has: All but overstrike. Has several kinds of standout, done right, including bold, blink, reverse, underline, blank. Can do double width or only tops or only bottoms of chars, on a line by line basis. Misfeatures: Ctrl key is the worst. Has vt100 "eat newline" glitch, which is slightly different from concept eat newline glitch. Many of the modes cannot be set by escape sequence. This is a very entertaining terminal to give a demo of. Hitting reset makes it do a song and dance. The horizontal scroll feature does what you might think it would do - pretty gross. Putting it in 50 HZ mode on a 60 HZ line works, but gives you a headache. Has zillions of modes like ooze mode (aka smooth scroll), various auto-newline modes, emulates vt52/100/132, but the extra memory can cause things to get scrolled onto the screen when blank lines should be showing up. Memory is done differently than other terminals - things like cursor addressing, home, etc only affect the 24 lines on the screen. You can view the terminal as a 24 line terminal that remembers a lot of history. This turns out to be exactly what vi wants, but you ITS people may not like this. Has "typomatic" mode (which automatically makes typos for you?) that some terminals call typamatic. Return key won't repeat. The terminal itself is very large (20 wide x 13 high x 20 deep), as is the keyboard (19 wide x 3 high x 8 deep). The screen, at 10x7 in, is quite easy to read in 80 col mode, and decent in 132 col mode. [Actualy I find it THE most readable 132 column mode on the market, including the Datagraphix terminals -CBF] The arrow keys always transmit (the code depends on the emulation mode). The shift arrow keys (scroll up, scroll down, home) always work in local. The repeat rate on the keyboard is about 30 cps. It has a clicker that can be turned off (sounds nice). Does x-on/x-off handshaking, which can be turned off. We found a few minor bugs in the terminal. 11 July 1981 Mark Horton (cxvax.Mark@Berkeley)  File: TERMS, Node: Tek4020, Previous: TAB132, Up: Top, Next: T1061 Tektronix 4024, 4025 & 4027 Support: :CRTSTY TK4025 or TK4027 Users: Greenfeld@BBN-TENEXD, FJW@MC for 4025, MIT Architecture for 4027 Price: 4024 (alphanumeric only) $3000 4025 (alpha & graphics) $4800 8K bytes graphics memory, $5500 16K bytes graphics memory, $6200 32K bytes graphics memory (maximum) 4027 (alpha & color graphics) $8000 Screen: 34 x 80 green, rather long persistence phosphor Chars: fairly large Keybrd: Detachable. Unfortunate feel, keys have lots of sideways play. Some 4027's I have seen have a stiffer (and better) version of the keboard. All keys (except shift & control) redefinable to transmit arbirtrary sequences! Speeds: 0,50,75,110,134,150,300,600,1200,1800,2400,4800,9600 supports split speeds fully. Has: Random, Tabs, InsChar, DelChar, InsLine, DelLine, Standout Lacks: Addr, CLEOL, CLEOS, Overstrike Misfeatures: Lack of full-duplex operation (see below). Bewildering number of features in local operating system. Must have a command character to introduce statements, which character then becomes unavailable for normal use. Verbose command language from transmission time point of view, commands are at least 4 characters, folowed by operands separated by delimiters. This terminals have an amazingly frustrating feature: If the user types on the keyboard while the terminal is recieving a command, the command will be garbled. You can practically forget about using these terminals for real-time editing because of this. The lack of direct cursor addressing and CLEOL make this terminal almost impossible to support reasonably. The DEL command, which might be used to simulate it, is actually slower than wiping out a line by transmitting spaces at 1200 baud! (On the 4025, the 4027 may be faster). In addition the insert line command works in a very non-standard fashion; it moves down one line before performing the operation. This makes it impossible to insert a line before the top line in the screen. In general the command structure of these terminals is designed along the bankrupt concept that the local terminal can provide a much better editing environment than the central timesharing system. That the central timesharing system is inflexible and unfriendly, and furthermore communicates in a lockstepped line at a time protocol. In other words, if you're using an IBM system (TSO, CMS) it might be advantages, for real time editing, forget it. If one is willing to get some other terminal for editing however, these terminals do have some merit as graphics devices (which is what Tektronix should probably stick to). The 4025 has a graphics resolution of 640 horizontal by 476 vertically displayable. However, they use this scheme whereby the screen is divided up into 8 by 14 cells. If any 8x14 cell does not contain any graphics (or characters) then it does not need to consume any memory. Ie. they have a sparse allocation technique based on character cells. In fact, once it is sent to the terminal, there is no difference between putting a character on the screen, or sending graphics commands to fill the cell with the character. Initially however, the terminal does not act on any graphics commands, until one defines the exact area of the screen that graphics overlap on. Ie. you can say, the graphics workspace is lines 10 through 20 from columns 40 through 60. The terminal is full of features like these to give the local user control of such things. The underlying presumption seems to be that he is dealing with canned programs on a central computer, which either cannot be modified, or have insufficient paramterization to allow the users of graphics systems to specify these things. Unfortunately, this sparse allocation scheme allows one to do graphics with less than a full screen worth of bit memory. In fact, the price for a full screen worth of bit memory is rather ridiculous, and serves to make the terminal overpriced as a simple graphics display. The color 4027 has more going for it. It uses a very good color monitor. It is organized in similar fashions to the 4025 in general. I would recommend that anyone who is looking for a terminal to do color information presentations examine this as a cost effective device. However, be aware you should probably use it a display only in line with the previous warnings about commands getting garbled during user type-in. 31 December 1978, CBF 25 April 1979, CBF updated.  File: TERMS, Node: T1061, Previous: Tek4020, Up: Top, Next: TVI912 Teleray 1061 Status: :TCTYP T1061 (+%TPMTA if you have the Meta key in place of Home) Users: T1062-FANS@MIT-MC Price: $1095 quantity 1-5, $1050 5-9, $1000 10-25. Screen: 24 x 80, P4 (white) phosphor, grey background Chars: 7 x 9 in an 8 x 12 field. Keybrd: Typewriter paired, detached, N-key rollover. The tactile feel and layout could both be much better. There is now a Meta key available. Speeds: 50,75,110,134.5,150,300,600,1200,1800,2000,2400,3600,4800,7200,9600 Has: Cursor, Addr, Tabs, CLEOL, CLEOS, InsChar, DelChar, InsLine, DelLine, Standout? Lacks: Overstrike Misfeatures: The standout features (inverse video, blinking, half intensity, underlining) are useless to any system, since the start of the standout feature (called "visuals") takes up a character position and causes a blank to appear on the screen. There is no way around this. There is supposed to be a newer similar modle for somewhat more money that does not have this misfeature Standard firmware will fail to scan the keyboard during extended insert/delete line or scrolling operations. You can order the "SIPB" or "Arpa" PROMs for $100 which fixes this, but slows down these operations somewhat. Timing requirements with SIPB PROMS are apx. 60 milliseconds to clear whole screen, and 2.5 milliseconds per line moved for line insert or delete. This is one of the most cost-effective, low risk terminal that posses all the necessary features. The Meta key replaces the home key (beneath the bottom row of the keyboard to the right) and operates like to shift key to set the 8th (or parity) bit. Beware, if you're system ever randomly lets control characters back to the terminal. If the terminal recieves a ^R it will turn on the printer port. The problem with this is whether there is a printer there or not, if the printer is set to slower baud rate than the line, the terminal will place characters on the screen only as fast as it can send them to the printer port, thus totally messing up screen output. The SIPB PROMs read disable the ^R command. Teleray has said they might make the disablement settable from one of the DIP switches in the back, but they have not done this to my knowledge. There is supposed to be a new model which fixes the standout mode problem, also have provision for multiple pages of memory, alternate character sets and APL. I have not yet seen this model or any documentation. Note for ITS usage the auto newline switch (switch 5) should be in the OFF position. 31 December 1978, CBF 22 February 1979, CBF 11 March 1979, CBF 30 September 1979, CBF 3 February 1980, CBF  File: TERMS, Node: TVI912, Previous: T1061, Up: Top, Next: TVI950 Tele-Video Inc. 912 and 920 Status: Upward compatible with 3A Users: Mike Powell (csvax.Powell@Berkeley) Price: $840 qty 1 for 912, $940 for 920. Qty 10 discount about $50. Screen: 80*24 Chars: 12 x 10, underlining does not overlap charactes. Keybrd: Model 912B and 920B: bit paired layout, no rollover, numeric pad, too many keys. By all accounts one of the worst tactile feels on the market. 920 has an extra row of function keys that 912 doesn't have for an extra $100. There are now models 912C and 920C which have typewriter paired layout for additional $75. The feel is still no better though. Speeds: 75..19200. Insert line said to not be able to handle 9600. Has: Cursor, Addr, Tabs, CLEOL, CLEOS, InsChar, DelChar, InsLine, DelLine, Standout. Lacks: Overstrike, except it can underline. Misfeatures: Standout modes take up a blank space to store the mode change, making them almost useless. Cannot underline and inverse video the same char, turning one on turns the other off. Other combinations of the 4 seem to be possible. Backtab key to left of ctrl, I kept hitting backtab D instead of ctrl D, but it didn't transmit anything. Comments: Insert line requires 33 milliseconds of padding per line moved! Comes standard with a switch for 50/60 hz. No fan. I saw a prototype, rs232 cable in back was at strange angle making it hard if not impossible to plug in. Manufacturer address: TeleVideo, Inc. 3190 Coronado Drive Santa Clara, CA 95051 (408) 246-5428. Most of the above is from Mark Horton (csvax.Mark@Berkeley). The following is from Dan Halbert (csvax.Halbert@Berkeley) who saw one somewhat latter: We were quoted ~$745 in qty. 10. Has underline, reverse video, blinking, half-bright. Has ins/del line/char When an insert or delete line is done, the screen is blanked for a long time (maybe 1/4 - 1/2 second). This is apprently so you don7t see glitches. The mode changing leaves a blank space (e.g. there is a space before anything that is underlined). This is like the Teleray 1061 bug. The keyboard is BY FAR the worst I have ever tried to use. It has a sort of light touch, and the keys are crowded together slightly, it seems, so fast slightly sloppy touch typing makes for extraneous characters as your fingers slide over the keys. Apparently the problem is that each key makes contact at too short a travel, though the entire key travel is reasonable. It does not have n-key rollover, and there is no clicker. Other people who were not touch typists tried it, and found it bad, but usable. Surprisingly, the keyboard feels OK when the terminal is off, though the keys do wiggle. The inside looks OK, and fairly clean, as far as I can tell. There is a space for an alternate character-set ROM. If this terminal didn't have such an awful keyboard, it would be fine as a cheap featureful terminal. Things like screen blanking I can live with. 22 February 1979 11 March 1979  File: TERMS, Node: TVI950, Previous: TVI912, Up: Top, Next: VIS200 TVI-950 Televideo model 950 Status: Not upward compatible with tvi 912, but upward compatible with adm3a (at least at Berkeley). Termcap descriptions exist. Users: ihuxi!ihuxg!grg@ucb, ucsbcgl!eggert@ucb, cory.probst@ucb Price: About $1000. Screen: looks very much like vt100. white phosphor, good display Chars: big - don't have it in front of me now but about 10x14. Keybrd: Same awful feel as other TVIs. Layout is typewriter. Detachable, but cord is compatible with telephone handset cord, not stereo phone plug. Typeamatic on all keys. Lots and lots of keys around the edges. Numeric pad. Speeds: Up to 9600 baud. Has: All but overstrike. Standout mode has "magic cookie" braindamage, like other tvis and telerays and adds. Has all major forms of standout (half bright, inverse, blink, underline, blank). Misfeatures: Worst misfeature is the "edit group" of keys, which includes backspace, tab, arrow keys, home, print, send, backtab, clear screen, insert/delete line/char. As a group, these either work in local or transmit their codes (escape sequence settable). Alas, if you want a tab and backspace key that transmit ^I and ^H, you give up having your print, send, and arrow keys work in local. If you have more than one page of memory (1, 2, or 4 are possible) you need the arrow keys to move back to previous pages. It might be possible to program around this with software. The other misfeature is that when you change page size, the screen (and memory) are cleared. Since cursor addressing is memory relative, and vi uses scrolling, you have to set the page size to 24 lines for vi to work, and go back to whatever mode you like outside vi. It would be nice to have your context left there, like on the concept. There appears to be NO padding needed at all, even at 9600 baud. This is true even if you turn off the xon-xoff handshaking. The manual is so full of typos and errors it's pretty hard to figure anything out from it, as of 6/81. It has a "user line" (25th line) that you can feed any message you want into, so we can run the h19sys program on it. But there are misfeatures. It just stores up to 80 bytes in the line - escape, tab, etc will display as control characters instead of doing what they should. Also, if you have the user line displayed instead of the "status line" (showing the modes, cursor position, etc) when you go into setup mode (like the vt100) you go blind, not being able to see what you're doing. I (Mark Horton) don't own one, but I borrowed it for a week to bring up the termcap description in June of 1981. Given a choice between a tvi and an h19 at the same price, I would choose the h19. The h19 is, however, up to $300 cheaper.  File: TERMS, Node: VIS200, Previous: TVI950, Up: Top, Next: VT Visual Technology Inc. Visual 200 Status: Supported by :TCTYP VT52 +%TOLID and :CRTSTY VIS200 Users: csvax.ucsfcgl!cmevax!rusty@berkeley Price: Suggested list is $1195. Some distributors may offer it at $950. Screen: 24x80 Chars: 7x9 Keybrd: ASCII layout. Detachable. N-key rollover. Typeamatic. Feel is not very good. Speeds: 110, 200, 300, 1200, 2400, 4800, 9600, 19200; no split speeds Has: Random, Address, Tabs, CLEOL, CLEOS, InsChar, DelChar, InsLine, DelLine, Standout Lacks: Overwrite Misfeatures: Beeps when character typed in insert character mode. CAPS ONLY turns ~ into ^, etc. The cursor is a non-blinking reverse video block, which is poor. The beeping in insert character mode is very annoying. Has a 256 character input buffer. Sends ^S^Q when it fills. Seems to require no padding at 9600 bps. May not listen to keyboard when it gets busy. EAK 14 January 1980 The whole screen can be put in reverse video by a switch in the back. When in this mode using half-bright (all you get) for standout seems fairly reasonable. The terminal can emulate the usual set: adm3a, vt52, haz1500, and one I can't think of now. On the sample I saw, the parity switches in the back did not work properly - if the port happened to generate the wrong parity it refuses to accept it, even when the switch is set to ignore parity! (Our dz's worked OK but dh's didn't). MRH 23 July 1980  File: TERMS, Node: VT, Previous: VIS200, Up: Top, Next: VT50 DEC Video Terminal (models VT50, VT50H, VT52, VT55, VT61, VT100, VT132) See individual entries. "VT" is a generic name which causes CRTSTY to figure out which model it is.  File: TERMS, Node: VT50, Previous: VT, Up: Top, Next: VT50H DEC VT-50 Status: Supported by ":TCTYP VT50 PADTAB 1" and ":CRTSTY VT50". Users: ? Price: $1250 list, $1K or less in quantity or through dealer Screen: 12 x 80, P4 (white) phosphor, grey background Chars: 7 x 7, no descenders, no lower case Keybrd: Very good. Has audible clicker. Speeds: (75,150,300)(110,600,1200,2400,4800,9600) for split speeds pick transmitting speeds from left list, receiving speeds from the right list. Above based on VT-50H manual, VT50 MIGHT be different. Has: Random, Tabs, CLEOL, CLEOS Lacks: Addr, InsChar, DelChar, InsLine, DelLine, Standout, Overstrike Misfeatures: Small screen size and lack of lower case.  File: TERMS, Node: VT50H, Previous: VT50, Up: Top, Next: VT52 DEC VT-50H Status: Supported by ":TCTYP VT50 PADTAB 3" and ":CRTSTY VT50H". Users: ? Price: probably $1400 list, usually less in quantity or through dealer Screen: 12 x 80, P4 (white) phosphor, grey background Chars: 7 x 7, no descenders, no lower case Keybrd: Very good. Has audible clicker. Speeds: (75,150,300)(110,600,1200,2400,4800,9600) for split speeds pick transmitting speeds from left list, receiving speeds from the right list. Has: Random, Tabs, CLEOL, CLEOS Lacks: InsChar, DelChar, InsLine, DelLine, Standout, Overstrike Misfeatures: Small screen size and lack of lower case.  File: TERMS, Node: VT52, Up: Top, Previous: VT50H, Next: VT55 DEC VT-52 Status: Supported by ":TCTYP VT52" and ":CRTSTY VT52". Users: everyone on the 8th & 9th floors of 545 Tech Sq. Price: $2195 list, down to $1250 quantity 100, average $1750 through dealers $1500 for anyone in "packs" of 4. Screen: 24 x 80, P4 (white) phosphor, grey background Chars: 7 x 7, no descenders Keybrd: Very good. Has audible clicker. Speeds: (75,150,300)(110,600,1200,2400,4800,9600) for split speeds pick transmitting speeds from left list, receiving speeds from the right list. Has: Random, Addr, Tabs, CLEOL, CLEOS Lacks: InsChar, DelChar, InsLine, DelLine, Standout, Overstrike Misfeatures: Direct Cursor Adressing at 960cps loses The main local terminal for ML and MC, this is understood by ITS quite well. The only reason to use CRTSTY with this is to take advantage of the alternate keypad, such that the blank colored keys become META/TOP/CTRL prefixes and so forth. If there are any good suggestions for the numeric keys, they can be implemented.  File: TERMS, Node: VT55, Up: Top, Previous: VT52, Next: VT61 DEC VT-55 Status: Supported by ":TCTYP VT52" and ":CRTSTY VT52". Users: ? Price: $2995 Screen: 24 x 80, P4 (white) phosphor, grey background Chars: 7 x 7, no descenders, no lower case Keybrd: Very good. Has audible clicker. Speeds: ...,9600. supports certain split speeds. Has: Random, Addr, Tabs, CLEOL, CLEOS Lacks: InsChar, DelChar, InsLine, DelLine, Standout, Overstrike Misfeatures: No lower case. This is supposedly a "graphics" terminal. That means it allows one to draw 2 "curves" horizontally across the screen. I.e. for each horizontal position on the screen, the user has the capability of turning on up to 2 points randomly in the vertical plane.  File: TERMS, Node: VT61, Up: Top, Previous: VT55, Next: VT100 DEC VT-61 Status: Supported by ":TCTYP VT52". Users: ? Price: $3300 Screen: 24 x 80, P4 (white) phosphor, grey background Chars: 7 x 8, descenders Keybrd: Very good. Has audible clicker. Speeds: ? Has: Random, Addr, Tabs, CLEOL, CLEOS, InsChar, DelChar, InsLine, DelLine Lacks: Standout, Overstrike Misfeatures: Too many features. It has been reliably reported to send ^S^Q sequences due to falling behind when run at 4800 baud or maybe even less. ITS does not know its line insertion and deletion codes. CRTSTY could be made to know them very quickly if anyone wants it to. Line insertion and deletion come in flavors that ripple up or down, but ITS only knows how to use down. Other interesting features like a readable checksum register, A few controllable LED indicators on the side, word wrapping mode and a host of other semi-useless features.  File: TERMS, Node: VT100, Up: Top, Previous: VT61, Next: VT132 DEC VT-100 Status: Supported by CRTSTY VT100; also emulates VT52 Users: Innumerable, about100,000 per year are being manufactured by DEC. Price: ~$2100 list, MIT used to pay $1230. Dealers might typically sell it for $1600 if it weren't in such tight supply. $270 "advanced video option" provides enough memory for 24 lines of 132 chars, more screen enhancement modes and place to mount extra character sets. Screen: 24x80 or 14x132 software selectable. 24x132 optional. 12 inch diagonal, (8x4.5 inch active) P4 phospher. Composite video input and output. Chars: 7x9 with descendsers. 80 column mode very readable, 132 somewhat. Keybrd: Very similar to office Selectric layout. Detachable. Control and delete keys are distant because of this. Keytops same as VT52/LA36; user can enable and disable clicker, which is electronically generated rather than mechanically like VT52. Speeds: 50,75,110,134.5,150,200,300,600,1200,1800,2000,2400,3600,4800,9600, 19200 in any combination of transmit and recieve speeds. Note terminal probably cannot keep up at 19200 without using DEC's losing ^S^Q protocol, and cannot keep up at many speeds if one uses smooth scroll. Has: Random, Addr, Tabs, CLEOL, CLEOS, Standout; Does not have InsLine or DelLine but they can be simulated by scrolling the scrolling region. Lacks: InsChar, DelChar, OverStrike. Misfeatures: Many attributes can be set from the keyboard but not from the computer, most unfortunately the ^S/^Q switch can only be set from keyboard and its state cannot be read from the computer. Features: 132 columns (14 lines of 132 columns standard, 24 lines optional $270). Smooth scrolling by raster line (can slow down terminal at high speeds (ie. 4800 baud or greater) if system doesn't support ^S^Q protocal this can cause lost characters). Mode setting using display and keyboard; modes can optionally be saved in non-volatile memory. Mode to select pound sign substitute for number sign (U.K. ASCII). Additional character sets available with advanced video option. User can presumably create own character sets in PROM, but note should beware that the entire matrix is not availble, if you want to do graphics. 4 Software controllable LED's on keyboard. This is the VT-52 replacement. There is a VT52 compatibility mode. This is the only 132 column terminal for less than $4000. A VT-100 with an lsi-11 is called a PDT-110 and costs about $3900, but does not use normal LSI-11 backplane. There is now a VT-103 which is a VT-100 with a real Q-bus backplane for $2995. Note that price does not include the LSI-11, just the backplane and power supply. There was an outfit doing this prior to DEC's announcement of the VT-103 and charging $2,395 for it, contact: Can/Am Transduction Inc; 1461-A Franklin Ave; Tustin, CA 92680 (714) 838-4742. There will be a VT-132 coming out around March 1979 which will have line/char insert/delete. This terminal will be aimed at the "forms-filling" market (primarily intended for use with non-DEC mainframes, although the Trax system might use it). Priced about $300 more than the VT-100. Note, the VT-105 is a fraud. It has VT-55 type graphics. The VT-100 has sparked many many imitators. Datamedia, Micro-term, Plessly, DataGraphix are a few companies that should be looked at before you buy the DEC version. Due to the continued short supply of VT-100's the imitators may not be more than $100 or $200 off of DEC's price, however most seem to offer the features of the $270 advanced video option standard within their price. Some also offer a printer port option CBF 6 July 1978 Moon 10 August 1978 CBF 20 October 1978 CBF 31 December 1978  File: TERMS, Node: VT132, Up: Top, Previous: VT100, Next: VT4800 DEC VT-132 Status: :CRTSTY VT132 Users: gmp@MIT-Multics Price: MIT $1550, list probably around $2500 or whatever market wil bear Screen: 24x80 or 24x132 software selectable. 12 inch diagonal, (8x4.5 inch active) P4 phospher. Composite video input and output. Chars: 7x9 with descendsers. 80 column mode very readable, 132 somewhat. Keybrd: Detached, typewriter paired layout, user enabled clicker, 2 key rollover. Speeds: 50,75,110,134.5,150,200,300,600,1200,1800,2000,2400,3600,4800,9600, 19200 in any combination of transmit and recieve speeds. Note terminal cannot seem to put normal characters on the screen at 9600 baud without 50% padding. Has: Random, Addr, Tabs, CLEOL, CLEOS, InsChar, DelChar, InsLine, DelLine, Standout. Lacks: OverStrike. Misfeatures: Current models have a couple of firmware bugs. Insert char does not seem to take effect in the first 4 columns when in 132 column mode. The function codes for the commands which the VT132 implemented which the VT100 did not implement have all been implemented with their meanings reverse of the ANSI standard. This is to be fixed in future models. The insert mode LED which is documented to light in insert mode does not. The terminal is grossly slow. Equivalent functions seem to take an average twice as long on the VT132 as the VT100. This includes simple displayable graphic characters. The VT100 can generally keep up at 9600 baud, the VT132 cannot, 4800 baud seems the best it can do. The insert and delete line functions take longer to execute than the method of simulating them via scroll region manipulation that one must do for the VT100. Features: 132 columns. Smooth scrolling by raster line (can slow down terminal at high speeds (ie. 4800 baud or greater) if system doesn't support ^S^Q protocal this can cause lost characters). Mode setting using display and keyboard; modes can optionally be saved in non-volatile memory. Mode to select pound sign substitute for number sign (U.K. ASCII). Additional character sets available. User can presumably create own character sets in PROM, but note should beware that the entire matrix is not availble, if you want to do graphics. 4 Software controllable LED's on keyboard. 22 February 1979 15 April 1980 CBF  File: TERMS, Node: VT4800, Up: Top, Previous: VT132 Video Terminal Technology VT-4800 (not DEC!) This is a transcript of a summary that I made after seeing a VT-4800 first-hand. Maybe later I'll fix it up a little, but not much. -- KLH Well, had a chance to look over a VT-4800 today. Basically my feeling is that the lack of char/line ins/del is quite deadly to any hopes of using it, particularly with 48 lines. It seemed clear that these functions don't exist at all. The salesman said that one customer was doing that by adding an 8080 or something; it does have a 800 Khz parallel input port with whch I assume you can hook up your favorite controller. Serial I/O is limited to 9600 baud. It's not clear why 19200 isn't supported, since he claims no pads are required even at the 800 Khz rate. Anyway, the model he brought was fixed at 4800 because the baud selector switch hadn't been wired up yet, and even then there seemed to e some problem with its timing, since 4800 from the IO-11 didn't work very well; lots of garbage on the screen. Perhaps the clock got out of adjustment or smething (crystal, 0.5% error). So we were limited mostly to local-mode stuff. Other things: It has video invert and lowercase but both ma be options; not sure. (by invert I mean single chars - but no blinking, etc). It has no CLEOF. Cursor addressing is 4 bytes; all control functions are strappable to any control character (except ^G,^M,^J fixed). It doesn't use escape sequence. ie mos stuff directly in hardware. although it has M6800 processor, I guess it doesn't do much. If you woner about the 4-byte addressing, well it has set-X and set-Y functions which take next byte as value. Other functions set video normal/inverse, cursor left and right, up and down (scroll both ays), home, clear, cleol, total-reset. It has provision for "pages" with functions to jump to page 0, show next page, previous page. Needs 4K memory per page, up to 16K. The keybaord is pretty good. N-key rollover, good feel, auto-repeat, lots of blank keys for progammed use. It is possible to get a metizer set up for the 8th bit; it seemed as if using CTRL and SHIFT together might be a pssible way of doing it, although I'm not sure I and salesman understood the other perfectly. The display is bit mapped and graphics hardware is supposed to be available, 2-plane so no interference with characters. These by the way are 5x7 which is how they fit 48 lines in, as oppoed to 7x9 etc; the monitor is conventional. Graphics will ave hardware vectors, 256x256 resolution. There are altogether too many jumpered options which would be nicer to have as switches such as oe finds on the ADM-3A. eg parity handling. But I suppose once you put your soldering iron away there won't be much need to change. Costs: 1195/kit to individual. they may not sell kits in future though. in groups of 5, $995 fully assembled and $895 kit. Over 25, he didn't know; something lower. No educational institution discounts. 2 boards, one for main stuff and the other is part of the kbd. Not really impressive I would say; 48 lines and decent keyboard are about the only things it has going for it over other terminals. Now if it had line ins/del I would perk up.