Lines Matching +refs:csh +refs:line +refs:to +refs:string

6 # Report bugs and new terminal descriptions to
14 # stopped updates several years ago); we have decided to not change the header
17 # To further muddy the waters, it is noted that changes to this file as part of
18 # maintenance of ncurses (since 1996) are generally conceded to be copyright
25 # It would also be a nuisance to split the file into unknown/known authorship
27 # have been small corrections to Raymond's translations to/from termcap format,
31 # which reflect creative effort.  Furthermore, the structure of entries to
32 # reuse common chunks also is creative (and subject to copyright).  Finally,
48 # Please e-mail changes to terminfo@thyrsus.com; the old termcap@berkeley.edu
65 # Pointers to related resources (including the ncurses distribution) may
72 # This file assumes a US-ASCII character set. If you need to fix this, start
89 # in that it admits a group of capabilities (prefixed `OT') equivalent to
91 # to stock terminfo simply by filtering with `sed "/OT[^,]*,/s///"'; but if
98 # string-table limit of archaic termcap libraries except where explicitly
106 # as of June 1995; they are encouraging everyone to migrate to ncurses.
110 # to termcap only has to carry over leading comments.  Also, no name field
114 # script (reorder) that moved the most common terminal types to the front of
128 # of a terminfo/termcap entry (this feature had to be sacrificed in order
129 # to allow standard terminfo and termcap syntax to be generated cleanly from
134 # the string "########") and minor sections (beginning with "####"); do
138 # to see a listing of section headings.  The intent of the divisions is
139 # (a) to make it easier to find things, and (b) to order the database so
143 # usually correspond to manufacturers or standard terminal classes.
145 # product line names used by that manufacturers.
154 # The part to the left of the dash, if a dash is present, describes the
155 # particular hardware of the terminal.  The part to the right may be used
168 #		uses magic cookies to support multiple attributes.
171 #	-ns	No status line - suppress status line
173 #	-s	Enable status line.
176 # If a name has multiple suffixes and one is a line height, that one should
179 # Entries with embedded plus signs are designed to be included through use/tc
188 # In many cases, these indicated obvious fixes to syntax garbled by the
189 # composers.  In a few cases, I was able to deduce corrected forms for garbled
198 # The System V Release 4 and XPG4 terminfo format defines ten string
200 # certain of these capabilities to describe functions which are not covered
203 #	u9	terminal enquire string (equiv. to ANSI/ECMA-48 DA)
205 #	u7	cursor position request (equiv. to VT100/ANSI/ECMA-48 DSR 6)
206 #	u6	cursor position report (equiv. to ANSI/ECMA-48 CPR)
208 # The terminal enquire string <u9> should elicit an answerback response
212 # The cursor position request (<u7>) string should elicit a cursor position
216 # answerback string.  The string may contain the following scanf(3)-like
222 # The cursor position report (<u6>) string must contain two scanf(3)-style
223 # %d format elements.  The first of these must correspond to the Y coordinate
224 # and the second to the %d.  If the string contains the sequence %i, it is
225 # taken as an instruction to decrement each value after reading it (this is
226 # the inverse sense from the cup string).  The typical CPR value is
234 # All the entries in this file have been edited to assume that the tabset
240 # is an issue, you will have to hand-patch the file locations before compiling
251 # For the terminal types still alive, I'd like to have manufacturer's
256 # UNIX hackers.  Ideally, I'd like the headers for each manufacturer to
258 # terminal types as possible to be tagged with information like years
264 # please go to the terminfo resource page, grab the UFO file (ufo.ti), and
267 # If you have been around long enough to contribute, please read the file
273 # California copyright with dates from 1980 to 1993.
285 # This file deliberately has no copyright.  It belongs to no one and everyone.
299 # Special "terminals".  These are used to label tty lines when you don't
310 lpr|printer|line printer,
325 # DEL and ^C are hardcoded to act as kill characters.
326 # ^D acts as a line break (just like newline).
350 ansi+erase|ANSI clear screen/line,
358 ansi+idl1|ANSI insert/delete one line,
388 # ECMA-48 does not specify scroll-regions, but most people consider it to be
394 # characters sent to the printer do not echo on the screen. DEC terminals
395 # can also be put into autoprinter mode, where each line is sent to the
396 # printer as you move off that line, e.g., by a carriage return.
404 # We use \E[11m for rmacs rather than \E[12m so the <acsc> string can use the
407 # will work with any Intel console, they all seem to have inherited \E[11m
415 # Highlight controls corresponding to the ANSI.SYS standard.  Most
434 # Highlight controls corresponding to the ANSI.SYS standard.  *All*
506 # This section lists entries in a least-capable to most-capable order.
535 # not to require any -- even at 9600 bps.  If you encounter problems,
540 # character set to specify.  ANSI 3.64 seems to make allowances for several.
541 # Please make the appropriate adjustments to fit your needs -- that is
546 # I would appreciate the results on other terminals sent to me.
548 # Please report comments, changes, and problems to:
557 # (Added vt100 <rc>,<sc> to quiet a tic warning --esr)
572 # <vpa>/<hpa> capabilities, forcing curses to use repetitions of <cuu1>,
573 # <cuf1>, <cud1> and <cub1>.  Also deleted <ich> and <ich1>, as QModem up to
575 # to confuse many emulators.  On the other hand, we can count on these programs
577 # <invis=\E[9m>, but <invis=\E[8m> now seems to be more common under
580 pcansi-m|pcansi-mono|ibm-pc terminal programs claiming to be ansi (mono mode),
596 pcansi|ibm-pc terminal programs claiming to be ansi,
630 # ansi-generic is a vanilla ANSI terminal. This is assumed to implement
632 # insert/delete line/char is there, so it won't work with
635 # can't do some of those. Padding is assumed to be zero, which
708 # does "beep".  ESC ESC i is used for Ins to avoid tn3270 ESC i for coltab.
710 # Caution: vi is limited to 256 string bytes, longer crashes or weirds out vi.
712 # actually useful because it sends ^@ O, which beeps and opens a line above).
722 # Adds ins/del line/character, hence vi reverse scrolls/inserts/deletes nicer.
876 # UniTerm terminal program for the Atari ST:  49-line VT220 emulation mode
884 # under tcsh/zsh/bash/sh/ksh/ash/csh when you run MiNT in `console' mode
902 # BeOS entry for Terminal program Seems to be almost ANSI
936 # * response to user requests, assumes kcbt=\E[Z, the ANSI/ECMA reverse-tab *
942 #	string F26 ="\033[Z"
944 # * This has to use a key slot which is unfortunate (any unused one will    *
945 # * do, F26 is the higher-numbered one).  The change ought to be built      *
987 # and it matters, turn off <ccc>.  The %02x escape used to implement this is
988 # not supposedly back-portable to older SV curses (although it has worked fine
1010 # The 2.2.x kernels add a private mode that sets the cursor type; use that to
1017 # Linux 2.6.x has a fix for SI/SO to work with UTF-8 encoding added here:
1022 #	'~' scan line 1
1023 #	'p' scan line 3
1024 #	'r' scan line 7
1025 #	'_' scan line 9
1044 # Changes to the Linux console driver broke bce model as reported in
1056 # This assumes you have used setfont(8) to load one of the Linux koi8-r fonts.
1077 # reason: cp437 (aka "straight to font") is not functional under luit.
1085 # This is based on the Linux console (relies on the console to perform some
1089 #	\E*	move cursor to home, as as \E[H
1091 #	\EE	move cursor to beginning of row
1094 # Note: The status-line support is buggy (dsl does not work).
1110 # The acsc string may be incorrect.
1133 # console driver to make it ignore SGR 21 because the ECMA-48 standard
1221 # Regarding compatibility to vt100:  hurd doesn't specify <xenl>, as we don't
1233 # gsbom/grbom are used to enable/disable real bold (not intensity bright) mode.
1351 #	allow an application running on a color console to behave as if it
1361 # (esr: commented out <scp> and <rmcup> to avoid warnings.)
1511 # make this easy to change...
1520 # change the original to keypad mode.
1522 # (iris-ansi: added rmam/smam based on init string -- esr)
1535 # \EOP to \EOS.  The shifted and control modifiers still do the same thing.
1550 iris-ansi|iris-ansi-net|IRIS emulating 40 line ANSI terminal (almost VT100),
1596 # Added several capabilities to pccon+base, reading wsemul_vt100_subr.c -TD
1597 # Changed kbs to DEL and removed keys that duplicate stty settings -TD
1657 # pcvt termcap database entries (corresponding to release 3.31)
1667 # (esr: added <civis> and <cnorm> to resolve NetBSD Problem Report #4583)
1748 # Terminfo entries to enable the use of the ncurses library in colour on a
1751 # modified codes for setf/setb to setaf/setab, then to klone+color, corrected
1789 # Entry for the DNARD OpenFirmware console, close to ANSI but not quite.
1820 # from Kermit's emulation of vt220, does not correspond to actual vt220.  At
1823 wsvt25|NetBSD wscons in 25 line DEC VT220 mode,
1833 wsvt25m|NetBSD wscons in 25 line DEC VT220 mode with Meta,
1840 # many differences to continue in that path.  However, test-results may be
1841 # useful to people curious about compatibility with xterm.
1857 #	(however, selective erase refers to DECSCA, SPA)
1859 # Uses spaces to simulate double-size characters
1870 netbsd6|NetBSD wscons in 25 line DEC VT100 mode,
1917 # Now el1 clears not only to the line beginning, but also a large chunk
1918 # of previous line. But there is another bug - ech does not work at all.
1924 # Bug? The ech and el1 attributes appear to move the cursor in some cases; for
1925 # instance el1 does if the cursor is moved to the right margin first.  Removed
1936 cons25w|ansiw|ansi80x25-raw|freebsd console (25-line raw mode),
1965 cons25|ansis|ansi80x25|freebsd console (25-line ansi mode),
1970 cons25-debian|freebsd console with debian backspace (25-line ansi mode),
1972 cons25-m|ansis-mono|ansi80x25-mono|freebsd console (25-line mono ansi mode),
1977 cons30|ansi80x30|freebsd console (30-line ansi mode),
1979 cons30-m|ansi80x30-mono|freebsd console (30-line mono ansi mode),
1981 cons43|ansi80x43|freebsd console (43-line ansi mode),
1983 cons43-m|ansi80x43-mono|freebsd console (43-line mono ansi mode),
1985 cons50|ansil|ansi80x50|freebsd console (50-line ansi mode),
1987 cons50-m|ansil-mono|ansi80x50-mono|freebsd console (50-line mono ansi mode),
1989 cons60|ansi80x60|freebsd console (60-line ansi mode),
1991 cons60-m|ansi80x60-mono|freebsd console (60-line mono ansi mode),
2006 cons50r-m|cons50-koi8r-m|freebsd console w/koi8-r cyrillic (50-line mono),
2010 cons60r-m|cons60-koi8r-m|freebsd console w/koi8-r cyrillic (60-line mono),
2025 cons50l1-m|cons50-iso-m|freebsd console w/iso 8859-1 chars (50-line mono),
2029 cons60l1-m|cons60-iso-m|freebsd console w/iso 8859-1 chars (60-line mono),
2033 # which is intended to be xterm-compatible.  See for example
2039 # The /etc/ttys entries for console and other ttys are all configured to set
2043 # There is no VT100 line-drawing (uses +'s and -'s)
2057 # Considering cons25 as a base, the line-drawing mostly works, but is missing
2058 # the cells which happen to have ASCII control-character values:
2064 # Those are removed from this entry's acsc string to avoid confusion.
2065 # The resulting description provides correct line-drawing and function-keys -TD
2109 # are described here.  This entry really ought to be upgraded.
2111 # "stty rows NN", e.g. to use 24 lines.
2154 # Actually (TD pointed this out at the time the acsc string was added):
2165 # The line-drawing happens to work in several terminal emulators, but should
2166 # not be used as a guide to the capabilities of the vt52.  Note in particular
2167 # that vt52 does not support line-drawing characters (the scan-X values refer
2168 # to a crude plotting feature) -TD
2191 # line and trademark, to SunRiver Data Systems.  SunRiver has since changed
2192 # its name to Boundless Technologies; see http://www.boundless.com.
2202 # weird state (concept at beginning of next line, vt100 at end
2203 # of this line) so all versions of vi before 3.7 don't handle
2204 # <xenl> right on vt100. The correct way to handle <xenl> is when
2206 # and then assume you are in column 1 of the next line. If <xenl>
2211 # that you set auto-nl to "on", if you set it off use vt100-nam
2218 # tab settings are in non-volatile memory and don't need to be
2220 # the screen annoyingly. You can type "reset" to get them set.
2224 # is the reset state, and is assumed to be the normal state.  Application
2226 # "Esc [ {code}" sequences, conforming to ANSI standards.  In Application
2228 # was provided primarily as an aid to the porting of VT52 applications.  It is
2230 # applications such as vi will always transmit the <smkx> string.  Therefore,
2231 # the definitions for the cursor keys are made to match what the terminal
2232 # transmits after the <smkx> string is transmitted.  If the <smkx> string
2233 # is a null string or is not defined, then cursor keys are assumed to be in
2236 # always transmit the <rmkx> string to the terminal before they exit.
2238 # The VT100 series terminals have an auxiliary keypad, commonly referred to as
2241 # Application Mode.  Numeric Mode is the reset state, and is assumed to be
2245 # can be configured to send either LF (\015) or CR LF).  In Application Mode,
2250 # will set the TERM environment variable to point to a terminfo entry which has
2251 # defined the <smkx> string to include the codes that switch the keypad into
2253 # fields to match the Application Mode control codes.  If the <smkx> string
2254 # is a null string or is not defined, then the keypad is assumed to be in
2255 # Numeric Mode.  If the <smkx> string switches the keypad into Application
2256 # Mode, it is expected that the <rmkx> string will contain the control codes
2257 # necessary to reset the keypad to "Normal" mode, and it is also expected that
2258 # applications which transmit the <smkx> string will also always transmit the
2259 # <rmkx> string to the terminal before they exit.
2262 # The top line is the name of the key (some DEC keyboards have the keys
2264 # the most "official" name).  The second line is the escape sequence it
2266 # character).  The third line contains two items, first the mapping of
2286 # terminfo guidelines.  That is a compromise used to assign the remaining
2287 # keys on the keypad to kf5-kf0, used on older systems with legacy termcap
2298 # A better adaptation to modern keyboards such as the PC's, which have a dozen
2299 # function keys and the keypad 2,4,6,8 keys are labeled with arrows keys, is to
2300 # use the 5-key arrangement to model the arrow keys as suggested in the
2362 # (vt100: I added <rmam>/<smam> based on the init string, also <OTbs>. -- esr)
2414 # vt100 with one of the 24 lines used as a status line.
2415 # We put the status line on the top.
2424 # Status line at bottom.
2425 # Clearing the screen will clobber status line.
2433 # This entry (or vt102-nsgr) is probably the right thing to use for
2442 # Many brain-dead PC comm programs that pretend to be `vt100-compatible'
2443 # fail to interpret the ^O and ^N escapes properly.  Symptom: the <sgr0>
2444 # string in the canonical vt100 entry above leaves the screen littered
2460 # (vt131: I added <rmam>/<smam> based on the init string, also <OTbs> -- esr)
2478 # vt132 - like vt100 but slower and has ins/del line and such.
2489 # This vt220 description maps F5--F9 to the second block of function keys
2490 # at the top of the keyboard.  The "DO" key is used as F10 to avoid conflict
2520 # A much better description of the VT200/220; used to be vt220-8
2521 # changed rmacs/smacs from shift-in/shift-out to vt200-old's explicit G0/G1
2522 # designation to accommodate bug in pcvt -TD
2619 # You probably don't want to use this on a VMS machine since VMS will think
2623 # (Added vt100 <rc>,<sc> to quiet a tic warning -- esr)
2653 # <kel> is end on a PC kbd.  Actually 'select' on a VT.  Mapped to
2654 # 'Erase to End of Field'... since nothing seems to use 'end' anyways...
2661 # to SMASH the 1k-barrier...
2694 vt320-nam|vt300-nam|dec vt320 7 bit terminal with no am to make SAS happy,
2698 # We have to init 132-col mode, not 80-col mode.
2710 #   host writable status line, yet another different DRCS matrix size,
2719 # Emacs v18 terminal modes to deal with the cursor keys in that the arrow
2726 # (vt340: string capability "sb=\E[M" corrected to "sr";
2727 # also, added <rmam>/<smam> based on the init string -- esr)
2728 vt340|dec-vt340|vt330|dec-vt330|dec vt340 graphics terminal with 24 line page,
2768 #    macros, and other features too numerous to remember right now.  TERMCAP
2772 # Emacs v18 terminal modes to deal with the cursor keys in that the arrow
2779 # (vt400: string capability ":sb=\E[M:" corrected to ":sr=\E[M:";
2780 # also, added <rmam>/<smam> based on the init string -- esr)
2843 # takes two parameters, the key and the string.  Translating the key is
2854 # The string must be the hexadecimal equivalent, e.g., "5052494E" for "PRINT".
2855 # There's no provision in terminfo for emitting a string in this format, so the
2856 # application has to know it.
2904 # The VT520 is a monochrome text terminal capable of managing up to
2913 # assigned to an unknown key, Set-Up may be entered by pressing
2924 # In trying to get the function keys to work, I had to cobble my own
2927 # I seem to get them all -Mike Gran
2936 # to telnetd.   Michael Deutschmann <ldeutsch@mail.netshop.net> informs us
2945 # Zstem340 is an (IMHO) excellent VT emulator for PC's.  I recommend it to
2950 # (z340: changed garbled \E[5?l to \E[?5l, DEC smooth scroll off -- esr)
2952 z340|zstem vt340 terminal emulator 132col 42line,
2956 z340-nam|zstem vt340 terminal emulator 132col 42line (no automatic margins),
2981 # "terminal.app" in GNUstep, but I believe it to be an unrelated
2982 # codebase and I have not attempted to describe it here.
3015 # seems to be somewhere between "vt102", "ncsa" and "xterm-color".
3022 # titlebar as a phony status line, and may produce warnings during
3025 # which need to position the cursor or do other fancy stuff inside the
3026 # status line won't work with these entries. They're primarily useful
3028 # status line. Please note that non-ASCII characters don't work right
3029 # in the status line, since Terminal.app incorrectly interprets their
3046 # The early versions were tailored to the NeXT character set. Sometime
3047 # after the Apple acquisition the encoding was switched to MacRoman
3048 # (initially with serious altcharset bugs due to incomplete conversion
3049 # of the old NeXT code,) and then later to UTF-8. Also sometime during
3050 # or just prior to the early days of Mac OS X, the Terminal grew ANSI
3055 # version 10.4) this suffered from the <bce> bug, but that seems to
3059 # would routinely crash under load. Many of these bugs seem to have
3061 # some may still remain. This change seems to correspond to
3075 # * Renamed the AppKit Terminal.app entry from "Apple_Terminal" to
3076 #   "nsterm" to comply with the name length and case conventions and
3096 # extensions. This terminfo was written to describe versions 41
3116 # and simulating enough cursor-key presses to move the cursor to the
3129 # depends on font sizes and window sizes, but 50 characters seems to
3136 # graphics emulation, which seems to think the character encoding is
3217 nsterm+s|AppKit Terminal.app v41+ status-line (window titlebar) support,
3302 # defaults to xterm-color.  Alternative selections are ansi, dtterm, rxvt,
3317 # The versions of Terminal.app in Mac OS X version 10.3.x seem to have
3322 # and D P Schreber comparing to nsterm-c-s-acs.
3339 #   Added those to nsterm-16color, which is the version used for OS X 10.6
3341 #   There is no way to press keypad-comma unless application mode is enabled
3347 #   xterm-256color.  However, it adds "nsterm", so it is possible to use the
3348 #   nsterm entry from this file to override the MacPorts (20110404) or
3351 #   dialog defaults to xterm-256color.  Alternative selections are ansi,
3353 #   the menu says "Declare terminal as" without promising to actually emulate
3364 #   + selecting "xterm" or "xterm-16color" sets TERM to "xterm-256color".
3373 # Earlier reports state that these differences also apply to OS X 10.7 (Lion),
3391 # + f1-f4 map to pf1-pf4
3398 # + 11.6.6 (test insert/delete char/line with bce) has several failures.
3404 #   mode 1047 fails to restore cursor position (do not use)
3405 #   mode 1049 fails to restore screen contents (do not use)
3439 # This is an alias which should always point to the "current" version
3449 # capabilities to nsterm-16color that I have derived this description from that
3451 # user-configurable, but I attempt only to describe the default configuration
3454 # According to its documentation, iTerm uses terminfo to obtain function key
3526 #	no improvement to XFree86 1047/1048 modes
3532 #	no improvement to ncurses 'k'
3542 # 2018/05/19: xterm+sm+1006 seems to work with 3.1.6beta -TD
3583 # prompt for user name, instead requiring an icon to be selected from
3586 # panel of the Login section of System Prefs to make the special
3587 # ">console" login accessible. The first option is to enable 'Show
3589 # "Other..." icon to the graphical login panel. Selecting "Other..."
3591 # is to change the 'Display Login Window as:' setting to 'Name and
3612 # (-m) entries,] but removing the (msgr) capability seemed to help.
3626 # description. The console driver appears to be ASCII-only, so (enacs)
3631 # this, change the "console" entry in /etc/ttys from "vt100" to
3852 # it is intended to be VT102 plus selected features.  By default, it sets $TERM
3853 # to xterm, which is incorrect, since several features are misimplemented:
3865 # PuTTY recognizes xterm's 1049 mode for switching to/from alternate screen,
3910 vt100-putty|Reset PuTTY to pure vt100,
3932 #    are unrelated to rxvt's actual settings.
3935 #    here to help users who expect that the selections do what is implied.
3968 # Use modifiers to obtain function keys past 12:
3992 # 2013 to 38ksloc in 2019.  That is still smaller than PuTTY (160ksloc), but
3995 # Version 3.0 responds to DA as a VT400, however it does not implement the
4011 #	setal    set (under)line color
4012 #	ol       set default (under)line color
4068 # emulation (testing with vttest and ncurses) is reasonably close to vt100 (no
4086 # No fonts are supplied with the program, so the acsc string is chosen to
4091 # "resize -s"), though it does not pass SIGWINCH to the application if the
4115 # Version 4.59 has regular vt100 line-drawing (so it is no longer necessary
4116 # to choose a Windows OEM font).
4120 #	- underline seems to work with color (modify ncv).
4150 #	+ responds to DECRQM and DECRQSS controls, but not consistent with DSR
4169 # 25x80.  This entry uses the 'Terminal' font, to get line-drawing characters.
4191 # b) The program sets $TERM to "ansi", which of course is inaccurate.
4204 #	The modifiers are represented as the codes listed above, prefixed to
4240 # The settings dialog does not work (unless the end user expects to open
4244 # - the program sets TERM to cygwin if the tab is set to PowerShell,
4245 #   and to xterm-256color if "Legacy".  However, in the latter, more tests
4246 #   fail in vttest, which does not pay attention to TERM.
4250 # - does not flush response to primary DA, leaving a ^M on the end when
4282 # This sets TERM to xterm-256color, which is a little more successful than
4287 # - claims to be a VT100 with AVO, but copies xterm #276's secondary response
4288 # - menu 8 (insert/delete char/line) has problem with delete-character
4318 # You can add the following line to your .Xdefaults to change the terminal type
4319 # set by the xterms you start up to my-xterm:
4324 # by adding a similar line to /usr/X11/lib/X11/app-defaults/XTerm.  In either
4326 # to the default of xterm.
4330 # (xterm: ":MT:" changed to ":km:"; added <smam>/<rmam> based on init string;
4332 # as these seem not to work -- esr)
4349 # corrected typos in rs2 string - TD
4408 # initially part of the xterm sources (in XFree86).  But "xterm" continued to
4412 # This entry restores the original intent, intentionally not an alias to
4463 # xterm to use terminfo-based descriptions with the titeInhibit resource.
4541 # bit to the parameter.
4561 #	A parameter for a function-key to represent a modifier is just more
4566 #	A parameterized string should (really) not begin with SS3 (\EO).
4572 # that might apply to xterm+pcfkeys with different values of that resource.
4579 #   * modify  xterm-new  terminfo  entry to use capabilities for shifted
4706 # PF1 is NumLock (which happens to be reserved for other use).  To work around
4708 # directly related to VT100.
4789 # If configured to support 88- or 256-colors (which is fairly common in 2009),
4813 # require a different approach to rs1 -TD
4859 # is of interest only to the numerically illiterate), is compatible with other
4897 # subparameter separators, but chose not to correct this in konsole.  As of
4924 # + PuTTY 0.70 seems to recognize xterm's original implementation but does
4933 # asked for some of xterm's advanced features to be added to its terminfo
4945 # Ss is used to set the cursor style as described by the DECSCUSR
4946 #	function to a block or underline.
4947 # Se resets the cursor style to the terminal power-on default.
4956 # To use it, your decTerminalID resource must be set to 200 or above.
5032 # sunKeyboard resource to true:
5036 #	+ maps numeric keypad "+" to ",".
5070 # setting the vt100Graphics resource to false.
5071 xterm-utf8|xterm with no VT100 line-drawing in UTF-8 mode,
5074 # These building-blocks allow access to the X titlebar and icon name as a
5075 # status line.  There are a few problems in using them in entries:
5077 # a) tsl should have a parameter to denote the column on which to transfer to
5078 #    the status line.
5082 #    don't want to mess with icon-name when using those window managers.
5086 xterm+sl|access X title line and icon name,
5089 xterm+sl-twm|access X title line (pacify twm-descended window managers),
5098 #	\E[1$}	Status line
5100 # DECSSDT (select status line type)
5101 #	\E[0$~	No status line
5102 #	\E[1$~	Indicator status line
5103 #	\E[2$~	Host-writable status line
5106 # status line (either the indicator, or status line).  That is because if no
5107 # status line is used, then the terminal makes that line part of the user
5110 # Because there is no "esl" (enable status line) capability, the "tsl"
5111 # capability ensures that the status line is host-writable.  A DEC terminal
5112 # will clear the status line when changing from indicator to host-writable
5115 # Once on the status line, the row part of cursor addressing is ignored.  Since
5116 # tsl expects a parameter (to specify the column), the shortest addressing that
5117 # can be used for this purpose is HPA, e.g., \E[5d to go to column 5.
5119 dec+sl|DEC VTxx status line,
5141 # In this section, two extended capabilities are used to illustrate the mouse
5142 # protocol: XM and xm.  The "XM" capability is recognized by ncurses to allow
5145 # information to make the mouse support completely data-driven.
5151 # sequence", referring to the fact that it does not correspond to a "real"
5171 # shift and control to other features.  However, they are important because
5173 # byte are used to encode the button number for both presses and releases.
5175 # bits in the byte).  Later work on XFree86 xterm used the remaining 3 bits to
5186 # mentioned the "emacs" ("T") response.  Comments in button.c referred to the
5192 # source-directory to xc/doc/specs/xterm, polishing the formatting but adding
5201 # A more complicated example could be constructed to account for the "t"
5212 # those changes and began improvements, e.g., to support color.  This was, by
5218 # xterm patch #83 (1998/10/7), added Jason Bacon's changes to provide an
5231 # xterm patch #116 (1999/9/25) added Stephen P Wall's changes to support DEC
5234 # xterm patch #120 (1999/10/28) added my change to support wheel mouse, by
5236 # available bits in the first byte of the response to encode buttons 4 and 5.
5237 # xterm patch #126 (2000/2/8) amended that change to avoid conflicting with
5240 # xterm patch #262 (2010/8/30) added Ryan Johnson's changes to provide a mode
5260 # to it as "SGR 1006" since the replies resemble the SGR control string:
5271 # (proper setting of enacs, smacs, rmacs makes kterm to use DEC Graphics
5289 # These (xtermc and xtermm) are distributed with Solaris.  They refer to a
5328 # color stuff, I also have a status line defined as the window manager
5329 # title bar. [I have translated it to terminfo -- ESR]
5330 xterm-pcolor|xterm with color used for highlights and status line,
5338 # before ECMA-64 color support was folded into the main-line xterm release.
5378 # csw (2002-05-15): make xterm-color primary instead of nxterm, to
5388 # because we want it to be seen as <kcpy>.
5421 # shift-f1 to shift-f10 are f11 to f20
5423 # NumLock changes the application keypad to approximate vt100 keypad, except
5446 # workable vt100 LNM, DECALN, DECSNM modes, making it possible to display
5452 # kf1 and kf10 are not tested since they're assigned (hardcoded?) to menu
5480 # For any "recent" version of gnome-terminal, it is futile to attempt to
5482 # is hardcoded to set $TERM to "xterm", and on startup, it builds a subset
5483 # of the keys (which more/less correspond to the termcap values), and will
5484 # interpret those according to the $TERM value, but others not in the
5485 # terminfo according to some constantly changing set of hacker guidelines -TD
5493 # In vttest, it claims to be a vt220 with national replacement character-sets,
5494 # but aside from the identifier string, implements only a small fraction of
5515 # Version 3.6.1 sets TERM to xterm-256color (still hardcoded), which has
5521 # "xterm" terminfo to supply key information which is not built into the
5524 # which did not correspond to any that xterm produced - still not solving the
5528 #	kf13-kf24	shift		2	F1 to F12
5529 #	kf25-kf36	control		5	F1 to F12
5530 #	kf37-kf48	shift/control	6	F1 to F12
5531 #	kf49-kf60	alt		3	F1 to F12
5532 #	kf61-kf63	shift-alt	4	F1 to F3
5535 # no parameters.  This appears to be rote copying based on xterm+pcc0.
5555 # Originally VTE was promoted as a library able to emulate any terminal by
5557 # ability to read definitions of special keys (function-, editing-, cursor).
5559 # Before 2014, VTE had a termcap reader (originally pointing to a private copy
5566 # gnome #730137) dealt with attempts to recast that termcap reader as library
5567 # calls, then attempting to adapt a chunk of code from ncurses (src/vteti.c),
5568 # abandoning that and finally constructing a table to match xterm's default
5600 # Since most of the terminfo-related behavior is due to the VTE library,
5612 # which (based on the default branch setting) seems to be a fork of vte
5623 # trimmed to eliminate those.  Also, since it is a slightly older version of
5669 # This program hardcodes $TERM to 'xterm', which is not accurate.  However, to
5671 # xterm-r6.  The default keyboard appears to be 'linux'.
5676 #    because its implementors are unaccustomed to reading documentation - as
5684 #    by sending a bogus code (in the source it says it's supposed to be a
5692 #    video option.  Perhaps that's intended to be a "mildly-broken vt102".
5713 #    (also overline, which is too rarely used to provide as an extension)
5739 # that.  The otherwise identical keytabs have definitions to model the VT52
5743 # 2001, and shortly after retitled to "XFree 3.x.x".  Both it and "vt100" were
5747 # and likewise retitled to "XFree 4".
5750 # and changing backspace to ^H, removing that keytab's attempt to model the
5753 # The developers made changes to the default and linux keytabs.  Comparing
5758 # Most of the change for the default keytab was to make konsole act more like
5767 # first 4 function keys are used in xterm to support the VT100 PF1-PF4 keypad
5783 #    might expect konsole's developers to followup.  Twelve years later that
5784 #    has yet to happen.
5813 # KDE's "vt100" keyboard has no relationship to any terminal that DEC made, but
5861 # that the control-modifier itself is used to spawn a new copy of mlterm (the
5862 # "-P" option).  So control/F1 to control/F12 may not be usable, depending on
5865 #				kf1 to kf12	\E[11~   to \E[24~
5866 #	shift			kf1 to kf12	\E[11;2~ to \E[24;2~
5867 #	alt			kf1 to kf12	\E[11;3~ to \E[24;3~
5868 #	shift/alt		kf1 to kf12	\E[11;4~ to \E[24;4~
5869 #	control			kf1 to kf12	\E[11;5~ to \E[24;5~ (maybe)
5870 #	control/shift		kf1 to kf12	\E[11;6~ to \E[24;6~
5871 #	control/alt		kf1 to kf12	\E[11;7~ to \E[24;7~
5872 #	control/shift/alt	kf1 to kf12	\E[11;8~ to \E[24;8~
5905 # The insert/delete/home/end keys do not respond to modifiers because mlterm
5906 # looks in its termcap to decide which string to send.  If it used terminfo
5938 # rxvt is normally configured to look for "xterm" or "xterm-color" as $TERM.
5974 # For the keypad, use Shift to temporarily override Application-Keypad
5975 # setting use Num_Lock to toggle Application-Keypad setting if Num_Lock
6041 # "ESC [ 35 ~" to "ESC [  49 ~".  Keyboards with more than 12 function keys
6043 # xterm+pcfkeys to define keys past F12.
6090 # Rxvt has been uploaded to /pub/Linux/Incoming/rxvt.tar.z and
6098 # to make it a little more compact, and to add and remove certain features.
6102 # Send submissions for comp.os.linux.announce to: linux-announce@tc.cornell.edu
6106 # was the author of the changes to use the bce ("new color model") which was
6112 # work on vttest, as well as bug reports to Mark Olesen.  For instance, the fix
6144 # This variant is supposed to work with rxvt 2.7.7 when compiled with
6156 # +	made custom description (below) to work, though it sets TERM=xterm.
6160 #	window collapses to a single line when running several of the screens
6163 # +	The vt52 test works properly, but this is an exception.  Due to the
6284 # It corresponds to emu's internal emulation:
6286 # emu's default sets TERM to "xterm", but that doesn't work well -TD
6287 # fixes: remove bogus rmacs/smacs, change oc to op, add bce, am -TD
6315 # vt220 Terminfo entry for the Emu emulation, corresponds to
6317 # with NumLock set (to make the keypad transmit kf0-kf9).
6355 #	has status line
6356 #	supports ANSI colors (except for 'op' string)
6388 # "mterm -type ansi" sets $TERM to "ansi"
6406 # mterm normally sets $TERM to "mterm"
6414 # "mterm -type decansi" sets $TERM to "decansi"
6516 #	there is some problem turning off line-drawing
6564 # because they are assigned to modifier-4.
6566 # The program assigns TERM to match the program name (the upstream source says
6567 # "st", but Debian renames it to "stterm").
6572 # because st's notion of "meta" does not correspond to the terminfo definition.
6646 # Note:  the original terminfo description uses leading blank to persuade
6647 # ncurses to use "st" as its name.  Proper fix for that is to use "st" as an
6660 # - sgr referred to unimplemented "invis" mode.
6691 # terminal wants to use TERM=stterm-256color, but that is longer than 14
6704 #    up to 114Mb after testing (no scrollback).
6706 #    (two copies, just in case the host happens to be Mac OS X).
6712 #    a line.
6731 # a) add sgr string
6732 # b) corrected sgr0 to reset alternate character set
6733 # c) modified smacs/rmacs to use SCS rather than SI/SO
6754 #	"F" thick-line characters do not display
6756 #	terminal does not respond to 80/132-column switching
6761 #	does not respond to xterm mouse controls
6769 terminator|Terminator no line wrap,
6809 #	resizing the window causes it to stop listening to the keyboard
6871 #	Changed from xterm+pcf0 to xterm+pcf2
6898 # Project started in 2016/02, uses Rust and OpenGL, and in contrast to (most X
6899 # terminal programs) is not designed to run with a remote server.
6940 # subparameter delimiter).  Refactored here to use ncurses building blocks -TD
6978 #	does not respond to "resize -s"
6984 #	claims to be vt200:
6997 #		no status-line
7003 #		line-drawing with DECCARA does not work
7028 #	In contrast to function-keys, some additional modifier combinations
7086 #	browser such as Electron, using Web Sockets to talk to the server.
7088 # it can connect to, and display in, a web browser, or as a standalone Qt
7137 # UNIX 4.0 and successors as line discipline 1 (?), but is
7225 # 'screen' defines extensions to termcap.  Some are used in its terminal
7230 #      S0   (str)   Switch charset 'G0' to the specified charset.
7231 #      E0   (str)   Switch charset 'G0' back to standard charset.
7235 # According to its manual page
7241 #      X3.64) and ISO 2022 standards (e.g.  insert/delete line and support for
7245 # program uses a table (rendlist) which equates the SGR codes to terminal
7251 # correspond to the rendlist table.
7274 # Without some redesign of screen, it is not possible to extend the set of
7279 # use the same value.  For example, changing standout to be the same as any of
7285 # defaults to using "screen".  For background, screen supported ecma+index
7317 # The bce and status-line entries are from screen 3.9.13 (and require some
7318 # changes to .screenrc).
7322 screen-s|VT 100/ANSI X3.64 virtual terminal with hardstatus line,
7327 # Those variations permit to benefit from 16 colors palette, and from
7336 screen-16color-s|GNU Screen with 16 colors and status line,
7342 screen-16color-bce-s|GNU Screen with 16 colors using BCE and status line,
7351 screen-256color-s|GNU Screen with 256 colors and status line,
7357 screen-256color-bce-s|GNU Screen with 256 colors using BCE and status line,
7378 #       When  screen  tries  to  figure  out  a  terminal name for
7388 screen+fkeys|function-keys according to screen,
7393 # do not assume screen supports the feature.  Add this tweak to entries which
7402 # (b)	screen's implementation of bw is incorrect according to tack.
7403 # (c)	screen appears to hardcode the strings for khome/kend, making it
7404 #	necessary to override the "use=" clause's values (screen+fkeys).
7405 # (d)	screen sets $TERMCAP to a termcap-formatted copy of the 'screen' entry,
7407 # (e)	when screen finds one of these customized entries, it sets $TERM to
7414 #	translation.  They are suppressed here to show what is tested by tack.
7468 # screen since 3.9.13, and when used, will require fewer characters to be sent
7469 # to the terminal for updates.
7474 # Adding these lines to your ".screenrc" file will allow using these customized
7535 # was last updated in 2009 to include 256-color support.  The most recent
7546 # mentions a change to implement italics which should be in a version 5,
7565 # such as "screen.$TERM" to correspond to the outer terminal's quirks.  The
7566 # various entries such as screen.xterm-new provide a way to more closely
7585 # + This uses ncurses to manage the display, including support for italics and
7590 #   Oddly enough, if $TERM contains "linux", it attempts to set the title.
7597 #   and may simply pass-through rxvt's, making it appear to work.
7665 #         answerback string: "ncsa-vt220-8"
7679 # Though it supports ANSI color, NCSA Telnet uses color to represent blinking
7682 # The status-line manipulation is a mapping of the xterm-compatible control
7684 # pairs, since the latter ends the string that is loaded to the window-title.
7723 # The documented function-key mapping refers to the Apple Extended Keyboard
7724 # (e.g., NCSA Telnet's F1 corresponds to a VT220 F6).  We use the VT220-style
7749 # project - an heavily stripped down Linux to be run on 16 bit
7750 # boxes or, eventually, to be used in embedded systems - and have been
7752 # and the latest improvements I know of date back to March 2000.
7755 # as an added bonus, this deals with all the capabilities common to
7774 # As a matter of fact, ELKS 0.0.83 on PCs defaults to ANSI emulation
7775 # instead of VT52, but the "elks" entry still refers to the latter.
7780 # Project SIBO (for Psion 3 palmtops) console is identical to the ELKS
7819 sun-il|Sun Microsystems console with working insert-line,
7836 # flake out on the last line.  Unfortunately, without them the terminal has no
7837 # way to scroll.
7840 # If you are using an SS5, change the sun definition to use sun-ss5.
7844 sun+sl|Sun Workstation window status line,
7849 sun-s|Sun Microsystems Workstation window with status line,
7855 sun-48|Sun 48-line window,
7857 sun-34|Sun 34-line window,
7859 sun-24|Sun 24-line window,
7861 sun-17|Sun 17-line window,
7863 sun-12|Sun 12-line window,
7865 sun-1|Sun 1-line window for sysline,
7878 # Most of the current references to sun-color are from users wondering why this
7883 # According to wscons manpage, color is supported only on IA systems.
7885 # does not list these.  It also sets ncv#3, however that corresponds to
7920 wsiris|iris40|iris emulating a 40 line visual 50 (approximately),
7936 # environment).   Note: these have nothing to do with Sony's News workstation
7937 # line.
7960 # some control characters you type will do strange things to the screen.
8067 nwp512-a|nwp514-a|news-a|news42|news40-a|sony vt100 emulator 42 line,
8190 # no_color_video to inform the application that standout(1), underline(2)
8217 # nice colors for Emacs (white on blue, mode line white on cyan)
8223 # nice colors for Emacs (white on black, mode line black on cyan)
8267 #                to MSDOS box?
8272 #   add    hs    [has status line]
8273 #   add    fsl   [return from status line]
8274 #   add    tsl   [go to status line]
8280 #   add    u8    [user string 8] \E[?6c
8281 #   add    el    [clear to end of line] \E[K
8292 #   xenl  [newline ignored after 80 cols] messes up last line? Ehud Karni
8299 #   civis [make cursor invisible] causes everything to stackdump? \E[?25l\E[?1c
8340 # are from linux.  These have been tested not to cause problems.  xenl was in
8574 # variable is set to 'on'.  While the Windows NT POSIX console is seldom used,
8580 # VI POSIX Editor for Windows NT 3.1".  True to Microsoft form, not only
8585 # To use this, you need to a bunch of environment variables:
8600 # <http://www.nentug.org/unix-to-nt/ntposix.htm>.
8636 # Later OpenNT was renamed to Interix.
8737 # Altos made a moderately successful line of UNIX boxes.  In 1990 they were
8756 # shift keys.  I have renamed them to keys 32 and up accordingly.  Also,
8793 # shift keys.  I have renamed them to keys 32 and up accordingly.  I have
8794 # also made this entry relative to adm12 in order to give it an <sgr>. The
8822 #	Vox: 1-(916)-785-4363	(Technical response line for VDTs)
8856 # user to use those keys as arrow keys rather than as function
8880 # Note: no <home> on HPs since that homes to top of memory, not screen.
8881 # Due to severe 2621 braindamage, the only way to get the arrow keys to
8882 # transmit anything at all is to turn on the function key labels
8883 # with <smkx>, and even then the user has to hold down shift!
8884 # The default 2621 turns off the labels except when it has to to
8889 # Note: there are newer ROMs for 2621's that allow you to set
8893 # Since there is no way to set/unset the strap with an escape
8900 # but inside vi, the function key labels appear. You have to
8901 # hold down shift to get them to xmit.
8925 hp2621-48|48 line 2621,
8959 # The 2624 has an "error line" to which messages can be sent.
8960 # This is CLOSE to what is expected for a "status line". However,
8961 # after a message is sent to the "error line", the next carriage
8962 # return is EATEN and the "error line" is turned back off again!
8979 # any window clears memory, it is probably not possible to use
8988 # extra slow on the last line of the window.
8999 # This entry is for sysline. It allocates a 23 line window with
9000 # a 115 line workspace for regular use, and a 1 line window for
9001 # the status line.
9004 # Turn off horizontal line, Create ws #1 with 115 lines,
9005 # Create ws #2 with 1 line, Create window #1 lines 1-23,
9006 # Create window #2 lines 24-24, Attach cursor to workspace #1.
9017 # Force terminal back to 24 lines after being 23.
9076 # software to support it.
9188 # these are closer to the new 26xx series than the other 2621b
9239 # <is2> sets the screen to be 80 columns wide
9302 # last line, and underline capabilities.
9381 # added empty <acsc> to avoid warnings re <smacs>/<rmacs> --esr)
9490 # was being held to ground would trigger a send of the top line on the screen.
9491 # A quick fix might be to drop back to a cheesy 4-wire cable with pin 22
9492 # hanging in the air. (Thanks to Eric Fischer, <eric@fudge.uchicago.edu>,
9520 # Most of these terminals required an option ROM to support lower case display.
9542 # expressed in the using entry.  We'd like to cook up an <sgr> but the
9548 # Status line additions from Stephen J. Muir <stephen%comp.lancs.ac.uk@ucl-cs>
9550 # be ^Z, according to his entry.
9571 # You will need to get into the supervisor setup before you can set
9579 # Ctrl-T tabs 1-80   use left&right to move and up to set and
9580 # Ctrl-V tabs 81-158 down to clear tab. Shift-Ctrl-M sets right margin at cursor
9586 # Ctrl-S Save both setup and functions keys. Takes from 6 to 10 seconds.
9590 # ADM-12+ supports hardware handshaking, but it is DTR/CTS as opposed to
9594 # The following null-modem cable should fix this and enable you to use
9714 #       OFF enables printer BUSY active HIGH - If set to this, ADM31 senses
9719 # sw8   ON causes selected attribute character to be displayed
9720 #       OFF causes SPACE to be displayed instead - Factory Setting
9739 # sw7   ON causes NULLS to be displayed as NULLS
9740 #       OFF causes NULLS to be displayed as SPACES - Factory Setting
9823 # "system line" at the bottom of the screen blank (for those who
9825 adm42-ns|lsi adm-42 with no system line,
9834 # not just the cursor line!
9896 # built to replace them, and a qvt119+ which was a 101+ with available wide
9912 # If you inherit a Qume without docs, try Ctrl-Shift-Setup to enter its
9913 # setup mode.  Shift-s should be a configuration save to NVRAM.
9918 # This used to have <cvvis=\E.2> but no <cnorm> or <civis>.  The BSD termcap
9921 # What seems to be going on here is that this entry was designed so that
9942 # (qvt103: added <rmam>/<smam> based on init string -- esr)
9996 #	be selected in the status line (setup line 9).
10017 # All of these terminals (912 to 970 and the tvipt) are discontinued.  Newer
10180 # ASCII, and were generally similar to adm3a but with attributes
10186 # are so slow as to be nearly unusable.
10192 # magic-cookie terminals similar to the 950.
10212 # and how you'd like to use the terminal:
10237 # At 9600 baud, the terminal is prone to overflow its input buffer
10239 # character/line/screen/page), and it does not signal this over the
10244 # running on a Linux box connected to a TVI-920C with a later-model
10246 # vary. The numbers below seem to give the terminal enough time so
10252 # If you want to use the FUNCT key on a tvi912[bc], use the
10319 # 3 Closed, 4 Open:	DTR is connected to RTS
10348 # appropriate scanf string, since we need to subtract %' ' from the
10350 # to synchronize during padding tests, and seems to work for that
10366 # useful in block mode, including "clear X to nulls" (vs. "clear X to
10372 # a.k.a. half-intensity outside of protect mode) is used to control
10376 # A49B1/A49C1; the major difference seems to be that the latter ROMs
10378 # memory page, and for sending whole pages back to the host (mainly
10383 # as in <cup>); read cursor position (\E/), which is analogous to <u9>
10396 # At least up to the A49B1 and A49C1 ROMs, there are no \Eb and \Ed
10401 # The back_tab (<cbt>) sequence (\EI) doesn't work according to
10408 # Extension print mode (<mc5>) echoes all characters to the printer
10409 # port [in addition to displaying them] except for the page print mode
10411 # definition for <mc5> but I don't expect it to cause problems.  We
10412 # reset to page print mode in <rs1> since it may have been enabled
10434 # converts all affected characters to spaces.
10450 # to give a consistent magic-cookie count. Also note that <sgr> uses
10451 # backspacing (in the TVI-supported order) to apply all requested
10462 # This uses the second page memory option to save & restore screen
10470 # This simulates flashing by briefly toggling to the other page
10565 # also added empty <acsc> to suppress tic warning -- esr)
10580 # also added empty <acsc> to suppress tic warning -- esr)
10599 # (tvi924: This used to have <dsl=\Es0>, <fsl=\031>.  I put the new strings
10770 # to avoid "magic cookie" standout glitch:
10783 # turn off status line (\Eg)	clear status line (\Ef\r)
10785 # edit mode (\Er)		load blank char to space (\Ee\040)
10786 # line edit mode (\EO)		enable buffer control (^O)
10788 # program unshifted send key to send line all (\E016)
10789 # program shifted send key to send line unprotected (\E004)
10790 # set the following to nulls:
10792 #	line delimiter (\Ex1\200\200)
10795 # set end of text delimiter to carriage return/null (\Ex4\r\200)
10869 # I also inserted <ich1> and <kich1>; the :ko: string indicated that <ich>
10870 # should be present and all tvi native modes use the same string for this.
10897 #	set 48 line page (\E\\2)
10898 #	place cursor at page 0, line 24, column 1 (\E-07 )
10902 #	when entering ex, set 24 line page (\E\\1)
10903 #	when exiting ex, reset 48 line page (\E\\2)
10916 #	set 96 line page (\E\\3)
10917 #	place cursor at page 0, line 24, column 1 (\E-07 )
10920 #	when entering ex, set 24 line page (\E\\1)
10921 #	when exiting ex, reset 96 line page (\E\\3)
10963 # the :rs: string, inserted the <ich> implied by the termcap :ko: string.  Note
10964 # the :ko: string had :cl: in it, which means that one of the original
10965 # <clear=\E*>, <kclr=\EY> had to be wrong; set <kclr=\E*> because that's what
10966 # the 950 has.   Finally, corrected the <kel> string to match the 950 and what
10993 # According to BRL we could have <rmkx>=\E>, <smkx>=\E= but I'm not sure what
10994 # it does to the function keys.  I deduced <rmam>/<smam>.
10995 # also added empty <acsc> to suppress tic warning,  -- esr)
11019 # per line (rather than 40), Esc K chooses the normal character set.  Not sure
11023 # its absence means <smam>=\Ev isn't safe to use. -- esr)
11093 # It's intended to run when the Visual 50 is in vt52 emulation mode
11098 # character inserts. The whole line gets painfully redrawn for each
11139 # Character insertion is kludged in order to get around the "beep" misfeature.
11142 # and <dl1> strings, but we seem to get along fine without them.
11161 # to use vi200-f.
11170 # the function keys are programmable but we don't reprogram them to their
11174 # (vi300: added <rmam>/<smam> based on init string -- esr)
11195 # The best place to look for the escape sequences is page A1-1 of the
11199 # The :xp: line below is so that emacs can understand the padding requirements
11202 # also added empty <acsc> to suppress tic warning -- esr)
11218 # and with 33 lines. clear screen is modified here to
11251 # They now own the Qume and Amdek brands, too.  So these are the people to
11257 # I made some entries relative to adm+sgr.
11263 #	it requires magic cookies to do so.  Many applications do not
11265 #	the protect mode to support one attribute (dim) without cookies.
11294 # (wy30-mc: added <smcup> to suppress tic warning --esr)
11308 #	i.e. change $<100/> to $<100>
11316 #	will be changed to be the same as the last attribute given.
11319 #	to correctly handle multiple attributes on a screen.
11355 #	i.e. change $<100/> to $<100>
11356 # (wy50-mc: added <smcup> to suppress tic warning --esr)
11379 #	The color attributes are designed to overlap the reverse, dim and
11389 #	to display both color and blink.  In the final analysis I am not
11396 #	i.e. change $<100/> to $<100>
11455 #	This terminal does not need padding up to 9600 baud!
11515 #	to follow the following outline:
11531 #	It may be useful to assign two function keys with the
11540 # (wy60: we use \E{ rather than ^^ for home (both are documented) to avoid
11616 #	does not have the 42/43 line mode.  In the Wyse-60 the "lines"
11660 #   is too much complex to be described);
11666 # vi deletes one character at the beginning of a line with tabs in it.
11762 #	to follow the following outline:
11774 #	graphics to text.  If "Display Memory = Unshared" then the
11777 # (wy160: we use \E{ rather than ^^ for home (both are documented) to avoid
11852 #	to be the same as the last attribute given.
11855 #	to correctly handle multiple attributes on a screen.
11918 #		24 line screen with status line.
11921 #	the escape key.  I strongly recommend that <f11> be set to
11923 #	The terminal may have to be set for 8 data bits and 2 stop
11924 #	bits for the arrow keys to work.
11979 # "What I did was change leave the terminal cursor keys set to Normal
11980 # (instead of application), and change \E[ to \233 for all the keys in
11983 # me), but I can't find that and the server under my bookmark to "Wyse
12027 #	as 24 data lines and a terminal status line (top or bottom) or
12028 #	25 data lines.  The 48 and 50 line modes change the page size
12196 #	Wyse 370 -- 24 line screen with status line.
12198 #	The terminal may have to be set for 8 data bits and 2 stop
12199 #	bits for the arrow keys to work.
12361 #               is3 -> If this string is empty then rs3 gets sent.
12364 #       - The BS key is programmed to generate BS in smcup since
12365 #         is2 doesn't seem to work.
12368 #       - Find   : delete to end of file
12369 #       - Select : clear a line
12372 #       - Bottom status line (host writable line) is used.
12374 #         keypad in Dec application mode which doesn't seem to work
12435 #       The DEL key is programmed to generate BS in is2.
12437 #       - 'End' key will clear till end of line on EPC keyboard
12440 #       - Delete : delete a character (have to change interrupt character
12441 #                  to CTRL-C: stty intr '^c') for it to work since the
12578 # Bugs: <ed>, <el>: do not work except at beginning of line!  <clear> does
12580 # line).
12588 # Cannot use line 25, now acts funny like ansi special scrolling region.
12589 # Initialization must escape from that region by cursor position to line 24.
12602 # Cannot use line 25, now acts funny like ansi special scrolling region.
12603 # Initialization must escape from that region by cursor position to line 24.
12642 # Please send changes with explanations to bug-gnu-emacs@prep.ai.mit.edu.
12643 # (vt320-k3: I added <rmam>/<smam> based on the init string -- esr)
12704 # These entries attempt to describe Avatar, a terminal emulation used with
12705 # MS-DOS bulletin-board systems.  It was designed to give ANSI-like
12715 # the facilities to test them.  Let me know if they work, or don't.
12718 # and more motivated than I am and can figure out how to wrap terminfo
12719 # around some of them, and because they are weird enough to be funny):
12721 # ^L		-- clear window/reset current attribute to default
12732 # ^V^J%p1%c%p2%c%p3%c%p4%c%p5%c	-- scroll (p2,p3) to (p4,p5) up by p1 lines
12733 # ^V^K%p1%c%p2%c%p3%c%p4%c%p5%c	-- scroll (p2,p3) to (p4,p5) down by p1 lines
12749 # ^V^T			-- change highlight at current cursor position to %c
12755 # (The <blink>/<bold>/<rev>/<smacs>/<smul>/<smso> capabilities exist only to
12761 # available documentation gives no clues for a workable string.
12786 # '90s), especially in the BBS world, and still has some loyal users due to
12787 # its very small memory footprint and to a cute macro language.
12827 # This line:
12834 # These entries add some 'sanity stuff' to the clear function.  That is, it
12853 # This section is devoted to older commercial terminal brands that are now
12854 # discontinued, but known to be still in use or represented by emulations.
12861 # The AT&T/Teletype terminals group was sold to SunRiver Data Systems (now
12889 # Seems upward compatible with vt100, plus ins/del line/char.
12891 # No check is made to make sure that only 3 parameters are output.
12944 # (v5410: added <rmam>/<smam> based on init string -- esr)
12984 # 5) set line wraparound
12987 # 8) program ENTER to transmit ^J,
12988 # We use \212 to program the ^J because a bare ^J will get translated by
12991 # <is3> set screen color to black,
12998 # smkx programs the SYS PF keys to send a set sequence.
13000 # This string causes them to send the strings <kf1>-<kf8>
13002 # (att4415: I added <rmam>/<smam> based on the init string -- esr)
13046 # However, when you program user pf labels you have to reselect
13047 # user pf keys to make them appear!
13189 # The following two lines are the comment originally attached to the entry:
13190 # This entry appears to avoid the top line - I have no idea why.
13206 # to change back to what's being used for the 5410. Many of the
13211 # (att5425: added <rmam>/<smam> based on the init string -- esr)
13281 # 9 corresponds to the leftmost touch target on the screen,
13282 # function key 16 corresponds to the rightmost.
13314 # Function key 9 corresponds to the leftmost touch target on the screen,
13315 # function key 16 corresponds to the rightmost.
13317 # There are problems with soft key labeling.  These are due to
13318 # strangenesses in the native terminal that are impossible to
13403 # printer must be set to EMUL ANSI to accept ESC codes
13445 # No delays specified; use "stty ixon -ixany" to enable DC3/DC1 flow control!
13463 # delays are specified; use "stty ixon -ixany" to enable DC3/DC1 flow control!
13464 # assumptions: <ind> (scroll forward one line) is only done at screen bottom
13497 # Entries for <kf15> thru <kf28> refer to the shifted system pf keys.
13499 # Entries for <kf29> thru <kf46> refer to the alternate keypad mode
13546 # (att610: I added <rmam>/<smam> based on the init string.  I also
13624 # (att620: I added <rmam>/<smam> based on the init string and
13696 # requirements.  Some capabilities assume a printer attached to the Aux EIA
13698 # delays are specified; use "stty ixon -ixany" to enable DC3/DC1 flow control!
13726 # terminal.  Comments are relative to changes from the 605V2 entry and
13730 # Here is what's going onm in the init string:
13732 # x	ESC [ 56;ps|	set lines to 24: ps=0; 40: ps=1 (plus status line)
13733 #	ESC [ 53;0|	set GenFlow to Xon/Xoff
13753 # was assumed the 730 entry to be more correct so rmso has changed.  The
13754 # 605V2 has no sequences to turn individual attributes off, thus its setting
13758 # to specifically turn off underscore, rather than return to all normal
13762 # capability as pfxl.  It was changed here to pfx since pfxl
13764 # allows strings to be parameters and label values must be programmed as
13766 # and strings to be parameters.  The 605V2 pfx entry should be examined later
13847 # att730 has status line of 80 chars
13856 # (att730: I added <rmam>/<smam> based on the init string -- esr)
13912 # not appear on the screen but are shown to reference the bezel buttons.
13914 # position relative to the screen.
13954 #	The character string sent by key 'kf26' may be user programmable
13955 #       to send either \E[16s, or \E[26s.
13956 #       The character string sent by key 'krfr' may be user programmable
13957 #       to send either \E[17s, or \E[27s.
13963 # "kf" functions adds carriage return to output string if terminal is in
13964 # 'new line' mode.
13983 #                       Pn= < Enter new line mode
13990 #                       Pn= < Exit new line mode
14004 #       Transparent Button String (TBS): \E[Pn1;Pn2;Pn3;{string
14005 #                       Pn1= Button number to be loaded
14006 #                       Pn2= Character count of "string"
14011 #                       String= Text string (15 chars max)
14038 #                       string= Phone number to be dialed
14041 #                       string= Label for phone buttons
14051 #       Programming The Function Buttons: \EPfPn;string\
14054 #                       string= Text to sent on button depression
14060 #       Local Directory Data to host: \EPp11;LOCAL...DIRECTORY...DATA\
14158 # (ampex219: I added <rmam>/<smam> based on the init string, added <cvvis>
14160 # and moved the padding to be *after* the caps -- esr)
14202 # allowing up to 76-character screen heights!  They were reachable at:
14219 # status line moved to top of screen, <flash> removed 5/82
14220 # Some unknown person at SCO then hacked the init strings to make them more
14231 #	    and the value used to test these termcaps)
14232 #	Note that many of these settings are irrelevant to the terminfo
14233 #	and are just set to the default mode of the terminal as shipped
14243 #	return and line feed/return for <cr> key *
14264 #	keyboard local/on line*
14273 #	transmit/do not transmit line separators to host*
14274 #	transmit/do not transmit page tab stops tabs to host*
14275 #	transmit/do not transmit column tab stop tabs to host*
14292 #	number of lines to print data on (printer) (56*)
14298 #	number of pad chars on new line to printer (0*)
14306 #	new line is: 01=LF,10=CR,11=CRLF*
14311 #	LF is newline/LF is down one line, same column*
14312 #	wrap to preceding line if move left from col 1*/don't wrap
14313 #	wrap to next line if move right from col 80*/don't wrap
14321 #	editing extent: 0=display,1=line*,2=field,3=area
14345 # to these capabilities.  This is the nonstandard GNU termcap scrolling
14510 guru+s|guru status line,
14541 guru-76-lp|guru-lp|guru with page bigger than line printer,
14570 # terminals was merged again.  Then AT&T sold the terminal business to
14571 # SunRiver, which later changed its  name to Boundless Technologies.  The
14582 # Their voice mail used to describe the place as "SunRiver (formerly ADDS)".
14587 # (regent: renamed ":bc:" to ":le:" -- esr)
14643 # Theory; the vp3a+ wants \E0%c to set highlights, where normal=01000000,
14666 # Erase to end of line            ESC,T
14667 # Erase to end of page            ESC,Y
14713 # Note:  emacs sends ei occasionally to insure the terminal is out of
14715 #        mode.  A hack to get around this is <ich1=\EF\s\EF^U>.  (Also,
14717 #   -    <xhp> indicates glitch that attributes stick to location
14718 #   -    <msgr> means it's safe to move in standout mode
14720 #               the status line
14753 # CIT 80  - vt-52 emulator, the termcap has been modified to remove
14764 # (cit101: added <rmam>/<smam> based on init string, merged this with c101 -- esr)
14779 # last two lines (with the capabilities in caps) are used by RM-cobol to allow
14800 # There is no need to use the initialization string is=... (by invoking
14802 # up (and the setup saved with ^S) to be compatible with this termcap.  To be
14804 # works is to set all the manually settable stuff to factory defaults
14812 # (cit101e-rv: added empty <rmcup> to suppress a tic warning. --esr)
14853 # Hardware tabs are assumed to be set every 8 columns; they can be set up
14855 # "stty ixon -ixany" to enable DC3/DC1 flow control!
14856 # (cit500: I added <rmam>/<smam> based on the init string -- esr)
14943 # Other switches may be set according to communication requirements.
14961 # Part of the long initialization string defines the "DOWN" key to the left
14962 # of the tab key to send an ESC.  The real ESC key is positioned way out
14965 # The termcap won't work in 132 column mode due to the way it it moves the
14966 # cursor.  Termcap doesn't have the capability (as far as I could tell) to
14969 # (cdc721: changed :ri: to :sr: -- esr)
14991 # `Geveke' and made async terminals; but (according to the company itself!)
14995 # to have been at least two models, the 33 and the 50.
14998 # The 50 seems to be a top end vt220 clone, with the addition of a higher
15028 	lf3=A insert line, lf4=A delete line, lf5=A clear,
15029 	lf6=A ce of/cf gn, lf7=A print, lf8=A on-line,
15047 # particular, the whole `Concept' line described here was discontinued long
15052 # Extensive changes to c108 by arpavax:eric Feb 1982
15053 # Some unknown person at SCO then translated it to terminfo.
15055 # There seem to be a number of different versions of the C108 PROMS
15063 # You should configure the C108 to send ^S/^Q before running this.
15066 # new status line display entries for c108-8p:
15069 # line of memory, set bkgnd stat mesg there, select window 0.
15071 # <tsl> - to status line - select window 2, home cursor, erase to
15072 # end-of-window, 1/2 bright on, goto(line#0, col#?)
15074 # <fsl> - from status line - 1/2 bright off, select window 0
15113 # relative. To get it to work right here, smcup/rmcup (which
15118 # we set the window size to zero ("\Ev    " in rmcup) which the
15119 # terminal recognizes as an error and resets the window to all
15138 # it was a bad idea to clobber their definitions.
15140 # The <mc5> sequence changes the escape character to ^^ so that
15141 # escapes will be passed through to the printer. Only trouble
15142 # is that ^^ won't be - ^^ was chosen to be unlikely.
15143 # Unfortunately, if you're sending raster bits through to be
15146 # \EQ"\EY(^W (send anything from printer to host, for xon/xoff)
15184 #	is2=.  Also, \E=124l in is2= could have been used to prevent needing
15185 #	to specify xenl:, but that would have rendered the last space on the
15186 #	last line useless.
15191 # dsl:	Go to window 2, go to the beginning of the line, use a line feed to
15192 #	scroll the window, and go back to window 1.
15193 # is2:	the string may cause a warning to be issued by tic that it
15194 #	found a very long line and that it suspects that a comma is missing
15196 #	once).  The initialization string contains the following commands:
15199 #		\E)0			set alternate character set to
15201 #		^O			set character set to default
15207 #		\E[90;3u		set Fkey definitions to "transmit"
15209 #		\E[92;3u		set cursor key definitions to
15211 #		\E[43;1u		set shift F13 to transmit...
15213 #		\E[44;1u		set shift F14 to transmit...
15215 #		\E[45;1u		set shift F15 to transmit...
15217 #		\E[46;1u		set shift F16 to transmit...
15219 #		\E[200;1u		set shift up to transmit...
15221 #		\E[201;1u		set shift down to transmit...
15223 #		\E[202;1u		set shift right to transmit...
15225 #		\E[203;1u		set shift left to transmit...
15227 #		\E[204;1u		set shift home to transmit...
15229 #		\E[212;1u		set backtab to transmit...
15231 #		\E[213;1u		set shift backspace to transmit...
15233 #		\E[214;1u		set shift del to transmit...
15236 #		\E[2!w			move to window 2
15237 #		\E[25;25w		define window as line 25 of memory
15238 #		\E[!w			move to window 1
15239 #		\E[2*w			show current line of window 2 as
15240 #						status line
15241 #		\E[2+x			set meta key to use high bit
15242 #		\E[;3+}			move underline to bottom of character
15244 #	All Fkeys are set to their default transmit definitions with \E[90;3u
15245 #	in is2=.  IMPORTANT:  to use this terminal definition, the "quit" stty
15249 #	necessary to change ^| to ^] in all of these definitions, and add
15250 #	\E[2;029!t to is2.
15251 # lines: is set to 24 because this terminal refuses to treat the 25th
15252 #	line normally.
15257 # rmso: Could use \E[1;7!{ to turn off only bold and reverse (leaving any
15258 #	other attributes alone), but some programs expect this to turn off
15260 # rmul: Could use \E[4!{ to turn off only underline (leaving any other
15261 #	attributes alone), but some programs expect this to turn off
15263 # sgr:	Attributes are set on this terminal with the string \E[ followed by
15274 #		effects to protected characters which make them inadvisable).
15275 #	 The mapping of terminfo parameters to attributes is as follows:
15285 #	 The code to do this is:
15310 # smul: The "underline" attribute is reconfigurable to an overline or
15311 #	strike-through, or (as done with \E[;3+} in is2=), to a line at the true
15312 #	bottom of the character cell.  This was done to allow for more readable
15313 #	underlined characters, and to be able to distinguish between an
15315 # xenl: Terminal can be configured to not need this, but this "glitch"
15320 # string modifies the 'O' in kf43 to kf46 to a '$'.
15328 # tsl:	Go to window 2, then do an hpa=.
15332 # There are two ways to flash the screen, both of which have their drawbacks.
15333 # The first is to set the bell mode to video, transmit a bell character, and
15334 # set the bell mode back - but to what?  There is no way of knowing what the
15335 # user's old bell setting was before we messed with it.  Worse, the command to
15336 # set the bell mode also sets the key click volume, and there is no way to say
15337 # "leave that alone", or to know what it's set to, either.
15338 # The second way to do a flash is to set the screen to inverse video, pad for a
15339 # tenth of a second, and set it back - but like before, there's no way to know
15342 # programs) assume that by defining flash=, you want the computer to use it
15349 # Windowing is possible, but not defined here because it is also used to
15350 # emulate status line functions.  Allowing a program to set a window could
15351 # clobber the status line or render it unusable.  There is additional memory,
15358 # available at any time, so taking the time to override the default is
15362 # Not defined, because it appears to have some strange side effects.
15366 #	 Available, but making them available to programs is inadvisable.
15367 #	 The code to do this is:
15376 #	 [next line applies to pfx only]
15380 #		%p2%s			OUTPUT	(string) as string
15392 # These sequences apply to the cursor and setup keys only, not to the
15394 # available to programs is inadvisable.
15397 # meaning to any other terminal.
15400 # Available, but making it available to programs is inadvisable.
15402 # Available, but making it available to programs is inadvisable.
15405 # Available, but making it available to programs is inadvisable.
15409 # Available, but making them available to programs is inadvisable.
15414 #	 "transparent print" (which doesn't copy data to the screen, and
15415 #	 therefore needs mc5i: specified to say so) and "auxiliary print"
15462 # <ht> through <el> included to specify padding needed in raw mode.
15463 # (avt-ns: added empty <acsc> to suppress a tic warning --esr)
15464 avt-ns|concept avt no status line,
15495 avt-rv-ns|concept avt in reverse video mode/no status line,
15498 avt-w-ns|concept avt in 132 column mode/no status line,
15501 avt-w-rv-ns|concept avt in 132 column mode/no status line/reverse video,
15505 # Concept AVT with status line. We get the status line using the
15506 # "Background status line" feature of the terminal. We swipe the
15507 # first line of memory in window 2 for the status line, keeping
15509 # The first line is used instead of the last so that this works
15513 avt+s|concept avt status line changes,
15554 # According to James Carlson <carlson@xylogics.com> writing in January 1995,
15558 # DG terminals have function keys that respond to the SHIFT and CTRL keys,
15670 # checked by "op", "seta[bf]", and "set[bf]" to refresh the attribute settings.
15747 # Initialization string 1 sets:
15758 # According to the 4.4BSD termcap file, the dg200 <cup> should be the
15797 # having a command to move straight down from any position on the bottom line
15823 # Ought to work for a Model 6242, Type D210 as well as a 605x.
15852 #	Reverse video, no insert/delete character/line, 7 bits/character only.
15854 # Initialization string 1 sets:
15872 # Like D200, but adds clear to end-of-screen and needs XON/XOFF.
15880 # Initialization string 2 sets:
15893 # Initialization string 2 sets:
15905 # Reset string 2 sets:
15918 # Enhanced DG mode with changes to be more UNIX compatible.
15947 # Initialization string 1 sets:
15953 # Reset string 1 sets:
15966 # Initialization string 3 sets:
15968 # Reset string 2 sets:
15996 # These add intelligent features like insert/delete to the D200 series.
15998 # Initialization string 2 sets:
16008 # Reset string 1 sets:
16010 # Reset string 2 sets:
16031 # Initialization string 1 sets:
16045 # Initialization string 2 sets:
16054 #	Reset string 1 sets:
16058 # Reset string 2 sets:
16079 # Initialization string 2 sets:
16087 # Reset string 2 sets:
16113 # Initialization string 1 sets:
16127 # Reset string 1 sets:
16173 d412-unix-s|d462-unix-s|d412+s|d462+s|Data General DASHER D412+/D462+ in Unix mode with status line,
16182 #	Relative cursor motions are confined to the current window,
16196 d413-unix-s|d463-unix-s|Data General DASHER D413/D463 in DG-UNIX mode with status line,
16207 d414-unix-s|d464-unix-s|Data General D414/D464 in DG-UNIX mode with status line,
16223 d430c-unix-s|d430-unix-s|Data General D430C in DG-UNIX mode with status line,
16233 d430c-unix-s-ccc|d430-unix-s-ccc|Data General D430C in DG-UNIX mode with status line and configurab…
16241 # Initialization string 1 sets:
16266 # Initialization string 2 sets:
16311 # Initialization string 1 sets:
16333 # to the factory in Pennsauken, NJ.  The factory was sold to a PCB board
16373 # (dmchat: ":MT:" changed to ":km:" -- esr)
16378 # (dm3025: ":MT:" changed to ":km:" -- esr)
16425 # dm80/1 is a vt100 lookalike, but it doesn't seem to need any padding.
16457 # These aren't end-all Excel termcaps; but do insert/delete char/line
16460 # E is for 'Excel') as # a name.  This was done to distinguish the entries
16461 # from the other Datamedias in use here, and yet to associate a model of
16512 # (ts100: I added <rmam>/<smam> based on the init string -- esr)
16585 # (f100: added empty <acsc> to suppress a tic warning;
16586 # made this relative to adm+sgr -- note that <invis> isn't
16587 # known to work for f100 but does on the f110. --esr)
16607 # is lost! It cannot be remapped in vi because it is necessary to enter
16608 # a ^V to to quote the ^V that is being remapped!!!
16610 # f110/f200 users will have to decide whether
16611 # to lose the down cursor key or the quoting capability. We will opt
16613 # is not generally applicable to most interactive applications
16631 # (f200: added <acsc> to suppress tic warnings re <smacs>/<rmacs> --esr)
16650 # The f200 has the ability to reprogram the down cursor key. The key is
16651 # reprogrammed to ^J (linefeed). This value is remembered in non-volatile RAM,
16652 # so powering the terminal off and on will not cause the change to be lost.
16670 # terminal, than can be switched to X graphics mode (driven over the serial
16671 # line) by an escape sequence.  No info on this beast yet.
16672 # (go140: I added <rmam>/<smam> based on the init string -- esr)
16691 # Hacked up vt200 termcap to handle GO-225/VT220
16693 # (go225: I added <rmam>/<smam> based on the init string -- esr)
16718 # so we must have early Superbee2 (Model 600, according to phone conversation
16728 # slow at low speeds; also spaces get converted to \040 which is
16731 # The <nl> string is designed for last line of screen ONLY; cup to
16732 # 25th line corrects the motion inherent in scrolling to Page 1.
16734 # There is one understood bug. It is that the screen appears to
16735 # pop to a new (blank) page after a <nel>, or leave a half-line
16736 # ellipsis to a quad that is the extra 48 memory locations. The
16737 # data received is dumped into memory but not displayed.  Not to
16740 # relative to MEMORY of window, nothing is lost; but beware of
16752 # into a backspace key. In use ESC must be pressed twice (to send)
16753 # and sending ^C must be prefixed by ESC to avoid that weird
16765 # The circuits MUST be modified to eliminate the SPOW latch.
16767 # the p.c. connection to Pin 5 and strap Pin 5 to Pin 8 of that
16772 # NOTE WELL that the rear panel switch should be set to CR/LF,
16801 # the switch in the back set to CRLF instead of AEP.  This description
16804 # the other two if you want to try to hit that tiny escape key.
16805 # This description is tricky: being able to use cup depends on there being
16806 # 2048 bytes of memory and the hairy <nl> string.
16848 # set tab is ^F, clear (one) tab is ^V, no way to clear all tabs.
16849 # good grief - does this entry make :sg:/:ug: when it doesn't have to?
16850 # look at those spaces in <rmso>/<smso>.  Seems strange to me...
16878 # (8675: changed k10, k11...k16 to k;, F1...F6 -- esr)
16884 # (8686: changed k10, k11...k16 to k;, F1...F6; fixed broken continuation
16898 # Hazeltine appears to be out of the terminal business as of 1995.  These
16907 # As late as 1993, manuals for the terminal product line could still be
16920 # Since <cuf1> is blank, when you want to erase something you
16921 # are out of luck.  You will have to do ^L's a lot to
16939 # freakout with out-of-range args and tn3270.  No hz since it needs to
16949 # h1510 assumed to be in sane escape mode.  Else use h1500.
16982 # Note: the h1552 appears to be the first Hazeltine terminal which
16984 # Be sure the auto lf/cr switch is set to cr.
17000 # to work ok. There is one problem though if one types a lot of garbage
17001 # characters very fast vi seems not able to keep up and hangs while trying
17002 # to insert. That's in insert mode while trying to insert in the middle of
17003 # a line. It might be because the Esprit doesn't have insert char and delete
17004 # char as a built in function. Vi has to delete to end of line and then
17005 # redraw the rest of the line.
17020 # Vi it seems always wants to send a control J for "do" and it turned out
17054 ibm327x|line mode IBM 3270 style,
17166 # (ibmmono: this had an unknown `sb' boolean, I changed it to `bs'.
17321 # "Megapel" refers to the display adapter, which was used with the IBM RT
17409 # (infoton: used to have the no-ops <lh#0>, <lw#0>, <nlab#0> -- esr)
17431 #ctrl-N          Disable xon/xoff to host
17432 #ctrl-O          Enable xon/xoff to host
17436 #ctrl-Z          Clear unprotected data to insert char
17444 #                p1 = scroll top    line:  20h - 37h
17445 #                p1 = scroll bottom line:  20h - 37h
17456 #ESC +           clear unprotected data to insert char
17457 #ESC ,           clear unprotected data to half intensity spaces
17458 #ESC - p1 p2 p3 p4     address cursor to page, row, column:
17477 #ESC 4           send unprotect line to cursor
17478 #ESC 5           send unprotect page to cursor
17479 #ESC 6           send line to cursor
17480 #ESC 7           send page to cursor
17487 #ESC :           clear unprotected data to null
17488 #ESC ;           clear unprotected data to insert char
17491 #ESC = p1 p2     address cursor to row, column
17504 #ESC E           line insert
17519 #ESC N           set page edit (clear line edit)
17520 #ESC O           set line edit (clear page edit)
17523 #ESC R           line delete
17525 #ESC T           erase line to insert char
17531 #ESC V 2 n       define line attribute:
17549 #ESC Y           erase page to insert char
17551 #ESC Z n         send user/status line:
17552 #                n = 0   send user line
17553 #                n = 1   send status line
17604 #ESC f text CR   load user status line with 'text'
17606 #ESC g           display user status line on 25th line
17607 #ESC h           display system status line on 25th line
17616 #ESC m           save current config to NVRAM
17622 #ESC o p1 p2     set characters/line and attribute:
17623 #                p1 = 0  80 chars/line
17625 #ESC o p1 p2     set characters/line and attribute:
17626 #                p1 = 0  80 chars/line
17627 #                p1 = 1  132 chars/line
17636 #ESC t           erase line to null
17641 #ESC y           erase page to null
17661 #                        (use Ctrl-P to escape ^P, ^Y )
17671 # This entry uses page 0 and is monochrome; I'm not brave enough to try
17672 # to make color work without a test terminal.  The <am> capability is a guess.
17673 # The initialization string sets conversation mode, blinking underline cursor,
17674 # full duplex, parallel attribute mode, display user status line, white
17697 # ISC used to sell OEMed and customized hardware to support ISC UNIX.
17735 # Kimtron seems to be history, but as March 1998 these people are still
17767 #    firmware revs prior to SP51
17768 # 2) Make sure to use abm85h entry if the terminal is in 85h mode and the
17776 #    <is2>.  Note that 920E mode does not have software commands to toggle
17779 # 4) <flash> attribute is nice, but seems too slow to work correctly
17783 # 6) auto new-line should be on (selectable from setup mode only)
17821 # Renamed TB=^I to :ta:, BE=^G to :bl:, BS=^H to :kb:, N to :kS: (based on the
17823 # identical to :mh:.  Removed :ES=\EGD: which is some kind of highlight
17843 # This was a line of terminals made by McDonnell-Douglas Information Systems.
17845 # to move primary names of the form p[0-9] * to aliases, and to comment out
17863 #       Limited numbers. Similar functionality to P5 (except attributes?).
17873 #       Latest range, functionally very similar to the P9.  The P14 has a
17908 #  Horizontal cursor qualifiers of NUL, XON and XOFF are mapped to the next
17928 #  Horizontal cursor qualifiers of NUL, XON and XOFF are mapped to the next
17930 #  Cursor key definitions removed because they interfere with vi and csh keys.
17962 # Similar definition to p4. Uses ANSI cursor motion to avoid network problems.
17965 #  Rev-index removed; can't send nulls to terminal in 8-bit modes.
17973 # Similar definition to p7. Uses ANSI cursor motion to avoid network problems.
17978 #  Rev-index removed; can't send nulls to terminal in 8-bit modes.
17990 #  Rev-index removed; can't send nulls to terminal in 8-bit modes.
18003 #  Fixes to deal with terminal firmware bugs:
18004 #  . 'ri' uses insert-line since rev index doesn't always
18006 #  . 'fsl' & 'dsl' use illegal char since cr is actioned wrong on line 25
18010 #  . Graphic characters (defaults correctly to vt100)
18057 # Similar to p8 definition.
18093 # Similar to p8 definition.
18127 # Similar to p8 definition.
18162 # The mime1 entries refer to the Microterm Mime I or Mime II.
18163 # The default mime is assumed to be in enhanced act iv mode.
18194 # uc was at one time disabled to get around a curses bug, be wary of it
18235 # scrolls down and insert mode works without redrawing the rest of the line
18236 # to the right of the cursor. This is done with a bit of a kludge using the
18237 # exit graphics mode to get out of insert, but it does not appear to hurt
18256 # also added <rmam>/<smam> based  on the init string -- esr)
18286 # (ergo400: added <rmam>/<smam> based on the init string -- esr)
18305 # For what happened to that group, see the ADDS section.
18312 # identical, except for case, to lower-case ones.  I also uncommented the acsc
18345 # Mapping to ASCII character set ('acsc' capability) can also be
18568 # (escape sequence) used to set color attributes, one of the fore/background
18569 # colors must be preset to a given value. I have set the background color to
18571 # 'setf' definition.  The escape sequence to set color attributes is
18574 # The capability 'op' sets colors to green on black (default combination).
18578 #	    capability and recompile if you wish to have it included.
18632 # Mapping to ASCII character set ('acsc' capability) can also be
18634 # In addition, color capabilities have been added to this file.  The drawback,
18635 # however, is that the background color has to be black.  The foreground colors
18639 #	    with the 'pairs' capability defined as below.  If you wish to
18694 # Mapping to ASCII character set ('acsc' capability) can also be
18871 # 1   - End of line entry disabled/enabled
18891 # multiplying the 0 or 1 by a correct factor to shift) so the bias character,
18892 # '@' is (effectively) "or"ed with each attribute to generate the proper third
18893 # character in the <ESC>0 sequence.  The <sgr> string implements the following
18904 # From <root@goliath.un.atlantaga.NCR.COM>, init string hacked by SCO.
18927 #	   The user can enter a CTRL-B to get out of this locked state.
18928 # In <hpa>, we want to output the character given by the formula:
18963 # keep the same keyboard layout), add an optional 25-line mode, replace the DIP
18964 # switches with a menu and remove the "lock line" feature (ESC !  1 and ESC !
18965 # 2), here is the NDR 9500.  Even the line-lock, albeit disabled, is
18992 ndr9500-nl|NDR 9500 with no status line,
18997 ndr9500-25|NDR 9500 with 25th line enabled,
19000 ndr9500-25-nl|NDR 9500 with 25 lines and no status line,
19014 ndr9500-mc-nl|NDR 9500 with magic cookies and no status line,
19019 ndr9500-25-mc-nl|NDR 9500 with 25 lines and magic cookies and no status line,
19079 # Sperry Univac has merged with Burroughs to form Unisys.
19084 # provided is comparable to the DEC vt100.
19085 # (uts30: I added <rmam>/<smam> based on the init string -- esr)
19107 # Tandem builds these things for use with its line of fault-tolerant
19131 # Tandy has a line of VDTs distinct from its microcomputers.
19197 # (tek4013: added <acsc> to suppress tic warnings re <smacs>/<rmacs> --esr)
19203 # (tek4015: added <acsc> to suppress tic warnings re <smacs>/<rmacs> --esr)
19209 # (tek4015-sm: added <acsc> to suppress tic warnings re <smacs>/<rmacs> --esr)
19214 # You need to have "stty nl2" in effect.  Some versions of tset(1) know
19215 # how to set it for you.
19220 # it to flash, change the letter 'H' to 'P' in the :so: field.
19230 # on keyboard don't work.  You have to hit BREAK twice to get
19236 # <el> was commented out since vi and rogue seem to work better
19258 tek4025-17|tek 4025 17 line window,
19260 tek4025-17-ws|tek 4025 17 line window in workspace,
19272 #	!COM 29			# NOTE: changes command character to GS (^])
19279 # Other modes may be set according to communication requirements.
19281 # Insert-character cannot be made to work on both top and bottom rows.
19282 # Clear-to-end-of-display emulation via !DLI 988 is too grotty to use, alas.
19283 # There also seems to be a problem with vertical motion, perhaps involving
19284 # delete/insert-line, following a typed carriage return.  This terminal sucks.
19285 # Delays not specified; use "stty ixon -ixany" to enable DC3/DC1 flow control!
19304 # Here's the command file that I use to get rogue to work on the 4025.
19306 # not try to scroll, or cursor addressing won't work.  Also, you can't
19338 # (tek4105-30: I added <rmam>/<smam> based on the init string -- esr)
19339 tek4105-30|4015 emulating 30 line vt100,
19378 # delays are specified; use "stty ixon -ixany" to enable DC3/DC1 flow control!
19426 # delays are specified; use "stty ixon -ixany" to enable DC3/DC1 flow control!
19478 # see the note attached to tek4207.
19489 # off the bottom line.  Out of the dialog area, ^K moves it up, but there
19490 # is no way to scroll.
19496 # 4112 in non-dialog area pretending to scroll. It really wraps
19497 # but vi is said to work (more or less) in this mode.
19518 tek4112-5|4112 in 5 line dialog area,
19520 # (tek4113: this used to have "<cuf1=\LM1\s\LM0>", someone's mistake;
19524 # to be 4-digit octal -- esr)
19525 tek4113|tektronix 4113 color graphics with 5 line dialog area,
19532 tek4113-34|tektronix 4113 color graphics with 34 line dialog area,
19535 # :ns: left off to allow vi visual mode. APL font (:as=\E^N:/:ae=\E^O:) not
19537 # :vb: needs enough delay to let you see the background color being toggled.
19548 # (tek4115: :bc: renamed to :le:, <rmam>/<smam> added based on init string -- esr)
19585 # Note that this entry uses all 34 lines and sets the cursor color to green.
19600 # note that if jcoker wasn't confused you may be able to use this one.
19634 # I added the is string - straight Unix has ESC ; in the login
19635 # string which sets a ct8500 into monitor mode (aka 4025 snoopy
19636 # mode). The is string here cleans up a few things (but not
19650 # is placed in the 80'th column, but when we are attempting to type
19651 # the 81'st character on the line.  (esr: hmm, this is like the vt100
19656 # table by setf/setb. All colors are reset to factory specifications by oc.
19657 # The <initc> cap uses RGB notation to define colors.  for arguments 1-3 the
19707 # The earliest UNIXes were designed to use these clunkers; nroff and a few
19708 # other programs still default to emitting codes for the Model 37.
19720 # There are known to be at least three flavors of the tty40, all seem more
19726 # it - it's null here to prevent it from showing the BL character.
19729 # to get crlf, even if <cr> is not ^M.)
19748 # You can add <is2=\E<> to put this 40-column mode, though I can't
19749 # for the life of me think why anyone would want to.
19764 # to program one...)
19811 # supposed to emulate a vt-100, but emulates the wraparound bug incorrectly,
19812 # doesn't support scrolling regions, ignores add line commands, and ignores
19813 # delete line commands. Consequently, the resulting behavior looks like a
19820 # I've found that my KayPro II, running MDM730, continues to emulate an
19882 # The modem interface is permitted to discard LF (maybe DC1), otherwise
19883 # passing characters to the 80-column firmware via COUT (PR#3 assumed).
19884 # Auto-wrap does not work right due to newline scrolling delay, which also
19938 # (apple-videx: this used to be called DaleApple -- esr)
19967 # language card.  SYSTEM.MISCINFO is assumed to be the same as that
19969 # using SETUP to 80 columns.  Note that the right arrow is not mapped in
19994 apple-videx2|Apple II+ w/ Videx card (similar to Datamedia h1520),
20028 # (lisa: changed <cvvis> to <cnorm> -- esr)
20050 # settings are in non-volatile memory and don't need to be reset upon login.
20052 # You can type "reset" to get them set.
20072 # Although MacTerminal has insert/delete line, it is commented out here
20073 # since it is much faster and cleaner to use the "lock scrolling region"
20074 # method of inserting and deleting lines due to the MacTerminal implementation.
20075 # Also, the "Insert/delete ch" strings have an extra character appended to them
20076 # due to a bug in MacTerminal V1.1.  Blink is disabled since it is not
20132 # C-128, VIC-20).  The C-64 is said to have been the most popular machine
20139 # to '\0xx' entries since a couple of people mentioned losing '^x' sequences.
20145 #     This one appears to fix a problem I always had with a line ending
20146 #     at 'width+1' (I think) followed by a blank line in vi. The blank
20147 #     line tended to disappear and reappear depending on how the screen
20149 #     something like a Dnet Fterm with the window sized to some peculiar
20151 # :k0=\E9~:	map F10 to k0 - could have F0-9 -> k0-9, but ... F10 was 'k;'
20153 # also added empty <acsc> to suppress a warning --esr)
20172 # (amiga: added empty <acsc> to suppress a warning.
20243 #	I'm trying to write a termcap for a commodore b-128, and I'm
20244 # having a little trouble. I've had to map most of my control characters
20245 # to something that unix will accept (my delete-char is a ctrl-t, etc),
20249 # Jove knows what I want, but I don't know what it's sending to me (it
20250 # isn't thats bound to next-line in jove).
20252 # DAG -- I changed his "^n" entries to "\n"; see if that works.
20279 # Osborne, but with the 80-column upgrade, it's too easy to
20284 # 52, 80, or 104 characters wide; default to 80 for compatibility
20287 # The tab is destructive on the Ozzie; make sure to 'stty -tabs'.
20306 # Similar to tvi920
20323 # machines that tried to emulate the UNIX look'n'feel.  Coherent and Venix
20325 # Memory-segmentation limits and a strong tendency to look like V7 long after
20327 # and Minix were ported to 32-bit Intel boxes, only to be run over by a
20328 # steamroller named `Linux' (which, to be fair, traces some lineage to Minix).
20398 # According to the Coherent 2.3 manual, the PC console is similar
20399 # to a z19. The differences seem to be (1) 25 lines, (2) no status
20400 # line, (3) standout is broken, (4) ins/del line is broken, (5)
20411 # According to the Venix 1.1 manual, the PC console is similar
20412 # to a DEC vt52.  Differences seem to be (1) arrow keys send
20413 # different strings, (2) enhanced standout, (3) added insert/delete line.
20430 # It may be used as a terminal by putting it in "line" mode as seen on
20432 # Initialization is similar to CIT80. <is2> will set ANSI mode for you.
20433 # Hardware tabs set by <if> at 8-spacing.  Auto line wrap causes glitches so
20463 # development system (which I used even in 1993 to program the steering
20489 # \E\:1}	switch to te'le'informatique mode (ascii terminal/ISO 6429)
20493 # \Ec		reset: G0 U.S. charset (to get #,@,{,},...), 80 cols, clear screen
20494 # \E)0		G1 DEC set (line graphics)
20805 # 2. Native brown color corrected to good yellow color.
20975 # David Breneman <daveb@dgtl.com> has this to say:
20982 #  Jerq to sleep, cloned its parts, and the Blit was brought forth unto the
20987 #  it started to be shown outside of the halls of the Bell Labs Research
20989 #  not remain. So it was renamed to be Blit. This was in late 1981.
21029 # display, and a 68000 to run it.  You could download code and run it on
21032 # the cpu to bitblt pixels to scroll, it took longer than the refresh
21036 # Beranek and Neuman) sold at most a few hundred of them to the real
21041 # with scrolling in vi can only be fixed by getting BBN to put
21045 # I always thought the problem was related to the terminal
21048 # paints lines in on the bottom line of the screen, so you get
21075 # (bg1.25: I added <rmam>/<smam> based on the init string -- esr)
21136 #	MPW map status line window:	^[PY99:98^[\
21137 #	SCP select status line:		^[[0;98v
21148 #	MC start routing to printer:	^[[5i
21149 #	MC stop routing to printer:	^M^[[4i
21153 # dku7102, tws2102, and tws models 2105 to 2112
21213 #  They are used in string capabilities with vt220-320 emulation mode.
21226 #	SCS select G1 = line-graphic:	esc ) 0
21235 #	SM LNM New line :		esc [ 2 0 h
21339 #	SCS select G1 = line-graphic:	esc ) 0
21348 #	SM LNM New line :		csi 2 0 h
21381 # (bq300-8: <cub1>,<cuf1>,<cuu1>,<cud1>,<dl1>,<il1> to get under 1024 --esr)
21517 #	SLS status line set		^[w
21518 #	SLR status line reset		^[v
21519 #	SLL status line lock		^[O
21588 # window back to be the whole screen, and puts the cursor at just
21589 # below the small window. I defined <cnorm> and <civis> to really turn
21640 # Datapoint is gone.  They used to be headquartered in Texas.
21655 # CONTROL-INT-INT to take the terminal off-line, and type (opt).
21658 # shift-F1 to shift-F5 as F6 to F10 (unshifted F1 to F5 are in
21675 #         ctrl-N           horizontal line
21676 #         ctrl-O           vertical line
21732 # (gigi: I added <rmam>/<smam> based on the init string, corrected cub1 -- esr)
21747 # DEC PRO-350 console (VT220-style).  The 350 was DEC's attempt to produce
21750 # a format program, so you had to buy pre-formatted floppies from DEC at
21831 # (ddr: I added <rmam>/<smam> based on the init string -- esr)
21860 # (ps300: changed ":pt@:" to "it@" -- esr)
21900 # 2 = wrap at end of line (0 = no wrap)
21911 # (h19: I added <rmam>/<smam> based on the init string;
21912 # also added empty <acsc> to suppress a tic warning -- esr)
21931 # also added empty <acsc> to suppress a tic warning --esr)
21940 # On the Z19, the only way I have found around this problem is to remove
21942 # help (I have tried up to 20000).  Removing <il1=\EL$> and <dl1=\EM$>
21967 # The problem with declaring an H19 to be synonymous with a Z29 is that
21969 # to put the Z29 into insert mode and insert text at 9600 baud. It
21970 # even loses worse if the program attempts to insert tabs at 9600
21971 # baud. Adding padding to text that is inserted loses because in
21972 # order to make the Z29 not die, one must add so much padding that
21973 # whenever the program tries to use insert mode, the effective
21976 # What program would want to put the terminal into insert mode
21979 # Emacs. Emacs seems to want to do the mathematically optimal
21981 # When it is about to output a line on top of a line that is
21982 # already on the screen, instead of just killing to the end of
21983 # the line and outputting the new line, it compares the old line
21984 # and the new line and if there are any similarities, it
21985 # constructs the new line by deleting the text on the old line
21987 # text into the line to transform it into the new line that is
21988 # to be displayed. The Z29 does not react kindly to this.
21993 # line will be really slow", you say. Well there is a sort of a
21994 # solution to that too. There is an insert character option on
21997 # character, and changing it back to H19 mode. All this takes 12
21998 # characters. Pretty expensive to insert one character, but it
21999 # works. Either Emacs doesn't try to use its inserting hack when
22003 # (z29: added empty <acsc> to suppress a tic warning, merged in
22004 # status line capabilities from BRL entry --esr)
22021 # the world is stable. <rs1> causes the terminal to be reset to the state
22118 # Nevada, that flourished from the mid-70s to mid-80s.  They made S-100
22119 # bus/Z80 hardware and a line of terminals called Ultimas.
22123 ims950-b|bare ims950 no init string,
22181 # past.  They used to be reachable at:
22218 # These people used to be reachable at:
22235 # dated 1984.  According to the manual, it featured Tek 4010/4014
22262 # If we set TERM=vt100, and set the Modgraph screen to 24 lines, setting a
22263 # mark and using delete-to-killbuffer work correctly.  However, we would
22264 # like normal mode of operation to be using a Modgraph with 48 line setting.
22266 # the setting mark and delete-to-killbuffer results in the deletion of only
22267 # the line the mark is set on.
22268 # We've discovered that the delete-to-killbuffer works correctly
22269 # with TERM=mod and screen set to 80x48 but it's not obvious.  Only
22270 # the first line disappears but a ctrl-l shows that it did work
22289 # S100-bus machines.  They used to be reachable at:
22357 # Hardware tabs are assumed to be every 8 columns; they can be set up by the
22360 # delays are specified; use "stty ixon -ixany" to enable DC3/DC1 flow control!
22406 # For use with graphics software, all graphics modes should be set to factory
22409 # to enable DC3/DC1 flow control!
22455 # to me, but it took a clue from their service department to make
22461 # one evening, tossing back a few and trying to decide what to
22462 # call their new company and what to use for a logo.
22501 # Synertek used to make ICs, various 6502-based single-board process
22504 # first to have a "terminal-on-a-keyboard", where the terminal itself
22508 # was a 40x24 terminal that could connect to a standard TV through a
22510 # could be upgraded to the 2/80 by adding 2 2114 SRAMs and a new ROM).
22522 # real broken, at least according to the folks I've talked to about it.
22525 # slow (it takes like 100ms for the KTM-2 to clear the screen...) And
22534 # (they've gone to ASICs and FPGAs), but I doubt they're in the computer
22538 # Tested, seems to work fine with vi.
22555 # <smkx>/<rmkx> have nothing to do with arrow keys.
22557 # Seems to be no way to get rid of status line.
22558 # The manual for this puppy was dated June 1981.  It claims to be VT52-
22586 # There was a newer line of Telerays (Model 7, Model 20, Model 30, and
22590 # to the front if you have either.  A dumb teleray with the cursor stuck
22622 # converts the keypad to programmable function keys, and has other goodies.
22636 # teleray 16 - map the arrow keys for vi/rogue, shifted to up/down page, and
22639 # for line/local. Do the magic appropriate to make the page shifts work.
22668 # Refer to:
22672 # (cpi) (80 characters per line) by default, and can be switched to/from 17.0
22811 # had just <smso>=\EG6 which I think means standout was supposed to be
22923 # I have a UNIX PC which I use as a terminal attached to my Linux PC.
22932 # uses an 80x24 window (and doesn't support seem to support a 132-column
22941 # machine, tam also supported VT100-compatible terminals attached to the
22947 # (If you're interested in adding some of the tam calls to ncurses, btw, I
22949 # used an extended terminfo format to describe key sequences, special
22955 # to use ncurses on the UNIX PC, since ncurses doesn't properly handle the
22957 # seem to have been built from the manual describing the VT100 sequences.
22962 # up to 8 custom fonts to be loaded at any given time. This means that
22966 # manpage), should you wish to do so:
22974 # Graphics for line drawing are not reliably found at *any* character
22977 # universal. So ASCII line drawing is in order if smacs/rmacs are enabled.
22991 #	Somewhat similar to a vt100-am (but different enough
22992 #	to redo this from scratch.)
23002 #	*     The call to ioctl with the argument WIOCLFONT is the key
23003 #	*     to this routine.  For more information, see window(7) in
23006 #	#include <string.h>		/* needed for strcpy call */
23013 #	*     similar to that of the Teletype 5425 terminal.  To view
23098 # supposed to be a high performance graphic accelerator, but it was
23099 # late to market and barely appeared before faster dumb frame buffers
23112 # Fortune made a line of 68K-based UNIX boxes that were pretty nifty
23126 # to force both magic cookie glitches off.  Once upon a time, I
23129 # EN/PD/PU/CO/CF/RT according to the XENIX/TC mappings, but not HM/DL/RF/RC.
23151 # Masscomp has gone out of business.  Their product line was purchased by a
23156 # (masscomp: ":MT:" changed to ":km:";  -- esr)
23197 # (ibmpcx: this entry used to be known as ibmx.
23231 # so we have to save their status in the static registers A, B and H
23232 # respectively, to be able to restore them when color changes
23235 # rather than simply  entering them.  Thus we have to check the
23236 # static register B and H to determine the status, before sending the
23238 # 4. <sgr0> now must set the status of all 3 register (A,B,H) to zero
23241 # we would have to use \E&bn sequence to turn off colors (as well as all
23243 # static variable.  If someone really needs this mode, they would have to
23253 # 8. turn on background: similar to turn on foreground above
23309 # (h19k: changed ":pt@:" to ":it@"
23337 # (xtalk: I added <rmam>/<smam> based on the init string.
23338 xtalk|IBM PC with xtalk communication program (versions up to 3.4),
23355 # Note - insert mode commented out - doesn't seem to work on AT&T PC.
23399 # I have been unable to get tabs set in all 96 lines - it always leaves at
23400 # least one line with no tabs in it, and once you tab through that line,
23402 # (dtc382: change <rmcup> to <smcup> -- it  just does a clear --esr)
23539 # design, but isn't.  The structure is similar to the Direct terminals,
23600 # NEXT SCREEN. The key definitions k6-k9 represent the PF1 to PF4 keys.
23603 # I had a couple of scopes (3221) like this once where I used to work, around
23605 # by an outfit called LANPAR Technologies, and were meant to me DEC VT 220
23609 # (3221) cost about $1500 each, and one was always broken -- had to be sent
23610 # back to the shop for repairs.
23615 # appear on the bottom.  I don't remember the VT-240s being able to do that.
23621 # I added <rmam>/<smam> based on the init string -- esr)
23639 # terminfo, you're not actually supposed to specify both ich/ich1 and rmir/smir
23647 # entries that suppress ich/ich1.  And upgrade to ncurses!
23653 # ECMA-48 standards are said to be almost identical, but are not the same
23656 # You can obtain ECMA-48 for free by sending email to helpdesk@ecma.ch
23665 # am not certain) that these are effectively identical to ECMA-6 and ECMA-35
23676 # discussion of some terminfo-related issues, and updates to capture ECMA-48
23682 # SPC for space.  Pn stands for a single numeric parameter to be inserted
23787 # RIS  Reset to Initial State    \E c                -         Fs     -
23850 # being assigned functions relevant to terminal control there (they
23851 # referred to other standards such as ISO 1745 or ECMA-35).  They are listed
23856 # (B) There seems to be some confusion abroad between CHA and HPA.  Most
23861 # (C) CHT corresponds to terminfo (tab).  Usually it has the value ^I.
23873 #	1 = set line tab,
23875 #	3 = clear line tab,
23876 #	4 = clear all char tabs on current line,
23878 #	6 = clear all line tabs.
23897 #	0 = clear to end,
23903 # (L) The GSM parameters are vertical and horizontal parameters to scale by.
23973 #	0 = start xfer to primary aux device,
23975 #	2 = start xfer to secondary aux device,
23977 #	4 = stop relay to primary aux device,
23978 #	5 = start relay to primary aux device,
23979 #	6 = stop relay to secondary aux device,
23980 #	7 = start relay to secondary aux device.
24041 #             an Escape sequence (F from 3/0 to 7/14 in the ASCII table)
24042 #             a control sequence (F from 4/0 to 7/14)
24045 #        2/0 to 7/14) in the ASCII table
24053 #        This is too complicated to explain fully here, so read Jim Fleming's
24058 #        (Fe ranges from 4/0 to 5/15)
24063 #        designated C0 and C1 control sets (Fs ranges from 6/0 to 7/14)
24065 # I      is an Intermediate character from 2/0 to 2/15 (inclusive) in the
24068 # P      is a parameter character from 3/0 to 3/15 (inclusive) in the ASCII
24071 # Pn     is a numeric parameter in a control sequence, a string of zero or
24072 #        more characters ranging from 3/0 to 3/9 in the ASCII table
24077 #        3/0 to 3/9 and includes 3/11
24079 # *      Not relevant to terminal control, listed for completeness only.
24083 # A format effector specifies how following output is to be displayed.
24084 # An editor function allows you to modify the display.  Informally
24088 # cursor or equivalent) one space to the left would be useful when you want to
24090 # overlaid. Control-H, the Backspace character, is actually supposed to be a
24092 # nonstandard fashion, as an editor function, deleting the character to the
24093 # left of the cursor and moving the cursor left. When Control-H is assumed to
24111 #      Erase from cursor to end of line           Esc [ 0 K    or Esc [ K
24112 #      Erase from beginning of line to cursor     Esc [ 1 K
24113 #      Erase line containing cursor               Esc [ 2 K
24114 #      Erase from cursor to end of screen         Esc [ 0 J    or Esc [ J
24115 #      Erase from beginning of screen to cursor   Esc [ 1 J
24118 # Some brain-damaged terminal/emulators respond to Esc [ J as if it were
24121 # The VT100 responds to receiving the DA (Device Attributes) control
24140 # where Pl is the line number and Pc is the column number (in decimal).
24168 #   supposed to enable bright background.
24171 #   when you try to set a "bright brown" (yellow) background with attribute
24176 # * Some older versions of ANSI.SYS have a bug that causes them to require
24184 # These recommendations are optional.  IBCS2 allows the leading escape to
24197 #	CSI <n>;<m>H	(cup)	cursor to line n and column m
24198 #	CSI <n>;<m>f		cursor to line n and column m
24208 #	CSI <n>K	(el)	erase in line
24209 #	CSI <n>L	(il)	insert line(s)
24215 #	CSI <n>`		cursor to column n on line
24217 #	CSI <n>d	(vpa)	cursor to line n
24224 #	CSI u			restore cursor position to saved value
24238 #	CSI 8		(rc)	(rc) restore cursor position to saved value
24240 #	CSI Q<n><string>	define function key string
24241 #				(string must begin and end with delimiter char)
24246 # everything iBSC2 has to say about terminal escape sequences; there is
24247 # no further discussion of their meaning or how to set the parameters
24257 # there.  We try to describe them here.
24294 #	GG			:as:/:ae: glitch (analogous to :sg:/:ug:)
24306 #	GV	Gv	vertical line             |
24307 #	GH	Gv	horizontal line       -   _
24318 # These were invented to take advantage of the IBM PC ROM character set.  One
24319 # can compose an acsc string from the single-width characters as follows
24327 # nonstandard capabilities.  Its signature is the KM capability, used to name
24330 # documentation, seem to establish that BO=:mr: (start reverse video), DS=:mh:
24337 # The HP library (as of mid-1995, their term.h file version 70.1) appears to
24338 # have the System V capabilities up to SVr1 level.  After that, it supports
24339 # two nonstandard caps meml and memu corresponding to the old termcap :ml:,
24352 # rvert, lvert.   Some of these are identical to XPG4/SVr4 equivalents:
24354 # renamed (to kcbt and kslt).  The places in the box[12] capabilities
24355 # correspond to acsc chars, here is the mapping:
24369 # The box2 characters are the double-line versions of these forms graphics.
24392 # infocmp to manipulate user-definable capabilities.  Those that are intended
24413 #       E0   (str)   Switch charset 'G0' back to standard charset.
24414 #       S0   (str)   Switch charset 'G0' to the specified charset.
24418 # AX is relatively straightforward; it is interpreted by ncurses to say that
24419 # SGR 39/49 reset the terminal's foreground and background colors to their
24426 # a) OSC 1 sets the title string, e.g., for the icon.  Recent versions of
24438 #    does not support.  Therefore, screen uses a hardcoded string to work
24442 # g) screen also uses the feature to decide whether to pay attention to other
24443 #    xterm-related features which are unrelated to the description in the
24447 # it is appropriate to use it in the derived terminal descriptions such as
24450 # The other ISO-2022 features are rarely used, but provided here to make
24458 # though they did not follow xterm's lead in xterm patch #167 (in 2002), to make
24473 # Cr is a string capability which resets the cursor color
24474 # Cs is a string capability which sets the cursor color to a given value.
24475 #    The single string parameter is the color name/number, according to the
24480 # Se resets the cursor style to the terminal power-on default.
24481 # Ss is a string capability with one numeric parameter.  It is used to set the
24482 #    cursor style as described by the DECSCUSR function to a block or
24484 # TS is a string capability which acts like "tsl", but uses no parameter and
24485 #    goes to the first column of the "status line".
24486 # XM is a string capability which overrides ncurses's built-in string which
24502 # gsbom/grbom are used to enable/disable real bold (not intensity bright) mode.
24510 #    support VT100 SI/SO when processing UTF-8 encoding.  Set this to a nonzero
24511 #    value to enable it.
24523 # capabilities has been moved to comments.  Some all-numeric names of older
24526 # I changed :MT: to :km: (the 4.4BSD name) everywhere.  I commented out some
24533 # making perfectly synchronized mistakes which I then failed to catch by
24536 # Major version number bumps correspond to major version changes in ncurses.
24561 #	* Emacs uses :so:, not :mr:, for its mode line.  Fix linux entry
24562 #	  to use reverse-video standout so Emacs will look right.
24563 #	* Added el1 capability to ansi.
24564 #	* Added smacs/rmacs to ansi.sys.
24579 #	* Fix linux & AT386 sgr strings to do A_ALTCHARSET turnoff correctly.
24581 #	  to force a particular height.
24603 #	* Added the usual set of function-key mappings to ANSI entry.
24609 #	* Change some \0 escapes to \200.
24614 #	* Changed mandatory to advisory padding in many entries.
24615 #	* Replaced HP entries up to hpsub with purpose-built ones.
24628 #	* Moved old superbee entry to superbee-xsb, pulled in new superbee
24631 #	* Added lm#0 to cbunix and virtual entries.
24648 #	* Changed tic -C logic to dump all capabilities used by GNU termcap.
24652 #	* Yet another tic change, and a couple of entry tweaks, to reduce the
24653 #	  number of long (> 1024) termcap translations back to 0.
24656 #	* Added kf13-kf20 to Linux entry.
24663 #	* Name field changes to shorten some long entries.
24669 #	* Change linux entry to use smacs=\E[11m and have an explicit acsc,
24686 #	* Add 50% cut mark as a desperate hack to reduce tic's core usage.
24687 #	* xterm doesn't try to use application keypad mode any more.
24691 #	* Added csr capability to linux entry.
24693 #	* Added vt102-nsgr to cope with stupid IBM PC `VT100' emulators.
24695 #	  for computing string-table lengths had a bug in it.
24696 #	* pcansi series modified to fit comm-program reality better.
24702 #	* Change vt100 standout to be normal reverse vide, not bright reverse;
24703 #	  this makes the Emacs status line look better.
24706 #	* Minor surgery, mostly on name strings, to shorten termcap version.
24710 #	* Added kmous to xterm for use with xterm's mouse-tracking facility.
24713 #	* Changed xterm lines to 24, the X11R6 default.
24715 #	* Added 7 newly discovered, undocumented acsc characters to linux
24719 #	* I finally figured out how to translate ko capabilities.  Done.
24726 #	  by making them relative to use capabilities
24727 #	* Added cuf1=^L to tvi925 from deleted variant tvi925a.
24729 #	* added it#8 to entries that used to have :pt: -- tvi912, vi200,
24731 #	* Translate all home=\E[;H capabilities to home=\E[H, they're
24736 #	* Added it#8 and ht=\t to *all* entries with :pt:; the ncurses tic
24748 #	* Linux kbs is default-mapped to delete for some brain-dead reason.
24753 #	* Added some longname fields in order to avoid warning messages from
24755 #	* According to ctlseqs.ms, xterm has a full vt100 graphics set.  Use
24759 #	* Add xon to a number of console entries, they're memory-mapped and
24771 #	* Added smacs to klone entries, just as documentation.
24782 #	* Added rmam/smam capabilities to many entries based on init strings.
24783 #	* Added correct hpa/vpa to linux.
24784 #	* Reduced several entries relative to vt52.
24786 #	* Exiled some utterly unidentifiable custom and homebrew types to the
24805 #	* Added kspd=\E[P, kcbt=\E[Z, to linux entry, changed kbs back to ^H.
24806 #	* Added kent=\EOM to xterm entry.
24810 #	* Restored cuf/cud1 to xterm, their apparent bugginess was due to
24812 #	* Corrected flash strings to have a uniform delay of .2 sec.  No
24823 #	* More flash string improvements.
24825 #	* Added dim to at386.
24827 #	  he's ready to start using the termcap generated from this one.
24832 #	* Changed hazeltine name prefix from h to hz.
24833 #	* Sent t500 to the UFI file.
24837 #	* Unfreeze, linux kbs needed to be fixed.
24840 #	* Fixes to ansi and klone capabilities (thank you, Aaron Ucko).
24850 #	* Added 1.3.x color-change capabilities to linux entry.
24856 #	* Changed linux op capability to avoid screwing up a background color
24865 #	* Added dg100 alias to dg6053 based on a comp.terminals posting.
24874 #	* Change klone+color setf/setb to simpler forms that can be
24878 #	* Added color support to bsdos.
24881 #	* Created ecma+color, linux can use it.  Also added ech to linux.
24883 #	* Add descriptions to FreeBSD console entries.  Also shorten
24884 #	  some aliases to <= 14 chars for portability.
24886 #	* Added OTbs to several VT-series entries.
24890 #	* kterm correction due to Kenji Rikitake.
24891 #	* ACS correction in vt320-kll due to Phillippe De Muyter.
24894 #	* Changed all DEC VT300 and up terminals to use VT300 tab set
24896 #	* Added smul to linux entry (we never noticed it was missing
24898 #	* Added rmln to hp+labels (deduced from other HP entries).
24899 #	* Added vt100 acsc capability to vt220, vt340, vt400, d800, dt80-sas,
24903 #	  this corresponds to reality and helps prevent some tic warnings.
24904 #	* Added sgr0 to c101, pcix, vt100-nav, screen2, oldsun, next, altos2,
24918 #	* xterm entry now corresponds to XFree86 3.1.2E, with color.
24919 #	* xtitle and xtitle-twm enable access to the X status line.
24922 #	* Move fos and apollo terminals to obsolete category.
24935 #	  (old osborne moved to osborne-w), tvi970-vb, tvi970-2p, tvi925-hi,
24938 #	* Paired-attribute fixes to various terminals.
24953 #	* updated xterm entry to cover 3.1.2E's new features.
24962 #	* Added acsc/rmacs/smacs to vt52.
24965 #	  added technical corrections to avoid warning messages.
24973 #	* Back off the xterm entry to use r6 as a base.
24977 #-(original-changelog-1996/12/29-to-1998/02/28-by-TD)---------------------------
24980 #	* Minor corrections to xterm entries.
24987 #	* Thorsten Lockert added termcap `bs' to a lot of types, working from
24994 #	* update xterm-xf86-v32 to match XFree86 3.2A (changes F1-F4)
25009 #	* add cbt to xterm-xf86-v32
25023 #	* add color, mouse support to kterm.
25029 #	* add u8,u9 to sun-il description
25036 #	* rename xhpterm back to hpterm.
25038 #	* change initc in linux-c-nc to use 0..1000 range.
25047 #	* rename rxvt-color to rxvt to match rxvt 2.4.5 distribution.
25049 #	* change xterm (xterm-xf86-v40), xterm-8bit rs1 to use hard reset.
25050 #	* rename xterm-xf86-v39t to xterm-xf86-v40
25059 #	* remove sgr string from qnx based on report by Xiaodan Tang
25062 #	* correct xterm-8bit to match XFree86 3.9Ad F1-F4.
25064 #	* add linux-koi8r to replace linux-koi8 (which uses a corrupt acsc,
25072 #	* Minor corrections to xterm entries.
25076 #	* Thorsten Lockert added termcap `bs' to a lot of types, working from
25081 #	* Switch base xterm entry to 3.3 level.
25095 #	* remove sgr string from qnx based on report by Xiaodan Tang
25098 #	* add linux-koi8r to replace linux-koi8 (which uses a corrupt acsc,
25101 #	* Corrected some erroneous \\'s to \.
25102 #	* 4.2 ncurses has been changed to use setaf/setab, consistent w/SysV.
25110 #	* Reverted `amiga'; to Kent Polk's version, as I'm told
25126 #	* modify rxvt terminfo description to clear alternate screen before
25127 #	  switching back to normal screen, for compatibility with applications
25129 #	* modify linux terminfo description to reset color palette (reported
25144 #	* modify ncsa telnet entry to reflect color, other capabilities based on
25148 #	* Corrected some erroneous \\'s to \ (eterm, osborne) - TD.
25158 #	* format most %'char' sequences to %{number}
25163 #	* update xterm-xfree86 to current (xterm patch #84), for is2/rs2 changes - TD
25164 #	* correct initialization string in xterm-r5, add misc other features
25165 #	  to correspond with xterm patch #84 - TD
25168 #	* update xterm-xfree86 to current (xterm patch #90), smcur/rmcur changes - TD
25173 #	* change linux to use ncv#2, since underline does not work with color - TD
25176 #	* add kbt to iris-ansi, document other shift/control functionkeys - TD
25177 #	* correct iris-ansi and iris-ansi-ap with respect to normal vs keypad
25178 #	  application modes, change kent to use the correct keypad code - TD
25185 #	* rename several entries used by BSDI: bsdos to bsdos-pc-nobold,
25186 #	  and bsdos-bold to bsdos-pc (Jeffrey C Honig)
25195 #	* adjust xterm-xfree86 khome/kend to match default PC-style keyboard
25201 #	* update entries for BSD/OS console to use klone+sgr and klone+color
25208 #	* add linux-lat, from RedHat patches to ncurses 4.2
25217 #	* add kmous to xterm-r5 -TD
25228 #	* corrections to beterm entry -TD
25237 #	* add acsc string to HP 70092 terminfo entry -Joerg Wunsch
25250 #	* add cnorm, cvvis to cons25w, and modify ncv to add 'dim' -TD
25251 #	* reorder ncsa entries to make ncsa-vt220 use the alternate function
25259 #	* add 'dim' to ncv mask for linux (report by Klaus Weide).
25263 #	* add hts to xterm-r6, and u6-u9 to xterm-r5 -TD
25271 #	* correct cup string for regent100 -TD
25276 #	* change OTrs capabilities to rs2 -TD
25277 #	* add obsolete and extended capabilities to 'screen' -TD
25282 #	* updated xterm-xf86-v40, making kdch1 correspond to vt220 'Remove',
25303 #	* move screen's AX extension to ecma+color, modify several entries to
25312 #	* add cud, ech, etc., to beterm based on feedback from Rico Tudor -TD
25313 #	* correct color definition for ibm3164, make minor changes to other
25319 #	* change cons25w to use rs2 for reset rather than rs1 -TD
25320 #	* add rc/sc to aixterm based on manpage -TD
25326 #	* add kmous capability to linux to use Joerg Schoen's gpm patch.
25343 #	* add xterm-basic, xterm-sco entries, update related entries to XFree86
25347 #	* add S0, E0 extensions to screen's entry -TD
25356 #	* rename cygwin to cygwinB19, adapt newer entry from Earnie Boyd -TD
25363 #	* modify kterm to use acsc via SCS controls.
25366 #	* screen 3.9.8 allows xterm mouse controls to pass-through
25372 #	* modify 'screen' khome/kend to match screen 3.09.08
25387 #	* add u6 (CSR) to Eterm (Michael Jennings).
25390 #	* renamed "Apple_Terminal" entries to "nsterm" to work with Solaris's
25396 #	* change kbs in mach entries to ^? (Marcus Brinkmann).
25407 #	* update gnome entry to Redhat 7.2 -TD
25410 #	* add kf13-kf48 strings to cons25w -TD
25414 #	* add scoansi-new entry corresponding to OpenServer 5.0.6
25417 #	* add kcbt to screen entry -TD
25424 #	  in turn linux (with cnorm, etc) inherit from linux-c-nc to reflect
25426 #	* scaled the linux-c terminfo entry to match linux-c-nc, i.e., the
25427 #	  r/g/b parameters of initc are in the range 0 to 1000 -TD
25440 #	* update xterm-xfree86 to current (xterm patch #170) -TD
25457 #	  primary instead of nxterm, to match XFree86's xterm.terminfo
25458 #	  usage and to prevent circular links.
25494 #	* add sgr strings to several common entries lacking them, e.g.,
25495 #	  screen, to make the entries more portable -TD
25505 #	* modify xterm-new aka xterm-xfree86 to accommodate luit, which relies
25520 #	* add rmam/smam to linux (Trevor Van Bremen)
25521 #	* change wyse acsc strings to use 'i' map rather than 'I' -TD
25531 #	* correct tsl string in kterm -TD
25535 #	* add sgr, rc/sc and ech to syscons entries -TD
25536 #	* add function-keys to decansi -TD
25537 #	* add sgr to mterm-ansi -TD
25538 #	* add sgr, civis, cnorm to emu -TD
25542 #	* modify DEC entries (vt220, etc), to add sgr string, and to use
25546 #	* rename xterm-pc-fkeys to xterm+pcfkeys -TD
25558 #	* modify several entries to ensure xterm mouse and cursor visibility
25559 #	  are reset in rs2 string:  hurd, putty, gnome, konsole-base, mlterm,
25565 #	* modify amiga-8bit to add khome/kend/knp/kpp -Pavel Fedin
25566 #	* corrected \E[5?l to \E[?5l in vt320 entries -TD
25572 #	* update pairs for xterm-88color and xterm-256color to reflect the
25576 #	* modify sgr/sgr0 in xterm-new to improve tgetent's derived "me" -TD
25577 #	* add aixterm-16color to demonstrate 16-color capability -TD
25580 #	* add media-copy to vt100 -TD
25581 #	* corrected acsc string for vt52 -TD
25589 #	* re-corrected acsc string for vt52 -TD
25592 #	* corrected sun-il sgr string which referred to bold and underline -TD
25596 #	* modify sgr0 in several entries to reset alternate-charset as in the
25597 #	  sgr string -TD
25598 #	* modify sgr string of prism9 to better match the individual
25608 #	* other minor fixes to cygwin based on tack -TD
25614 #	* remove setf/setb from xterm-256color to match xterm #209 -TD
25615 #	* update mlterm entry to 2.9.2 -TD
25618 #	* fixes to make nsterm-16color match report
25638 #	* add some notes regarding copyright to terminfo.src -TD
25647 #	* update xterm-sun and xterm-sco entries to match xterm #216 -TD
25649 #	* modify xterm-24 to inherit from "xterm" -TD
25657 #	* update xterm+pcfkeys to match xterm #216 -TD
25666 #	* correct acsc string in kterm -TD
25671 #	  that implement the feature (or have not been shown to lack it) -TD
25674 #	* add ka2, kb1, kb3, kc2 to vt220-keypad as an extension -TD
25675 #	* minor improvements to rxvt+pcfkeys -TD
25682 #	  status line (Alain Bench).
25691 #	* restore section of pre-ncurses-4.2 changelog to fix attribution -TD
25701 #	* update mlterm entry to 2.9.3 -TD
25705 #	* move kLFT, kRIT, kind and kri capabilities from xterm-new to
25706 #	  xterm+pcc0, etc., to make the corresponding building blocks reflect
25715 #	* use xterm-xf86-v44 for "xterm-xfree86", reflecting changes to
25718 #	  to match xterm #230 -TD
25719 #	* extend xterm+pccX entries to match xterm #230 -TD
25739 #	  original to teraterm2.3 -TD
25740 #	* update "gnome" to 2.22.3 -TD
25741 #	* update "konsole" to 1.6.6 -TD
25746 #	* change several \E[2g (clear tab at current column) to \E[3g
25747 #	  (clear all tabs) to match definition for tbc capability -TD
25758 #	* move function- and cursor-keys from emx-base to ansi.sys, and create
25765 #	* change ncv and op capabilities in sun-color to match Sun's entry for
25771 #	* remove unnecessary kcan assignment to ^C from putty (Sven Joachim)
25775 #	* add ccc and initc capabilities to xterm-16color -TD
25784 #	* add bw (auto-left-margin) to nsterm* entries (Benjamin Sittler)
25785 #	* rename minix to minix-1.7, add minix entry for Minix3 -TD
25789 #	* minor fix to rxvt+pcfkeys -TD
25792 #	* update mrxvt to 0.5.4, add mrxvt-256color -TD
25798 #	* modify screen-bce.XXX entries to exclude ech, since screen's color
25802 #	* rename atari and st52 to atari-old, st52-old, use newer entries from
25809 #	* add hard-reset for rs2 to wsvt25 to help ensure that reset ends
25816 #	* add invis to tek4115 sgr -TD
25819 #	* reformat acsc strings to canonical format -TD
25822 #	* add "XT" capability to entries for terminals that support both
25831 #	* add parameterized cursor-controls to linux-basic (report by Dae) -TD
25852 #	* add U8 feature to denote entries for terminal emulators which do not
25860 #	* update minix entry to minix 3.2 (Thomas Cort).
25870 #	* add E3 extended capability to linux-basic (Miroslav Lichvar)
25871 #	* add linux2.2, linux2.6, linux3.0 entries to give context for E3 -TD
25872 #	* add SI/SO change to linux2.6 entry (Debian #515609) -TD
25875 #	* add kich1 to sun (Yuri Pankov)
25900 #	* add mach-gnu-color, tweaks to mach-gnu -TD
25925 #	* revert 2011-07-16 change to "linux" alias, return to "linux2.2" -TD
25929 #	* add XT to some places to improve usefulness for other applications
25930 #	  than screen, which would like to pretend that xterm's title is
25931 #	  a status-line. -TD
25936 #	* add msgr to vt420, similar DEC vtXXX entries -TD
25939 #	* change xterm+sl and xterm+sl-twm to include only the status-line
25946 #	* add eslok flag to dec+sl -TD
25947 #	* dec+sl applies to vt320 and up -TD
25950 #	* add ansi+tabs to vt520 -TD
25951 #	* add ansi+enq to vt220-vt520 -TD
25956 #	* modify env230 sgr/sgr0 to match other capabilities -TD
25957 #	* modify smacs/rmacs in bq300-8 to match sgr/sgr0 -TD
25972 #	* add kdch1 to wsvt25 entry from NetBSD CVS (reported by David Lord,
25974 #	* add cnorm/civis to wsvt25 entry from NetBSD CVS (report/analysis by
25976 #	* add kdch1 aka "Remove" to vt220 and vt220-8 entries -TD
25977 #	* add kdch1, etc., to qvt108 -TD
25978 #	* add dl1/il1 to some entries based on dl/il values -TD
25979 #	* add dl to simpleterm -TD
25982 #	* modify some older xterm entries to align with xterm source -TD
25986 #	* add E3 to xterm-basic and putty -TD
26003 #	* correct typo in sgr string for sun-color,
26006 #	* correct typo in sgr string for terminator -TD
26007 #	* add blink to the attributes masked by ncv in linux-16color (report
26013 #	* further improvements to linux-16color (Benjamin Sittler)
26016 #	* move nsterm-related entries out of "obsolete" section to more
26021 #	* added note to clarify Terminal.app's non-emulation of the various
26025 #	* use TS extension to describe xterm's title-escapes -TD
26026 #	* modify terminator and nsterm-s to use xterm+sl-twm building block -TD
26027 #	* update hurd.ti, add xenl to reflect 2011-03-06 change in
26037 #	* add teken entry to show actual properties of FreeBSD's "xterm"
26053 #	* updated "st" entry (leaving the 0.1.1 version as "simpleterm") to
26067 #	* update xterm-new to xterm patch #305 -TD
26068 #	+ change screen's smso to use SGR 7 (ECMA-80 reverse) rather than SGR 3
26075 #	+ revert the change to screen; see notes for why this did not work -TD
26076 #	+ cancel sitm/ritm for entries which extend "screen", to work around
26080 #	+ modify sgr for screen.xterm-new to support dim capability -TD
26081 #	+ add dim capability to nsterm+7 -TD
26083 #	+ add dim, invis capabilities to vte-2012 -TD
26084 #	+ add sitm/ritm to konsole-base and mlterm3 -TD
26101 #	+ add 'dim' capability to screen entry (report by Leonardo B Schenkel)
26102 #	+ add several key definitions to nsterm-bce to match preconfigured
26108 #	+ split-out recent change to nsterm-bce as nsterm-build326, and add
26109 #	  nsterm-build342 to reflect changes with successive releases of OSX
26111 #	+ add xon, ich1, il1 to ibm3161 (patch by Stephen Powell,
26116 #	+ add several screen.XXX entries to support the respective variations
26124 #	+ add vte-2014, update vte to use that -TD
26131 #	+ add status line to tmux via xterm+sl (patch by Nicholas Marriott).
26135 #	+ updated minitel entries to fix kel problem with emacs, and add
26142 #	+ add bold to pccon+sgr+acs and pccon-base (Tati Chevron).
26143 #	+ add keys f12-f124 to pccon+keys (Tati Chevron).
26154 #	+ add putty-noapp entry, and amend putty entry to use application mode
26159 #	+ add 'oc' capability to xterm+256color, allowing palette reset for
26163 #	+ modify linux2.6 entry to improve line-drawing -TD
26167 #	+ modify rs1 for xterm-16color, xterm-88color and xterm-256color to
26168 #	  reset palette using "oc" string as in linux -TD
26171 #	+ use ANSI reply for u8 in xterm-new, to reflect vt220-style responses
26179 #	+ updated minitel entries to use status line with screen(1), as well as
26185 #	+ correct rs2 string for vt100-nam -TD
26188 #	+ modify linux-16color to not mask dim, standout or reverse with the
26190 #	+ add 0.1sec mandatory delay to flash capabilities using the VT100
26193 #	  entry, to avoid conflict with the user-defined mapping.  The reset
26200 #	+ modify flash capability for linux and wyse entries to put the delay
26204 #	+ minor comment-fixes to help automate links to bug-urls -TD
26206 #	+ add settings corresponding to xterm-keys option to tmux entry to
26207 #	  reflect upcoming change to make that option "on" by default
26220 #	+ modify vt100 rs2 string to reset vt52 mode and scrolling regions
26225 #	+ add smso for env230, to match sgr -TD
26229 #	+ add rmxx/smxx ECMA-48 strikeout extension to tmux and xterm-basic
26233 #	+ correct missing comma-separator between string capabilities in
26236 #	+ restore rmir/smir in ansi+idc to better match original ansiterm+idc,
26240 #	+ reformatted using hexadecimal numbers to improve readability -TD
26245 #	+ add "rep" to xterm-new, available since 1997/01/26 -TD
26246 #	+ move SGR 24 and 27 from vte-2014 to vte-2012 (request by Alain
26260 #	+ add "op" to xterm+256setaf -TD
26265 #	+ modify old terminology entry and a few other terminal emulators to
26267 #	+ correct sgr string for tmux, which used screen's "standout" code
26270 #	  those correspond to the non-sgr settings where they differ, but
26281 #	+ modify xterm+256color and xterm+256setaf to use correct number of
26283 #	+ add rs1 capability to xterm-256color -TD
26284 #	+ modify xterm-r5, xterm-r6 and xterm-xf86-v32 to use xterm+kbs to
26295 #	+ add XT, hpa, indn, and vpa to screen, and invis, E3 to tmux (patch by
26306 #	+ modify iterm to use xterm+sl-twm building block -TD
26309 #	+ expanded the history section of konsole to explain its flawed
26315 #	+ update vte to vte-2017 -TD
26316 #	+ add ecma+strikeout to vte-2017 -TD
26337 #	+ add acsc string to vi200 (Nibby Nebbulous)
26338 #	  add right/down-arrow to vi200's acsc -TD
26342 #	+ add wy50 and wy60 shifted function-keys as kF1 to kF16 -TD
26343 #	+ remove ansi+rep mis-added to interix in 2018-02-23 -TD
26353 #	+ added function-key definitions to agree with Televideo 950 manual -TD
26354 #	+ add bel to tvi950 -TD
26355 #	+ add shifted function-keys to regent60 -TD
26356 #	+ renumber regent40 function-keys to match manual -TD
26357 #	+ add cd (clr_eos) to adds200 -TD
26365 #	+ add nsterm-build309 to replace nsterm-256color, assigning the latter
26366 #	  as an alias of nsterm, to make mouse work with nsterm-256color -TD
26379 #	+ add "screen5", to mention italics (report by Stefan Assmann)
26380 #	+ modify description of xterm+x11hilite to eliminate unused p5 -TD
26383 #	+ update xterm-new to xterm patch #345 -TD
26390 #	+ add rs1 to konsole, mlterm -TD
26395 #	+ comment-out some user-defined capabilities in mintty+common to allow
26408 #	+ amend the change to screen, because tmux relies upon that entry
26419 #	+ modify linux-16color to accommodate Linux console driver change in
26423 #	+ add "xterm-mono" to help packagers (report by Sven Joachim) -TD
26426 #	+ drop ich1 from rxvt-basic, Eterm and mlterm to improve compatibility