2 A Hacker's Guide to NCURSES
7 * Objective of the Package
9 + How to Design Extensions
10 * Portability and Configuration
11 * Documentation Conventions
13 * A Tour of the Ncurses Library
18 + Output and Screen Updating
19 * The Forms and Menu Libraries
20 * A Tour of the Terminfo Compiler
21 + Translation of Non-use Capabilities
22 + Use Capability Resolution
23 + Source-Form Translation
25 * Style Tips for Developers
30 This document is a hacker's tour of the ncurses library and utilities.
31 It discusses design philosophy, implementation methods, and the
32 conventions used for coding and documentation. It is recommended
33 reading for anyone who is interested in porting, extending or
34 improving the package.
36 Objective of the Package
38 The objective of the ncurses package is to provide a free software API
39 for character-cell terminals and terminal emulators with the following
42 * Source-compatible with historical curses implementations
43 (including the original BSD curses and System V curses.
44 * Conformant with the XSI Curses standard issued as part of XPG4 by
46 * High-quality -- stable and reliable code, wide portability, good
47 packaging, superior documentation.
48 * Featureful -- should eliminate as much of the drudgery of C
49 interface programming as possible, freeing programmers to think at
50 a higher level of design.
52 These objectives are in priority order. So, for example, source
53 compatibility with older version must trump featurefulness -- we
54 cannot add features if it means breaking the portion of the API
55 corresponding to historical curses versions.
59 We used System V curses as a model, reverse-engineering their API, in
60 order to fulfill the first two objectives.
62 System V curses implementations can support BSD curses programs with
63 just a recompilation, so by capturing the System V API we also capture
66 More importantly for the future, the XSI Curses standard issued by
67 X/Open is explicitly and closely modeled on System V. So conformance
68 with System V took us most of the way to base-level XSI conformance.
70 How to Design Extensions
72 The third objective (standards conformance) requires that it be easy
73 to condition source code using ncurses so that the absence of
74 nonstandard extensions does not break the code.
76 Accordingly, we have a policy of associating with each nonstandard
77 extension a feature macro, so that ncurses client code can use this
78 macro to condition in or out the code that requires the ncurses
81 For example, there is a macro NCURSES_MOUSE_VERSION which XSI Curses
82 does not define, but which is defined in the ncurses library header.
83 You can use this to condition the calls to the mouse API calls.
85 Portability and Configuration
87 Code written for ncurses may assume an ANSI-standard C compiler and
88 POSIX-compatible OS interface. It may also assume the presence of a
89 System-V-compatible select(2) call.
91 We encourage (but do not require) developers to make the code friendly
92 to less-capable UNIX environments wherever possible.
94 We encourage developers to support OS-specific optimizations and
95 methods not available under POSIX/ANSI, provided only that:
97 * All such code is properly conditioned so the build process does
98 not attempt to compile it under a plain ANSI/POSIX environment.
99 * Adding such implementation methods does not introduce
100 incompatibilities in the ncurses API between platforms.
102 We use GNU autoconf(1) as a tool to deal with portability issues. The
103 right way to leverage an OS-specific feature is to modify the autoconf
104 specification files (configure.in and aclocal.m4) to set up a new
105 feature macro, which you then use to condition your code.
107 Documentation Conventions
109 There are three kinds of documentation associated with this package.
110 Each has a different preferred format:
112 * Package-internal files (README, INSTALL, TO-DO etc.)
114 * Everything else (i.e., narrative documentation).
116 Our conventions are simple:
118 1. Maintain package-internal files in plain text. The expected viewer
119 for them more(1) or an editor window; there's no point in
121 2. Mark up manual pages in the man macros. These have to be viewable
122 through traditional man(1) programs.
123 3. Write everything else in HTML.
125 When in doubt, HTMLize a master and use lynx(1) to generate plain
126 ASCII (as we do for the announcement document).
128 The reason for choosing HTML is that it's (a) well-adapted for on-line
129 browsing through viewers that are everywhere; (b) more easily readable
130 as plain text than most other mark-ups, if you don't have a viewer;
131 and (c) carries enough information that you can generate a
132 nice-looking printed version from it. Also, of course, it make
133 exporting things like the announcement document to WWW pretty trivial.
137 The reporting address for bugs is bug-ncurses@gnu.org. This is a
138 majordomo list; to join, write to bug-ncurses-request@gnu.org with a
139 message containing the line:
140 subscribe <name>@<host.domain>
142 The ncurses code is maintained by a small group of volunteers. While
143 we try our best to fix bugs promptly, we simply don't have a lot of
144 hours to spend on elementary hand-holding. We rely on intelligent
145 cooperation from our users. If you think you have found a bug in
146 ncurses, there are some steps you can take before contacting us that
147 will help get the bug fixed quickly.
149 In order to use our bug-fixing time efficiently, we put people who
150 show us they've taken these steps at the head of our queue. This means
151 that if you don't, you'll probably end up at the tail end and have to
154 1. Develop a recipe to reproduce the bug.
155 Bugs we can reproduce are likely to be fixed very quickly, often
156 within days. The most effective single thing you can do to get a
157 quick fix is develop a way we can duplicate the bad behavior --
158 ideally, by giving us source for a small, portable test program
159 that breaks the library. (Even better is a keystroke recipe using
160 one of the test programs provided with the distribution.)
161 2. Try to reproduce the bug on a different terminal type.
162 In our experience, most of the behaviors people report as library
163 bugs are actually due to subtle problems in terminal descriptions.
164 This is especially likely to be true if you're using a traditional
165 asynchronous terminal or PC-based terminal emulator, rather than
166 xterm or a UNIX console entry.
167 It's therefore extremely helpful if you can tell us whether or not
168 your problem reproduces on other terminal types. Usually you'll
169 have both a console type and xterm available; please tell us
170 whether or not your bug reproduces on both.
171 If you have xterm available, it is also good to collect xterm
172 reports for different window sizes. This is especially true if you
173 normally use an unusual xterm window size -- a surprising number
174 of the bugs we've seen are either triggered or masked by these.
175 3. Generate and examine a trace file for the broken behavior.
176 Recompile your program with the debugging versions of the
177 libraries. Insert a trace() call with the argument set to
178 TRACE_UPDATE. (See "Writing Programs with NCURSES" for details on
179 trace levels.) Reproduce your bug, then look at the trace file to
180 see what the library was actually doing.
181 Another frequent cause of apparent bugs is application coding
182 errors that cause the wrong things to be put on the virtual
183 screen. Looking at the virtual-screen dumps in the trace file will
184 tell you immediately if this is happening, and save you from the
185 possible embarrassment of being told that the bug is in your code
186 and is your problem rather than ours.
187 If the virtual-screen dumps look correct but the bug persists,
188 it's possible to crank up the trace level to give more and more
189 information about the library's update actions and the control
190 sequences it issues to perform them. The test directory of the
191 distribution contains a tool for digesting these logs to make them
192 less tedious to wade through.
193 Often you'll find terminfo problems at this stage by noticing that
194 the escape sequences put out for various capabilities are wrong.
195 If not, you're likely to learn enough to be able to characterize
196 any bug in the screen-update logic quite exactly.
197 4. Report details and symptoms, not just interpretations.
198 If you do the preceding two steps, it is very likely that you'll
199 discover the nature of the problem yourself and be able to send us
200 a fix. This will create happy feelings all around and earn you
201 good karma for the first time you run into a bug you really can't
202 characterize and fix yourself.
203 If you're still stuck, at least you'll know what to tell us.
204 Remember, we need details. If you guess about what is safe to
205 leave out, you are too likely to be wrong.
206 If your bug produces a bad update, include a trace file. Try to
207 make the trace at the least voluminous level that pins down the
208 bug. Logs that have been through tracemunch are OK, it doesn't
209 throw away any information (actually they're better than
210 un-munched ones because they're easier to read).
211 If your bug produces a core-dump, please include a symbolic stack
212 trace generated by gdb(1) or your local equivalent.
213 Tell us about every terminal on which you've reproduced the bug --
214 and every terminal on which you can't. Ideally, sent us terminfo
215 sources for all of these (yours might differ from ours).
216 Include your ncurses version and your OS/machine type, of course!
217 You can find your ncurses version in the curses.h file.
219 If your problem smells like a logic error or in cursor movement or
220 scrolling or a bad capability, there are a couple of tiny test frames
221 for the library algorithms in the progs directory that may help you
222 isolate it. These are not part of the normal build, but do have their
223 own make productions.
225 The most important of these is mvcur, a test frame for the
226 cursor-movement optimization code. With this program, you can see
227 directly what control sequences will be emitted for any given cursor
228 movement or scroll/insert/delete operations. If you think you've got a
229 bad capability identified, you can disable it and test again. The
230 program is command-driven and has on-line help.
232 If you think the vertical-scroll optimization is broken, or just want
233 to understand how it works better, build hashmap and read the header
234 comments of hardscroll.c and hashmap.c; then try it out. You can also
235 test the hardware-scrolling optimization separately with hardscroll.
237 There's one other interactive tester, tctest, that exercises
238 translation between termcap and terminfo formats. If you have a
239 serious need to run this, you probably belong on our development team!
241 A Tour of the Ncurses Library
245 Most of the library is superstructure -- fairly trivial convenience
246 interfaces to a small set of basic functions and data structures used
247 to manipulate the virtual screen (in particular, none of this code
248 does any I/O except through calls to more fundamental modules
249 described below). The files
251 lib_addch.c lib_bkgd.c lib_box.c lib_chgat.c lib_clear.c
252 lib_clearok.c lib_clrbot.c lib_clreol.c lib_colorset.c lib_data.c
253 lib_delch.c lib_delwin.c lib_echo.c lib_erase.c lib_gen.c
254 lib_getstr.c lib_hline.c lib_immedok.c lib_inchstr.c lib_insch.c
255 lib_insdel.c lib_insstr.c lib_instr.c lib_isendwin.c lib_keyname.c
256 lib_leaveok.c lib_move.c lib_mvwin.c lib_overlay.c lib_pad.c
257 lib_printw.c lib_redrawln.c lib_scanw.c lib_screen.c lib_scroll.c
258 lib_scrollok.c lib_scrreg.c lib_set_term.c lib_slk.c
259 lib_slkatr_set.c lib_slkatrof.c lib_slkatron.c lib_slkatrset.c
260 lib_slkattr.c lib_slkclear.c lib_slkcolor.c lib_slkinit.c
261 lib_slklab.c lib_slkrefr.c lib_slkset.c lib_slktouch.c lib_touch.c
262 lib_unctrl.c lib_vline.c lib_wattroff.c lib_wattron.c lib_window.c
264 are all in this category. They are very unlikely to need change,
265 barring bugs or some fundamental reorganization in the underlying data
268 These files are used only for debugging support:
270 lib_trace.c lib_traceatr.c lib_tracebits.c lib_tracechr.c
271 lib_tracedmp.c lib_tracemse.c trace_buf.c
273 It is rather unlikely you will ever need to change these, unless you
274 want to introduce a new debug trace level for some reasoon.
276 There is another group of files that do direct I/O via tputs(),
277 computations on the terminal capabilities, or queries to the OS
278 environment, but nevertheless have only fairly low complexity. These
281 lib_acs.c lib_beep.c lib_color.c lib_endwin.c lib_initscr.c
282 lib_longname.c lib_newterm.c lib_options.c lib_termcap.c lib_ti.c
283 lib_tparm.c lib_tputs.c lib_vidattr.c read_entry.c.
285 They are likely to need revision only if ncurses is being ported to an
286 environment without an underlying terminfo capability representation.
288 These files have serious hooks into the tty driver and signal
291 lib_kernel.c lib_baudrate.c lib_raw.c lib_tstp.c lib_twait.c
293 If you run into porting snafus moving the package to another UNIX, the
294 problem is likely to be in one of these files. The file lib_print.c
295 uses sleep(2) and also falls in this category.
297 Almost all of the real work is done in the files
299 hardscroll.c hashmap.c lib_addch.c lib_doupdate.c lib_getch.c
300 lib_mouse.c lib_mvcur.c lib_refresh.c lib_setup.c lib_vidattr.c
302 Most of the algorithmic complexity in the library lives in these
303 files. If there is a real bug in ncurses itself, it's probably here.
304 We'll tour some of these files in detail below (see The Engine Room).
306 Finally, there is a group of files that is actually most of the
307 terminfo compiler. The reason this code lives in the ncurses library
308 is to support fallback to /etc/termcap. These files include
310 alloc_entry.c captoinfo.c comp_captab.c comp_error.c comp_hash.c
311 comp_parse.c comp_scan.c parse_entry.c read_termcap.c write_entry.c
313 We'll discuss these in the compiler tour.
319 All ncurses input funnels through the function wgetch(), defined in
320 lib_getch.c. This function is tricky; it has to poll for keyboard and
321 mouse events and do a running match of incoming input against the set
322 of defined special keys.
324 The central data structure in this module is a FIFO queue, used to
325 match multiple-character input sequences against special-key
326 capabilities; also to implement pushback via ungetch().
328 The wgetch() code distinguishes between function key sequences and the
329 same sequences typed manually by doing a timed wait after each input
330 character that could lead a function key sequence. If the entire
331 sequence takes less than 1 second, it is assumed to have been
332 generated by a function key press.
334 Hackers bruised by previous encounters with variant select(2) calls
335 may find the code in lib_twait.c interesting. It deals with the
336 problem that some BSD selects don't return a reliable time-left value.
337 The function timed_wait() effectively simulates a System V select.
341 If the mouse interface is active, wgetch() polls for mouse events each
342 call, before it goes to the keyboard for input. It is up to
343 lib_mouse.c how the polling is accomplished; it may vary for different
346 Under xterm, however, mouse event notifications come in via the
347 keyboard input stream. They are recognized by having the kmous
348 capability as a prefix. This is kind of klugey, but trying to wire in
349 recognition of a mouse key prefix without going through the
350 function-key machinery would be just too painful, and this turns out
351 to imply having the prefix somewhere in the function-key capabilities
352 at terminal-type initialization.
354 This kluge only works because kmous isn't actually used by any
355 historic terminal type or curses implementation we know of. Best guess
356 is it's a relic of some forgotten experiment in-house at Bell Labs
357 that didn't leave any traces in the publicly-distributed System V
358 terminfo files. If System V or XPG4 ever gets serious about using it
359 again, this kluge may have to change.
361 Here are some more details about mouse event handling:
363 The lib_mouse()code is logically split into a lower level that accepts
364 event reports in a device-dependent format and an upper level that
365 parses mouse gestures and filters events. The mediating data structure
366 is a circular queue of event structures.
368 Functionally, the lower level's job is to pick up primitive events and
369 put them on the circular queue. This can happen in one of two ways:
370 either (a) _nc_mouse_event() detects a series of incoming mouse
371 reports and queues them, or (b) code in lib_getch.c detects the kmous
372 prefix in the keyboard input stream and calls _nc_mouse_inline to
373 queue up a series of adjacent mouse reports.
375 In either case, _nc_mouse_parse() should be called after the series is
376 accepted to parse the digested mouse reports (low-level events) into a
377 gesture (a high-level or composite event).
379 Output and Screen Updating
381 With the single exception of character echoes during a wgetnstr() call
382 (which simulates cooked-mode line editing in an ncurses window), the
383 library normally does all its output at refresh time.
385 The main job is to go from the current state of the screen (as
386 represented in the curscr window structure) to the desired new state
387 (as represented in the newscr window structure), while doing as little
390 The brains of this operation are the modules hashmap.c, hardscroll.c
391 and lib_doupdate.c; the latter two use lib_mvcur.c. Essentially, what
392 happens looks like this:
394 The hashmap.c module tries to detect vertical motion changes between
395 the real and virtual screens. This information is represented by the
396 oldindex members in the newscr structure. These are modified by
397 vertical-motion and clear operations, and both are re-initialized
398 after each update. To this change-journalling information, the hashmap
399 code adds deductions made using a modified Heckel algorithm on hash
400 values generated from the line contents.
402 The hardscroll.c module computes an optimum set of scroll, insertion,
403 and deletion operations to make the indices match. It calls
404 _nc_mvcur_scrolln() in lib_mvcur.c to do those motions.
406 Then lib_doupdate.c goes to work. Its job is to do line-by-line
407 transformations of curscr lines to newscr lines. Its main tool is the
408 routine mvcur() in lib_mvcur.c. This routine does cursor-movement
409 optimization, attempting to get from given screen location A to given
410 location B in the fewest output characters posible.
412 If you want to work on screen optimizations, you should use the fact
413 that (in the trace-enabled version of the library) enabling the
414 TRACE_TIMES trace level causes a report to be emitted after each
415 screen update giving the elapsed time and a count of characters
416 emitted during the update. You can use this to tell when an update
417 optimization improves efficiency.
419 In the trace-enabled version of the library, it is also possible to
420 disable and re-enable various optimizations at runtime by tweaking the
421 variable _nc_optimize_enable. See the file include/curses.h.in for
422 mask values, near the end.
424 The Forms and Menu Libraries
426 The forms and menu libraries should work reliably in any environment
427 you can port ncurses to. The only portability issue anywhere in them
428 is what flavor of regular expressions the built-in form field type
429 TYPE_REGEXP will recognize.
431 The configuration code prefers the POSIX regex facility, modeled on
432 System V's, but will settle for BSD regexps if the former isn't
435 Historical note: the panels code was written primarily to assist in
436 porting u386mon 2.0 (comp.sources.misc v14i001-4) to systems lacking
437 panels support; u386mon 2.10 and beyond use it. This version has been
438 slightly cleaned up for ncurses.
440 A Tour of the Terminfo Compiler
442 The ncurses implementation of tic is rather complex internally; it has
443 to do a trying combination of missions. This starts with the fact
444 that, in addition to its normal duty of compiling terminfo sources
445 into loadable terminfo binaries, it has to be able to handle termcap
446 syntax and compile that too into terminfo entries.
448 The implementation therefore starts with a table-driven, dual-mode
449 lexical analyzer (in comp_scan.c). The lexer chooses its mode (termcap
450 or terminfo) based on the first `,' or `:' it finds in each entry. The
451 lexer does all the work of recognizing capability names and values;
452 the grammar above it is trivial, just "parse entries till you run out
455 Translation of Non-use Capabilities
457 Translation of most things besides use capabilities is pretty
458 straightforward. The lexical analyzer's tokenizer hands each
459 capability name to a hash function, which drives a table lookup. The
460 table entry yields an index which is used to look up the token type in
461 another table, and controls interpretation of the value.
463 One possibly interesting aspect of the implementation is the way the
464 compiler tables are initialized. All the tables are generated by
465 various awk/sed/sh scripts from a master table include/Caps; these
466 scripts actually write C initializers which are linked to the
467 compiler. Furthermore, the hash table is generated in the same way, so
468 it doesn't have to be generated at compiler startup time (another
469 benefit of this organization is that the hash table can be in
470 shareable text space).
472 Thus, adding a new capability is usually pretty trivial, just a matter
473 of adding one line to the include/Caps file. We'll have more to say
474 about this in the section on Source-Form Translation.
476 Use Capability Resolution
478 The background problem that makes tic tricky isn't the capability
479 translation itself, it's the resolution of use capabilities. Older
480 versions would not handle forward use references for this reason (that
481 is, a using terminal always had to follow its use target in the source
482 file). By doing this, they got away with a simple implementation
483 tactic; compile everything as it blows by, then resolve uses from
486 This won't do for ncurses. The problem is that that the whole
487 compilation process has to be embeddable in the ncurses library so
488 that it can be called by the startup code to translate termcap entries
489 on the fly. The embedded version can't go promiscuously writing
490 everything it translates out to disk -- for one thing, it will
491 typically be running with non-root permissions.
493 So our tic is designed to parse an entire terminfo file into a
494 doubly-linked circular list of entry structures in-core, and then do
495 use resolution in-memory before writing everything out. This design
496 has other advantages: it makes forward and back use-references equally
497 easy (so we get the latter for free), and it makes checking for name
498 collisions before they're written out easy to do.
500 And this is exactly how the embedded version works. But the
501 stand-alone user-accessible version of tic partly reverts to the
502 historical strategy; it writes to disk (not keeping in core) any entry
503 with no use references.
505 This is strictly a core-economy kluge, implemented because the
506 terminfo master file is large enough that some core-poor systems swap
507 like crazy when you compile it all in memory...there have been reports
508 of this process taking three hours, rather than the twenty seconds or
509 less typical on the author's development box.
511 So. The executable tic passes the entry-parser a hook that immediately
512 writes out the referenced entry if it has no use capabilities. The
513 compiler main loop refrains from adding the entry to the in-core list
514 when this hook fires. If some other entry later needs to reference an
515 entry that got written immediately, that's OK; the resolution code
516 will fetch it off disk when it can't find it in core.
518 Name collisions will still be detected, just not as cleanly. The
519 write_entry() code complains before overwriting an entry that
520 postdates the time of tic's first call to write_entry(), Thus it will
521 complain about overwriting entries newly made during the tic run, but
522 not about overwriting ones that predate it.
524 Source-Form Translation
526 Another use of tic is to do source translation between various termcap
527 and terminfo formats. There are more variants out there than you might
528 think; the ones we know about are described in the captoinfo(1) manual
531 The translation output code (dump_entry() in ncurses/dump_entry.c) is
532 shared with the infocmp(1) utility. It takes the same internal
533 representation used to generate the binary form and dumps it to
534 standard output in a specified format.
536 The include/Caps file has a header comment describing ways you can
537 specify source translations for nonstandard capabilities just by
538 altering the master table. It's possible to set up capability aliasing
539 or tell the compiler to plain ignore a given capability without
540 writing any C code at all.
542 For circumstances where you need to do algorithmic translation, there
543 are functions in parse_entry.c called after the parse of each entry
544 that are specifically intended to encapsulate such translations. This,
545 for example, is where the AIX box1 capability get translated to an
550 The infocmp utility is just a wrapper around the same entry-dumping
551 code used by tic for source translation. Perhaps the one interesting
552 aspect of the code is the use of a predicate function passed in to
553 dump_entry() to control which capabilities are dumped. This is
554 necessary in order to handle both the ordinary De-compilation case and
555 entry difference reporting.
557 The tput and clear utilities just do an entry load followed by a
558 tputs() of a selected capability.
560 Style Tips for Developers
562 See the TO-DO file in the top-level directory of the source
563 distribution for additions that would be particularly useful.
565 The prefix _nc_ should be used on library public functions that are
566 not part of the curses API in order to prevent pollution of the
567 application namespace. If you have to add to or modify the function
568 prototypes in curses.h.in, read ncurses/MKlib_gen.sh first so you can
569 avoid breaking XSI conformance. Please join the ncurses mailing list.
570 See the INSTALL file in the top level of the distribution for details
573 Look for the string FIXME in source files to tag minor bugs and
574 potential problems that could use fixing.
576 Don't try to auto-detect OS features in the main body of the C code.
577 That's the job of the configuration system.
579 To hold down complexity, do make your code data-driven. Especially, if
580 you can drive logic from a table filtered out of include/Caps, do it.
581 If you find you need to augment the data in that file in order to
582 generate the proper table, that's still preferable to ad-hoc code --
583 that's why the fifth field (flags) is there.
589 The following notes are intended to be a first step towards DOS and
590 Macintosh ports of the ncurses libraries.
592 The following library modules are `pure curses'; they operate only on
593 the curses internal structures, do all output through other curses
594 calls (not including tputs() and putp()) and do not call any other
595 UNIX routines such as signal(2) or the stdio library. Thus, they
596 should not need to be modified for single-terminal ports.
598 lib_addch.c lib_addstr.c lib_bkgd.c lib_box.c lib_clear.c
599 lib_clrbot.c lib_clreol.c lib_delch.c lib_delwin.c lib_erase.c
600 lib_inchstr.c lib_insch.c lib_insdel.c lib_insstr.c lib_keyname.c
601 lib_move.c lib_mvwin.c lib_newwin.c lib_overlay.c lib_pad.c
602 lib_printw.c lib_refresh.c lib_scanw.c lib_scroll.c lib_scrreg.c
603 lib_set_term.c lib_touch.c lib_tparm.c lib_tputs.c lib_unctrl.c
606 This module is pure curses, but calls outstr():
610 These modules are pure curses, except that they use tputs() and
613 lib_beep.c lib_color.c lib_endwin.c lib_options.c lib_slk.c
616 This modules assist in POSIX emulation on non-POSIX systems:
621 The following source files will not be needed for a
622 single-terminal-type port.
624 alloc_entry.c captoinfo.c clear.c comp_captab.c comp_error.c
625 comp_hash.c comp_main.c comp_parse.c comp_scan.c dump_entry.c
626 infocmp.c parse_entry.c read_entry.c tput.c write_entry.c
628 The following modules will use open()/read()/write()/close()/lseek()
629 on files, but no other OS calls.
632 used to read/write screen dumps
635 used to write trace data to the logfile
637 Modules that would have to be modified for a port start here:
639 The following modules are `pure curses' but contain assumptions
640 inappropriate for a memory-mapped port.
643 assumes there may be multiple terminals
646 assumes acs_map as a double indirection
649 assumes cursor moves have variable cost
652 assumes there may be multiple terminals
655 assumes there may be multiple terminals
657 The following modules use UNIX-specific calls:
671 various tty-manipulation and system calls
674 various tty-manipulation calls
677 various tty-manipulation calls
680 various tty-manipulation calls
683 signal-manipulation calls
686 gettimeofday(), select().
687 _________________________________________________________________
690 Eric S. Raymond <esr@snark.thyrsus.com>
692 (Note: This is not the bug address!)