3 This file describes various problems that have been encountered
4 in compiling, installing and running XEmacs. It has been updated for
7 This file is rather large, but we have tried to sort the entries by
8 their respective relevance for XEmacs, but may have not succeeded
9 completely in that task. The file is divided into four parts:
11 - Problems with building XEmacs
12 - Problems with running XEmacs
13 - Compatibility problems
16 Use `C-c C-f' to move to the next equal level of outline, and
17 `C-c C-b' to move to previous equal level. `C-h m' will give more
18 info about the Outline mode.
20 Also, Try finding the things you need using one of the search commands
21 XEmacs provides (e.g. `C-s').
25 WATCH OUT for your init file! (~/.xemacs/init.el or ~/.emacs) If
26 you observe strange problems, invoke XEmacs with the `-vanilla'
27 option and see if you can repeat the problem.
29 Note that most of the problems described here manifest at RUN
30 time, even those described as BUILD problems. It is quite unusual
31 for a released XEmacs to fail to build. So a "build problem"
32 requires you to tweak the build environment, then rebuild XEmacs.
33 A "runtime problem" is one that can be fixed by proper
34 configuration of the existing build. Compatibility problems and
35 Mule issues are generally runtime problems, but are treated
36 separately for convenience.
39 * Problems with building XEmacs
40 ===============================
44 Much general information is in INSTALL. If it's covered in
45 INSTALL, we don't repeat it here.
47 *** How do I configure to get the buffer tabs/progress bars?
49 These features depend on support for "native widgets". Use the
50 --with-widgets option to configure. Configuration of widgets is
51 automatic for "modern" toolkits (MS Windows, GTK, and Motif), but if
52 you are using Xt and the Athena widgets, you will probably want to
53 specify a "3d" widget set. See configure --usage, and don't forget to
54 install the corresponding development libraries.
56 *** I know I have libfoo installed, but configure doesn't find it.
58 Typical of Linux systems with package managers. To link with a shared
59 library, you only need the shared library. To compile objects that
60 link with it, you need the headers---and distros don't provide them with
61 the libraries. You need the additional "development" package, too.
63 *** When using gcc, you get the error message "undefined symbol __fixunsdfsi".
64 When using gcc, you get the error message "undefined symbol __main".
66 This means that you need to link with the gcc library. It may be called
67 "gcc-gnulib" or "libgcc.a"; figure out where it is, and define LIB_GCC in
68 config.h to point to it.
70 It may also work to use the GCC version of `ld' instead of the standard one.
72 *** Excessive optimization with pgcc can break XEmacs
74 It has been reported on some systems that compiling with -O6 can lead
75 to XEmacs failures. The workaround is to use a lower optimization
76 level. -O2 and -O4 have been tested extensively.
78 All of this depends heavily on the version of pgcc and the version
79 of libc. Snapshots near the release of pgcc-1.0 have been tested
80 extensively and no sign of breakage has been seen on systems using
83 *** src/Makefile and lib-src/Makefile are truncated--most of the file missing.
85 This can happen if configure uses GNU sed version 2.03. That version
86 had a bug. GNU sed version 2.05 works properly.
88 *** When compiling with X11, you get "undefined symbol _XtStrings".
90 This means that you are trying to link emacs against the X11r4 version of
91 libXt.a, but you have compiled either Emacs or the code in the lwlib
92 subdirectory with the X11r5 header files. That doesn't work.
94 Remember, you can't compile lwlib for r4 and emacs for r5, or vice versa.
97 *** test-distrib says that the distribution has been clobbered
98 or, temacs prints "Command key out of range 0-127"
99 or, temacs runs and dumps xemacs, but xemacs totally fails to work.
100 or, temacs gets errors dumping xemacs
102 This can be because the .elc files have been garbled. Do not be
103 fooled by the fact that most of a .elc file is text: these are binary
104 files and can contain all 256 byte values.
106 In particular `shar' cannot be used for transmitting GNU Emacs. It
107 typically truncates "lines". (this does not apply to GNU shar, which
108 uses uuencode to encode binary files.)
110 If you have a copy of Emacs that has been damaged in its nonprinting
111 characters, you can fix them by running:
115 This will rebuild all the needed .elc files.
117 ** Intel Architecture General
119 *** Don't use -O2 or -O3 with Cygwin 1.0, CodeFusion-99070 or gcc 2.7.2 on x86
120 without also using `-fno-strength-reduce'.
122 gcc will generate incorrect code otherwise. This bug is present in at
123 least 2.6.x and 2.7.[0-2]. This bug has been fixed in GCC 2.7.2.1 and
124 later. This bug is O/S independent, but is limited to x86 architectures.
126 This problem is known to be fixed in egcs (or pgcc) 1.0 or later.
128 Unfortunately, later releases of Cygnus-released compilers (not the
129 Net-released ones) have a bug with the same `problem signature'.
131 If you're lucky, you'll get an error while compiling that looks like:
133 event-stream.c:3189: internal error--unrecognizable insn:
134 (insn 256 14 15 (set (reg/v:SI 24)
135 (minus:SI (reg/v:SI 25)
136 (const_int 2))) -1 (insn_list 11 (nil))
140 If you're unlucky, your code will simply execute incorrectly.
142 *** Don't use -O2 with gcc 2.7.2 under Intel architectures without also
143 using `-fno-caller-saves'.
145 gcc will generate incorrect code otherwise. This bug is still
146 present in gcc 2.7.2.3. There have been no reports to indicate the
147 bug is present in egcs 1.0 (or pgcc 1.0) or later. This bug is O/S
148 independent, but limited to x86 architectures.
150 This problem is known to be fixed in egcs (or pgcc) 1.0 or later.
152 *** `compress' and `uncompress' not found and XFree86
154 XFree86 installs a very old version of libz.a by default ahead of where
155 more modern version of libz might be installed. This will cause problems
156 when attempting to link against libMagick. The fix is to remove the old
157 libz.a in the X11 binary directory.
162 Motif is the X11 version of the Gnus torture test: if there's a way to
163 crash, Motif will find it. With the open source release of Motif, it
164 seems like a good idea to collect all Motif-related issues in one
167 You should also look in your OS's section, as it may not be Motif's
170 *** XEmacs visibly repaints itty-bitty rectangles very slowly.
172 This should only be visible on a slow X connection (ISDN, maybe T1).
174 At least some versions of Motif apparently do not implement
175 XtExposeCompressMaximal properly, so it is disabled. If you wish to
176 experiment, you can remove the #ifdef LWLIB_NEEDS_MOTIF at line 238
177 (or so) of src/EmacsFrame.c, leaving only the line
179 /* compress_exposure */ XtExposeCompressMaximal | XtExposeNoRegion,
181 and recompile. This enables exposure compression, giving a 10:1 or
182 better speedup for some users. However, on some Motif platforms (Red
183 Hat Linux 9.0 and Solaris 2.8, at least), this causes XEmacs to hang
184 while displaying the progress bar (eg, in font-lock). A workaround
185 for that problem is to setq `progress-feedback-use-echo-area' to `t'.
187 *** XEmacs crashes on exit (#1).
189 The backtrace is something like:
192 #0 0xfeb9a480 in _libc_kill () from /usr/lib/libc.so.1
193 #1 0x000b0388 in fatal_error_signal ()
194 #2 <signal handler called>
195 #3 YowIter (ht=0xb, id=0x0, v=0x74682074, client=0x47e3c0)
197 #4 0xff26cc5c in _LTHashTableForEachItem (ht=0x4725e8,
198 iter=0xff26dda0 <YowIter>, ClientData=0x47e3c0) at Hash.c:671
199 #5 0xff2a4664 in destroy (w=0x496550) at Screen.c:352
200 #6 0xfef92118 in Phase2Destroy () from /usr/openwin/lib/libXt.so.4
201 #7 0xfef91940 in Recursive () from /usr/openwin/lib/libXt.so.4
202 #8 0xfef91e44 in XtPhase2Destroy () from /usr/openwin/lib/libXt.so.4
203 #9 0xfef91ae8 in _XtDoPhase2Destroy () from /usr/openwin/lib/libXt.so.4
204 #10 0xfef918cc in XtDestroyWidget () from /usr/openwin/lib/libXt.so.4
205 #11 0xfef91438 in CloseDisplay () from /usr/openwin/lib/libXt.so.4
206 #12 0xfef91394 in XtCloseDisplay () from /usr/openwin/lib/libXt.so.4
207 #13 0x0025b8b0 in x_delete_device ()
208 #14 0x000940b0 in delete_device_internal ()
209 #15 0x000806a0 in delete_console_internal ()
211 This is known to happen with Lesstif version 0.93.36. Similar
212 backtraces have also been observed on HP/UX and Solaris. There is a
213 patch for Lesstif. (This is not a solution; it just stops the crash.
214 It may or may not be harmless, but "it works for the author".)
216 Note that this backtrace looks a lot like the one in the next item.
217 However, this one is invulnerable to the Solaris patches mentioned there.
219 Frank McIngvale <frankm@hiwaay.net> says:
221 Ok, 0.93.34 works, and I tracked down the crash to a section
222 marked "experimental" in 0.93.36. Patch attached, "works for me".
224 diff -u -r lesstif-0.93.36/lib/Xm/ImageCache.c lesstif-0.93.36-mod/lib/Xm/ImageCache.c
225 --- lesstif-0.93.36/lib/Xm/ImageCache.c 2002-08-05 14:53:24.000000000 -0500
226 +++ lesstif-0.93.36-mod/lib/Xm/ImageCache.c 2002-11-11 11:13:12.000000000 -0600
227 @@ -1166,5 +1166,4 @@
228 DEBUGOUT(_LtDebug0(__FILE__, NULL, "_LtImageCacheScreenDestroy (XmGetPixmapByDepth) %p\n",
231 - (void) _LTHashTableForEachItem(PixmapCache, YowIter, (XtPointer)s);
234 *** XEmacs crashes on exit (#2)
236 Especially frequent with multiple frames. Crashes that produce C
237 backtraces like this:
239 #0 0xfec9a118 in _libc_kill () from /usr/lib/libc.so.1
240 #1 0x77f48 in fatal_error_signal (sig=11)
241 at /codes/rpluim/xemacs-21.4/src/emacs.c:539
242 #2 <signal handler called>
243 #3 0xfee929f4 in XFindContext () from /usr/openwin/lib/libX11.so.4
244 #4 0xfee92930 in XFindContext () from /usr/openwin/lib/libX11.so.4
245 #5 0xff297e54 in DisplayDestroy () from /usr/dt/lib/libXm.so.4
246 #6 0xfefbece0 in XtCallCallbackList () from /usr/openwin/lib/libXt.so.4
247 #7 0xfefc486c in XtPhase2Destroy () from /usr/openwin/lib/libXt.so.4
248 #8 0xfefc45d0 in _XtDoPhase2Destroy () from /usr/openwin/lib/libXt.so.4
249 #9 0xfefc43b4 in XtDestroyWidget () from /usr/openwin/lib/libXt.so.4
250 #10 0x15cf9c in x_delete_device (d=0x523f00)
252 are caused by buggy Motif libraries. Installing the following patches
253 has been reported to solve the problem on Solaris 2.7:
257 For information (although they have not been confirmed to work), the
258 equivalent patches for Solaris 2.8 are:
262 *** On HP-UX 11.0 XEmacs causes excessive X11 errors when running.
263 (also appears on AIX as reported in comp.emacs.xemacs)
265 Marcus Thiessel <marcus@xemacs.org>
267 Unfortunately, XEmacs releases prior to 21.0 don't work with
268 Motif2.1. It will compile but you will get excessive X11 errors like
270 xemacs: X Error of failed request: BadGC (invalid GC parameter)
272 and finally XEmacs gets killed. A workaround is to use the
273 Motif1.2_R6 libraries. You can the following line to your call to
276 --x-libraries="/usr/lib/Motif1.2_R6 -L/usr/lib/X11R6"
278 Make sure /usr/lib/Motif1.2_R6/libXm.sl is a link to
279 /usr/lib/Motif1.2_R6/libXm.3.
281 *** On HP-UX 11.0: Object "" does not have windowed ancestor
283 Marcus Thiessel <marcus@xemacs.org>
285 XEmacs dies without core file and reports:
287 Error: Object "" does not have windowed ancestor.
289 This is a bug. Please apply the patch PHSS_19964 (check if
290 superseded). The other alternative is to link with Motif1.2_R6 (see
293 *** Motif dialog boxes lose on Irix.
295 Larry Auton <lda@control.att.com> writes:
296 Beware of not specifying
298 --with-dialogs=athena
300 if it builds with the motif dialogs [boom!] you're a dead man.
304 *** IBM compiler fails: "The character # is not a valid C source character."
306 Most recently observed in 21.5.9, due to USE_KKCC ifdefs (they just
307 happen to tickle the implementation).
309 Valdis Kletnieks says:
311 The problem is that IBM defines a *MACRO* called 'memcpy', and we
312 have stuck a #ifdef/#endif inside the macro call. As a workaround,
313 try adding '-U__STR__' to your CFLAGS - this will cause string.h to
314 not do a #define for strcpy() to __strcpy() - it uses this for
315 automatic inlining support.
317 (For the record, the same issue affects a number of other functions
318 defined in string.h - basically anything the compiler knows how to
321 *** On AIX 4.3, you must specify --with-dialogs=athena with configure
323 *** The libXt shipped with AIX 4.3 up to 4.3.2 is broken. This causes
324 xemacs -nw to fail in various ways. The official APAR is this:
326 APAR NUMBER: <IX89470> RESOLVED AS: PROGRAM ERROR
329 <IX89470>: LIBXT.A INCORRECT HANDLING OF EXCEPTIONS IN XTAPPADDINPUT
331 The solution is to install X11.base.lib at version >=4.3.2.5.
333 *** On AIX, you get this compiler error message:
335 Processing include file ./XMenuInt.h
336 1501-106: (S) Include file X11/Xlib.h not found.
338 This means your system was installed with only the X11 runtime i.d
339 libraries. You have to find your sipo (bootable tape) and install
342 *** On AIX 4.1.2, linker error messages such as
343 ld: 0711-212 SEVERE ERROR: Symbol .__quous, found in the global symbol table
344 of archive /usr/lib/libIM.a, was not defined in archive member shr.o.
346 This is a problem in libIM.a. You can work around it by executing
347 these shell commands in the src subdirectory of the directory where
350 cp /usr/lib/libIM.a .
354 Then change -lIM to ./libIM.a in the command to link temacs (in
357 *** Excessive optimization on AIX 4.2 can lead to compiler failure.
359 Valdis.Kletnieks@vt.edu writes:
360 At least at the b34 level, and the latest-and-greatest IBM xlc
361 (3.1.4.4), there are problems with -O3. I haven't investigated
366 *** Don't use -O2 with gcc 2.8.1 and egcs 1.0 under SPARC architectures
367 without also using `-fno-schedule-insns'.
369 gcc will generate incorrect code otherwise, typically resulting in
370 crashes in the function skip-syntax-backward.
372 *** Don't use gcc-2.95.2 with -mcpu=ultrasparc on Solaris 2.6.
374 gcc will assume a 64-bit operating system, even though you've
375 merely told it to assume a 64-bit instruction set.
377 *** Dumping error when using GNU binutils / GNU ld on a Sun.
379 Errors similar to the following:
381 Dumping under the name xemacs unexec():
382 dldump(/space/rpluim/xemacs-obj/src/xemacs): ld.so.1: ./temacs:
383 fatal: /space/rpluim/xemacs-obj/src/xemacs: unknown dynamic entry:
386 are caused by using GNU ld. There are several workarounds available:
388 In XEmacs 21.2 or later, configure using the new portable dumper
391 Alternatively, you can link using the Sun version of ld, which is
392 normally held in /usr/ccs/bin. This can be done by one of:
394 - building gcc with these configure flags:
395 configure --with-ld=/usr/ccs/bin/ld --with-as=/usr/ccs/bin/as
397 - adding -B/usr/ccs/bin/ to CFLAGS used to configure XEmacs
398 (Note: The trailing '/' there is significant.)
400 - uninstalling GNU ld.
402 The Solaris2 FAQ claims:
404 When you install gcc, don't make the mistake of installing
405 GNU binutils or GNU libc, they are not as capable as their
406 counterparts you get with Solaris 2.x.
408 *** Link failure when using acc on a Sun.
410 To use acc, you need additional options just before the libraries, such as
412 /usr/lang/SC2.0.1/values-Xt.o -L/usr/lang/SC2.0.1/cg87 -L/usr/lang/SC2.0.1
414 and you need to add -lansi just before -lc.
416 The precise file names depend on the compiler version, so we
417 cannot easily arrange to supply them.
419 *** Problems finding X11 libraries on Solaris with Openwindows
421 Some users have reported problems in this area. The reported solution
422 is to define the environment variable OPENWINHOME, even if you must set
423 it to `/usr/openwin'.
425 *** Sed problems on Solaris 2.5
427 There have been reports of Sun sed truncating very lines in the
428 Makefile during configuration. The workaround is to use GNU sed or,
429 even better, think of a better way to generate Makefile, and send us a
432 *** On Solaris 2 I get undefined symbols from libcurses.a.
434 You probably have /usr/ucblib/ on your LD_LIBRARY_PATH. Do the link with
435 LD_LIBRARY_PATH unset. Generally, avoid using any ucb* stuff when
438 *** On Solaris 2 I cannot make alloc.o, glyphs.o or process.o.
440 The SparcWorks C compiler may have difficulty building those modules
441 with optimization level -xO4. Try using only "-fast" optimization
442 for just those modules. (Or use gcc).
444 *** Solaris 2.3 /bin/sh coredumps during configuration.
446 This only occurs if you have LANG != C. This is a known bug with
447 /bin/sh fixed by installing Patch-ID# 101613-01. Or, you can use
448 bash by setting the environment variable CONFIG_SHELL to /bin/bash
450 *** Solaris 2.x configure fails: ./config.status: test: argument expected
452 This is a known bug with /bin/sh and /bin/test, i.e. they do not
453 support the XPG4 standard. You can use bash as a workaround or an
454 XPG4-compliant Bourne shell such as the Sun-supplied /usr/xpg4/bin/sh
455 by setting the environment variable CONFIG_SHELL to /usr/xpg4/bin/sh
457 *** On SunOS, you get linker errors
459 _get_wmShellWidgetClass
460 _get_applicationShellWidgetClass
462 The fix to this is to install patch 100573 for OpenWindows 3.0
463 or link libXmu statically.
465 *** On Sunos 4, you get the error ld: Undefined symbol __lib_version.
467 This is the result of using cc or gcc with the shared library meant
468 for acc (the Sunpro compiler). Check your LD_LIBRARY_PATH and delete
469 /usr/lang/SC2.0.1 or some similar directory.
471 *** Undefined symbols when linking on Sunos 4.1.
473 If you get the undefined symbols _atowc _wcslen, _iswprint, _iswspace,
474 _iswcntrl, _wcscpy, and _wcsncpy, then you need to add -lXwchar after
475 -lXaw in the command that links temacs.
477 This problem seems to arise only when the international language
478 extensions to X11R5 are installed.
480 *** On a Sun running SunOS 4.1.1, you get this error message from GNU ld:
482 /lib/libc.a(_Q_sub.o): Undefined symbol __Q_get_rp_rd referenced from text segment
484 The problem is in the Sun shared C library, not in GNU ld.
486 The solution is to install Patch-ID# 100267-03 from Sun.
488 *** SunOS 4.1.2: undefined symbol _get_wmShellWidgetClass
490 Apparently the version of libXmu.so.a that Sun ships is hosed: it's missing
491 some stuff that is in libXmu.a (the static version). Sun has a patch for
492 this, but a workaround is to use the static version of libXmu, by changing
493 the link command from "-lXmu" to "-Bstatic -lXmu -Bdynamic". If you have
494 OpenWindows 3.0, ask Sun for these patches:
495 100512-02 4.1.x OpenWindows 3.0 libXt Jumbo patch
496 100573-03 4.1.x OpenWindows 3.0 undefined symbols with shared libXmu
498 *** Random other SunOS 4.1.[12] link errors.
500 The X headers and libraries that Sun ships in /usr/{include,lib}/X11 are
501 broken. Use the ones in /usr/openwin/{include,lib} instead.
505 See also Intel Architecture General, above.
507 *** egcs-1.1 on Alpha Linux
509 There have been reports of egcs-1.1 not compiling XEmacs correctly on
510 Alpha Linux. There have also been reports that egcs-1.0.3a is O.K.
512 *** Under Linux, you get "too many arguments to function `getpgrp'".
514 You have probably installed LessTiff under `/usr/local' and `libXm.so'
515 could not be found when linking `getpgrp()' test program, making XEmacs
516 think that `getpgrp()' takes an argument. Try adding `/usr/local/lib'
517 in `/etc/ld.so.conf' and run `ldconfig'. Then run XEmacs's `configure'
518 again. As with all problems of this type, reading the config.log file
519 generated from configure and seeing the log of how the test failed can
522 *** `Error: No ExtNode to pop!' on Linux systems with Lesstif.
524 This error message has been observed with lesstif-0.75a. It does not
525 appear to cause any harm.
527 *** xemacs: can't resolve symbol '__malloc_hook'
529 This is a Linux problem where you've compiled the XEmacs binary on a libc
530 5.4 with version higher than 5.4.19 and attempted to run the binary against
531 an earlier version. The solution is to upgrade your old library.
535 *** More coredumping in Irix (6.5 known to be vulnerable)
537 No fix is known yet. Here's the best information we have:
539 Valdis Kletnieks <Valdis.Kletnieks@vt.edu> writes:
541 Were xemacs and [any 3rd party, locally-compiled] libraries [you use]
542 all compiled with the same ABI ( -o32, -n32, -64) and
543 mips2/mips3/mips4 flags, and are they appropriate for the machine in
544 question? I know the IP30 implies an Octane, so it should be an R10K
545 chipset and above such nonsense, but I've seen the most astoundingly
546 bizzare crashes when somebody managed to compile with -mips4 and get
547 it to run on an R4400 or R5K system. ;)
549 Also, since you're using gcc, try re-running fixincludes and *then*
550 rebuilding xemacs and [any] libraries - mismatched headers can do that
551 sort of thing to you with little or no clue what's wrong (often you
552 get screwed when one routine does an malloc(sizeof(foo_struct)) and
553 passes the result to something that things foo_struct is a bit bigger,
556 Here's typical crash backtrace. With --pdump, this occurs usually at
557 startup under X windows and xemacs -nw at least starts, while without
558 --pdump a similar crash is observed during build.
560 #0 0x0fa460b8 in kill () at regcomp.c:637
561 637 regcomp.c: No such file or directory.
564 #0 0x0fa460b8 in kill () at regcomp.c:637
565 #1 0x10087f34 in fatal_error_signal ()
568 This is confusing because there is no such file in the XEmacs
569 distribution. This is seen on (at least) the following configurations:
571 uname -a: IRIX64 oct202 6.5 01091821 IP30
572 XEmacs 21.4.9 "Informed Management" configured for `mips-sgi-irix6.5'.
573 XEmacs 21.5-b9 "brussels sprouts" configured for `mips-sgi-irix6.5'.
575 *** On Irix 6.5, the MIPSpro compiler gets an internal compiler error
577 The MIPSpro Compiler (at least version 7.2.1) can't seem to handle the
578 union type properly, and fails to compile src/glyphs.c. To avoid this
579 problem, always build ---use-union-type=no (but that's the default, so
580 you should only see this problem if you're an XEmacs maintainer).
582 *** Linking with -rpath on IRIX.
584 Darrell Kindred <dkindred@cmu.edu> writes:
585 There are a couple of problems [with use of -rpath with Irix ld], though:
587 1. The ld in IRIX 5.3 ignores all but the last -rpath
588 spec, so the patched configure spits out a warning
589 if --x-libraries or --site-runtime-libraries are
590 specified under irix 5.x, and it only adds -rpath
591 entries for the --site-runtime-libraries. This bug was
592 fixed sometime between 5.3 and 6.2.
594 2. IRIX gcc 2.7.2 doesn't accept -rpath directly, so
595 it would have to be prefixed by -Xlinker or "-Wl,".
596 This would be fine, except that configure compiles with
597 ${CC-cc} $CFLAGS $LDFLAGS ...
598 rather than quoting $LDFLAGS with prefix-args, like
599 src/Makefile does. So if you specify --x-libraries
600 or --site-runtime-libraries, you must use --use-gcc=no,
601 or configure will fail.
603 *** On Irix 6.3, the SGI ld quits with segmentation fault when linking temacs
605 This occurs if you use the SGI linker version 7.1. Installing the
606 patch SG0001872 fixes this problem.
608 *** On Irix 6.0, make tries (and fails) to build a program named unexelfsgi
610 A compiler bug inserts spaces into the string "unexelfsgi . o"
611 in src/Makefile. Edit src/Makefile, after configure is run,
612 find that string, and take out the spaces.
614 Compiler fixes in Irix 6.0.1 should eliminate this problem.
616 *** On Irix 5.2, unexelfsgi.c can't find cmplrs/stsupport.h.
618 The file cmplrs/stsupport.h was included in the wrong file set in the
619 Irix 5.2 distribution. You can find it in the optional fileset
620 compiler_dev, or copy it from some other Irix 5.2 system. A kludgy
621 workaround is to change unexelfsgi.c to include sym.h instead of
624 *** Coredumping in Irix 6.2
626 Pete Forman <gsez020@compo.bedford.waii.com> writes:
627 A problem noted by myself and others (I've lost the references) was
628 that XEmacs coredumped when the cut or copy toolbar buttons were
629 pressed. This has been fixed by loading the SGI patchset (Feb 98)
630 without having to recompile XEmacs.
632 My versions are XEmacs 20.3 (problem first noted in 19.15) and IRIX
633 6.2, compiled using -n32. I'd guess that the relevant individual
634 patch was "SG0002580: multiple fixes for X libraries". SGI recommends
635 that the complete patch set be installed rather than parts of it.
637 ** Digital UNIX/OSF/VMS
638 *** On Digital UNIX, the DEC C compiler might have a problem compiling
641 In particular, src/extents.c and src/faces.c might cause the DEC C
642 compiler to abort. When this happens: cd src, compile the files by
643 hand, cd .., and redo the "make" command. When recompiling the files by
644 hand, use the old C compiler for the following versions of Digital UNIX:
645 - V3.n: Remove "-migrate" from the compile command.
646 - V4.n: Add "-oldc" to the compile command.
648 A related compiler bug has been fixed by the DEC compiler team. The
649 new versions of the compiler should run fine.
651 *** Under some versions of OSF XEmacs runs fine if built without
652 optimization but will crash randomly if built with optimization.
654 Using 'cc -g' is not sufficient to eliminate all optimization. Try
657 *** Compilation errors on VMS.
659 Sorry, XEmacs does not work under VMS. You might consider working on
660 the port if you really want to have XEmacs work under VMS.
663 *** On HPUX, the HP C compiler might have a problem compiling some files
666 Richard Cognot <cognot@ensg.u-nancy.fr> writes:
668 Had to drop once again to level 2 optimization, at least to
669 compile lstream.c. Otherwise, I get a "variable is void: \if"
670 problem while dumping (this is a problem I already reported
671 with vanilla hpux 10.01 and 9.07, which went away after
672 applying patches for the C compiler). Trouble is I still
673 haven't found the same patch for hpux 10.10, and I don't
674 remember the patch numbers. I think potential XEmacs builders
675 on HP should be warned about this.
677 *** I don't have `xmkmf' and `imake' on my HP.
679 You can get these standard X tools by anonymous FTP to
680 hpcvaaz.cv.hp.com. Essentially all X programs need these.
682 *** On HP-UX, problems with make
684 Marcus Thiessel <marcus@xemacs.org>
686 Some releases of XEmacs (e.g. 20.4) require GNU make to build
687 successfully. You don't need GNU make when building 21.x.
689 *** On HP-UX 9.05 XEmacs won't compile or coredump during the build.
691 Marcus Thiessel <marcus@xemacs.org>
693 This might be a sed problem. For your own safety make sure to use
694 GNU sed while dumping XEmacs.
698 *** Native cc on SCO OpenServer 5 is now OK. Icc may still throw you
699 a curve. Here is what Robert Lipe <robertl@arnet.com> says:
701 Unlike XEmacs 19.13, building with the native cc on SCO OpenServer 5
702 now produces a functional binary. I will typically build this
703 configuration for COFF with:
705 /path_to_xemacs_source/configure --with-gcc=no \
706 --site-includes=/usr/local/include --site-libraries=/usr/local/lib \
707 --with-xpm --with-xface --with-sound=nas
709 This version now supports ELF builds. I highly recommend this to
710 reduce the in-core footprint of XEmacs. This is now how I compile
711 all my test releases. Build it like this:
713 /path_to_XEmacs_source/configure --with-gcc=no \
714 --site-includes=/usr/local/include --site-libraries=/usr/local/lib \
715 --with-xpm --with-xface --with-sound=nas --dynamic
717 The compiler known as icc [ supplied with the OpenServer 5 Development
718 System ] generates a working binary, but it takes forever to generate
719 XEmacs. ICC also whines more about the code than /bin/cc does. I do
720 believe all its whining is legitimate, however. Note that you do
721 have to 'cd src ; make LD=icc' to avoid linker errors.
723 The way I handle the build procedure is:
725 /path_to_XEmacs_source/configure --with-gcc=no \
726 --site-includes=/usr/local/include --site-libraries=/usr/local/lib \
727 --with-xpm --with-xface --with-sound=nas --dynamic --compiler="icc"
729 NOTE I have the xpm, xface, and audio libraries and includes in
730 /usr/local/lib, /usr/local/include. If you don't have these,
731 don't include the "--with-*" arguments in any of my examples.
733 In previous versions of XEmacs, you had to override the defaults while
734 compiling font-lock.o and extents.o when building with icc. This seems
735 to no longer be true, but I'm including this old information in case it
736 resurfaces. The process I used was:
739 [ procure pizza, beer, repeat ]
741 make CC="icc -W0,-mP1COPT_max_tree_size=3000" font-lock.o extents.o
744 If you want sound support, get the tls566 supplement from
745 ftp.sco.com:/TLS or any of its mirrors. It works just groovy
748 The M-x manual-entry is known not to work. If you know Lisp and would
749 like help in making it work, e-mail me at <robertl@dgii.com>.
750 (UNCHECKED for 19.15 -- it might work).
752 In earlier releases, gnuserv/gnuclient/gnudoit would open a frame
753 just fine, but the client would lock up and the server would
754 terminate when you used C-x # to close the frame. This is now
757 In etc/ there are two files of note. emacskeys.sco and emacsstrs.sco.
758 The comments at the top of emacskeys.sco describe its function, and
759 the emacstrs.sco is a suitable candidate for /usr/lib/keyboard/strings
760 to take advantage of the keyboard map in emacskeys.sco.
762 Note: Much of the above entry is probably not valid for XEmacs 21.0
767 *** XEmacs complains "No such file or directory, diff"
769 or "ispell" or other commands that seem related to whatever you just
772 There are a large number of common (in the sense that "everyone has
773 these, really") Unix utilities that are not provided with XEmacs. The
774 GNU Project's implementations are available for Windows in the the
775 Cygwin distribution (http://www.cygwin.com/), which also provides a
776 complete Unix emulation environment (and thus makes ports of Unix
777 utilities nearly trivial). Another implementation is that from MinGW
778 (http://www.mingw.org/msys.shtml).
780 *** Weird crashes in pdump load or shortly after pdump load.
782 This can happen with incremental linking. Check if you have set
783 SUPPORT_EDIT_AND_CONTINUE to non-zero in config.inc, which must allow
784 incremental linking to be enabled (otherwise it's disabled). Either turn
785 this off, execute `nmake -f xemacs.mak clean', or manually remove
786 `temacs.exe' and `xemacs.exe'.
790 See also Intel Architecture General, above.
792 *** Signal 11 when building or running a dumped XEmacs.
794 This appears to happen when using the traditional dumping mechanism and
795 the system malloc. Andy Piper writes:
797 Traditional dumping on Cygwin relies on using gmalloc (there are specific
798 hacks in our version of gmalloc to support this), I suspect using sysmalloc
801 Try configuring with pdump or without system malloc.
803 *** In general use etc/check_cygwin_setup.sh to trap environment problems.
805 The script etc/check_cygwin_setup.sh will attempt to detect whether
806 you have a suitable environment for building. This script may not work
807 correctly if you are using ash instead of bash (see below).
809 *** Syntax errors running configure scripts, make failing with exit code 127
810 in inexplicable situations, etc.
812 [[ This may be because you are using the default Cygwin shell, under old
813 versions of Cygwin. The default Cygwin shell (/bin/sh.exe) is ash, which
814 appears to work in most circumstances but has some weird failure modes.
815 You may need to replace the symlink with bash.exe. ]] This doesn't appear
816 to affect Cygwin any longer, and /bin/sh.exe is no longer a symlink in
819 *** Lots of compile errors, esp. on lines containing macro definitions
820 terminated by backslashes.
822 Your partition holding the source files is mounted binary. It needs
823 to be mounted text. (This will not screw up any binary files because
824 the Cygwin utilities specify explicitly whether they want binary or
825 text mode when working with source vs. binary files, which overrides
826 the mount type.) To fix this, you just need to run the appropriate
827 mount command once -- afterwards, the settings are remembered in the
830 *** Errors from make like /c:not found.
832 Make sure you set the environment variable MAKE_MODE to UNIX in your
833 .bashrc, Control Panel (Windows 2000/NT), or AUTOEXEC.BAT (Windows
836 *** The info files will not build.
838 makeinfo that ships with old versions of Cygwin doesn't work.
839 Upgrade to the latest Cygwin version.
841 *** XEmacs hangs while attempting to rebuild the .elc files.
843 Check to make sure you're not configuring with rel-alloc. The relocating
844 allocator does not currently work under Cygwin due to bugs in Cygwin's
847 *** Trying to build with X, but X11 not detected.
849 This is usually because xmkmf is not in your path or because you are
850 using the default Cygwin shell. (See above.)
853 * Problems with running XEmacs
854 ==============================
857 *** XEmacs consistently crashes in a particular strange place.
859 One known case is on Red Hat Linux, compiled with GCC, attempting to
860 render PNG images. The problem is that XEmacs code is not compliant
861 with ANSI rules about aliasing. Adding -fno-strict-aliasing to CFLAGS
862 may help (or the equivalent for your compiler). (Some versions of
863 XEmacs may already do this automatically, but if you specify CFLAGS or
864 --cflags yourself, you will have to add this flag by hand.)
866 If you diagnose this bug for some other symptoms or systems, please
867 let us know (if you can send mail from the affected system, use M-x
868 report-xemacs-bug) so we can update this entry.
870 *** Changes made to .el files do not take effect.
872 You may have forgotten to recompile them into .elc files. Then the
873 old .elc files will be loaded, and your changes will not be seen. To
874 fix this, do `M-x byte-recompile-directory' and specify the directory
875 that contains the Lisp files.
877 Note that you will get a warning when loading a .elc file that is
878 older than the corresponding .el file.
880 *** VM appears to hang in large folders.
882 This is normal (trust us) when upgrading to VM-6.22 from earlier
883 versions. Let VM finish what it is doing and all will be well.
885 *** Starting with 21.4.x, killing text is absurdly slow.
887 See FAQ Q3.10.6. Should be available on the web near
888 http://www.xemacs.org/faq/xemacs-faq.html#SEC160.
890 *** Whenever I try to retrieve a remote file, I have problems.
892 A typical error: FTP Error: USER request failed; 500 AUTH not understood.
893 Thanks to giacomo boffi <giacomo.boffi@polimi.it> on comp.emacs.xemacs:
895 tell your ftp client to not attempt AUTH authentication (or do not
896 use FTP servers that don't understand AUTH)
898 and notes that you need to add an element (often "-u") to
899 `efs-ftp-program-args'. Use M-x customize-variable, and verify the
900 needed flag with `man ftp' or other local documentation.
902 *** gnuserv is running, some clients can connect, but others cannot.
904 The code in gnuslib.c respects the value of TMPDIR. If the server and
905 the client have different values in their environment, you lose.
906 One program known to set TMPDIR and manifest this problem is exmh.
907 You can defeat the use of TMPDIR by unsetting USE_TMPDIR at the top of
908 gnuserv.h at build time.
912 *** You type Control-H (Backspace) expecting to delete characters.
914 Emacs has traditionally used Control-H for help; unfortunately this
915 interferes with its use as Backspace on TTY's. As of XEmacs 21,
916 XEmacs looks at the "erase" setting of TTY structures and maps C-h to
917 backspace when erase is set to C-h. This is sort of a special hack,
918 but it makes it possible for you to use the standard:
922 to get your backspace key to erase characters. The erase setting is
923 recorded in the Lisp variable `tty-erase-char', which you can use to
924 tune the settings in your .emacs.
926 A major drawback of this is that when C-h becomes backspace, it no
927 longer invokes help. In that case, you need to use f1 for help, or
928 bind another key. An example of the latter is the following code,
929 which moves help to Meta-? (ESC ?):
931 (global-set-key "\M-?" 'help-command)
933 *** At startup I get a warning on stderr about missing charsets:
935 Warning: Missing charsets in String to FontSet conversion
937 You need to specify appropriate charsets for your locale (usually the
938 value of the LANG environment variable) in .Xresources. See
939 etc/Emacs.ad for the relevant resources (mostly menubar fonts and
940 fontsets). Do not edit this file, it's purely informative.
942 If you have no satisfactory fonts for iso-8859-1, XEmacs will crash.
944 It looks like XFree86 4.x (the usual server on Linux and *BSD) has
945 some braindamage where .UTF-8 locales will always generate this
946 message, because the XFree86 (font)server doesn't know that UTF-8 will
947 use the ISO10646-1 font registry (or a Cmap or something).
949 If you are not using a .UTF-8 locale and see this warning for a
950 character set not listed in the default in Emacs.ad, please let
951 xemacs-beta@xemacs.org know about it, so we can add fonts to the
952 appropriate fontsets and stifle this warning. (Unfortunately it's
953 buried in Xlib, so we can't easily get rid of it otherwise.)
955 *** Mail agents (VM, Gnus, rmail) cannot get new mail
957 rmail and VM get new mail from /usr/spool/mail/$USER using a program
958 called `movemail'. This program interlocks with /bin/mail using the
959 protocol defined by /bin/mail.
961 There are two different protocols in general use. One of them uses
962 the `flock' system call. The other involves creating a lock file;
963 `movemail' must be able to write in /usr/spool/mail in order to do
964 this. You control which one is used by defining, or not defining, the
965 macro MAIL_USE_FLOCK in config.h or the m- or s- file it includes. IF
966 YOU DON'T USE THE FORM OF INTERLOCKING THAT IS NORMAL ON YOUR SYSTEM,
969 If your system uses the lock file protocol, and fascist restrictions
970 prevent ordinary users from writing the lock files in /usr/spool/mail,
971 you may need to make `movemail' setgid to a suitable group such as
972 `mail'. To do this, use the following commands (as root) after doing
978 Installation normally copies movemail from the build directory to an
979 installation directory which is usually under /usr/local/lib. The
980 installed copy of movemail is usually in the directory
981 /usr/local/lib/emacs/VERSION/TARGET. You must change the group and
982 mode of the installed copy; changing the group and mode of the build
983 directory copy is ineffective.
985 *** Things which should be bold or italic (such as the initial
986 copyright notice) are not.
988 The fonts of the "bold" and "italic" faces are generated from the font
989 of the "default" face; in this way, your bold and italic fonts will
990 have the appropriate size and family. However, emacs can only be
991 clever in this way if you have specified the default font using the
992 XLFD (X Logical Font Description) format, which looks like
994 *-courier-medium-r-*-*-*-120-*-*-*-*-*-*
996 if you use any of the other, less strict font name formats, some of
999 lucidasanstypewriter-12
1003 then emacs won't be able to guess the names of the "bold" and "italic"
1004 versions. All X fonts can be referred to via XLFD-style names, so you
1005 should use those forms. See the man pages for X(1), xlsfonts(1), and
1008 *** The dumped Emacs crashes when run, trying to write pure data.
1010 Two causes have been seen for such problems.
1012 1) On a system where getpagesize is not a system call, it is defined
1013 as a macro. If the definition (in both unexec.c and malloc.c) is wrong,
1014 it can cause problems like this. You might be able to find the correct
1015 value in the man page for a.out (5).
1017 2) Some systems allocate variables declared static among the
1018 initialized variables. Emacs makes all initialized variables in most
1019 of its files pure after dumping, but the variables declared static and
1020 not initialized are not supposed to be pure. On these systems you
1021 may need to add "#define static" to the m- or the s- file.
1023 *** Reading and writing files is very very slow.
1025 Try evaluating the form (setq lock-directory nil) and see if that helps.
1026 There is a problem with file-locking on some systems (possibly related
1027 to NFS) that I don't understand. Please send mail to the address
1028 xemacs-beta@xemacs.org if you figure this one out.
1030 *** When emacs starts up, I get lots of warnings about unknown keysyms.
1032 If you are running the prebuilt binaries, the Motif library expects to find
1033 certain thing in the XKeysymDB file. This file is normally in /usr/lib/X11/
1034 or in /usr/openwin/lib/. If you keep yours in a different place, set the
1035 environment variable $XKEYSYMDB to point to it before starting emacs. If
1036 you still have the problem after doing that, perhaps your version of X is
1037 too old. There is a copy of the MIT X11R5 XKeysymDB file in the emacs `etc'
1038 directory. Try using that one.
1040 *** My X resources used to work, and now some of them are being ignored.
1042 Check the resources in .../etc/Emacs.ad (which is the same as the file
1043 sample.Xresources). Perhaps some of the default resources built in to
1044 emacs are now overriding your existing resources. Copy and edit the
1045 resources in Emacs.ad as necessary.
1047 *** I have focus problems when I use `M-o' to switch to another screen
1048 without using the mouse.
1050 The focus issues with a program like XEmacs, which has multiple
1051 homogeneous top-level windows, are very complicated, and as a result,
1052 most window managers don't implement them correctly.
1054 The R4/R5 version of twm (and all of its descendants) had buggy focus
1055 handling. Sufficiently recent versions of tvtwm have been fixed. In
1056 addition, if you're using twm, make sure you have not specified
1057 "NoTitleFocus" in your .tvtwmrc file. The very nature of this option
1058 makes twm do some illegal focus tricks, even with the patch.
1060 It is known that olwm and olvwm are buggy, and in different ways. If
1061 you're using click-to-type mode, try using point-to-type, or vice
1064 In older versions of NCDwm, one could not even type at XEmacs windows.
1065 This has been fixed in newer versions (2.4.3, and possibly earlier).
1067 (Many people suggest that XEmacs should warp the mouse when focusing
1068 on another screen in point-to-type mode. This is not ICCCM-compliant
1069 behavior. Implementing such policy is the responsibility of the
1070 window manager itself, it is not legal for a client to do this.)
1072 *** Emacs spontaneously displays "I-search: " at the bottom of the screen.
1074 This means that Control-S/Control-Q (XON/XOFF) "flow control" is being
1075 used. C-s/C-q flow control is bad for Emacs editors because it takes
1076 away C-s and C-q as user commands. Since editors do not output long
1077 streams of text without user commands, there is no need for a
1078 user-issuable "stop output" command in an editor; therefore, a
1079 properly designed flow control mechanism would transmit all possible
1080 input characters without interference. Designing such a mechanism is
1081 easy, for a person with at least half a brain.
1083 There are three possible reasons why flow control could be taking place:
1085 1) Terminal has not been told to disable flow control
1086 2) Insufficient padding for the terminal in use
1087 3) Some sort of terminal concentrator or line switch is responsible
1089 First of all, many terminals have a set-up mode which controls whether
1090 they generate XON/XOFF flow control characters. This must be set to
1091 "no XON/XOFF" in order for Emacs to work. Sometimes there is an
1092 escape sequence that the computer can send to turn flow control off
1093 and on. If so, perhaps the termcap `ti' string should turn flow
1094 control off, and the `te' string should turn it on.
1096 Once the terminal has been told "no flow control", you may find it
1097 needs more padding. The amount of padding Emacs sends is controlled
1098 by the termcap entry for the terminal in use, and by the output baud
1099 rate as known by the kernel. The shell command `stty' will print
1100 your output baud rate; `stty' with suitable arguments will set it if
1101 it is wrong. Setting to a higher speed causes increased padding. If
1102 the results are wrong for the correct speed, there is probably a
1103 problem in the termcap entry. You must speak to a local Unix wizard
1104 to fix this. Perhaps you are just using the wrong terminal type.
1106 For terminals that lack a "no flow control" mode, sometimes just
1107 giving lots of padding will prevent actual generation of flow control
1108 codes. You might as well try it.
1110 If you are really unlucky, your terminal is connected to the computer
1111 through a concentrator which sends XON/XOFF flow control to the
1112 computer, or it insists on sending flow control itself no matter how
1113 much padding you give it. Unless you can figure out how to turn flow
1114 control off on this concentrator (again, refer to your local wizard),
1115 you are screwed! You should have the terminal or concentrator
1116 replaced with a properly designed one. In the mean time, some drastic
1117 measures can make Emacs semi-work.
1119 You can make Emacs ignore C-s and C-q and let the operating system
1120 handle them. To do this on a per-session basis, just type M-x
1121 enable-flow-control RET. You will see a message that C-\ and C-^ are
1122 now translated to C-s and C-q. (Use the same command M-x
1123 enable-flow-control to turn *off* this special mode. It toggles flow
1126 If C-\ and C-^ are inconvenient for you (for example, if one of them
1127 is the escape character of your terminal concentrator), you can choose
1128 other characters by setting the variables flow-control-c-s-replacement
1129 and flow-control-c-q-replacement. But choose carefully, since all
1130 other control characters are already used by emacs.
1132 IMPORTANT: if you type C-s by accident while flow control is enabled,
1133 Emacs output will freeze, and you will have to remember to type C-q in
1136 If you work in an environment where a majority of terminals of a
1137 certain type are flow control hobbled, you can use the function
1138 `enable-flow-control-on' to turn on this flow control avoidance scheme
1139 automatically. Here is an example:
1141 (enable-flow-control-on "vt200" "vt300" "vt101" "vt131")
1143 If this isn't quite correct (e.g. you have a mixture of flow-control hobbled
1144 and good vt200 terminals), you can still run enable-flow-control
1147 I have no intention of ever redesigning the Emacs command set for the
1148 assumption that terminals use C-s/C-q flow control. XON/XOFF flow
1149 control technique is a bad design, and terminals that need it are bad
1150 merchandise and should not be purchased. Now that X is becoming
1151 widespread, XON/XOFF seems to be on the way out. If you can get some
1152 use out of GNU Emacs on inferior terminals, more power to you, but I
1153 will not make Emacs worse for properly designed systems for the sake
1154 of inferior systems.
1156 *** Control-S and Control-Q commands are ignored completely.
1158 For some reason, your system is using brain-damaged C-s/C-q flow
1159 control despite Emacs's attempts to turn it off. Perhaps your
1160 terminal is connected to the computer through a concentrator
1161 that wants to use flow control.
1163 You should first try to tell the concentrator not to use flow control.
1164 If you succeed in this, try making the terminal work without
1165 flow control, as described in the preceding section.
1167 If that line of approach is not successful, map some other characters
1168 into C-s and C-q using keyboard-translate-table. The example above
1169 shows how to do this with C-^ and C-\.
1171 *** Control-S and Control-Q commands are ignored completely on a net
1174 Some versions of rlogin (and possibly telnet) do not pass flow
1175 control characters to the remote system to which they connect.
1176 On such systems, emacs on the remote system cannot disable flow
1177 control on the local system.
1179 One way to cure this is to disable flow control on the local host
1180 (the one running rlogin, not the one running rlogind) using the
1181 stty command, before starting the rlogin process. On many systems,
1182 `stty start u stop u' will do this.
1184 Some versions of tcsh will prevent even this from working. One way
1185 around this is to start another shell before starting rlogin, and
1186 issue the stty command to disable flow control from that shell.
1188 If none of these methods work, the best solution is to type
1189 `M-x enable-flow-control' at the beginning of your emacs session, or
1190 if you expect the problem to continue, add a line such as the
1191 following to your .emacs (on the host running rlogind):
1193 (enable-flow-control-on "vt200" "vt300" "vt101" "vt131")
1195 See the entry about spontaneous display of I-search (above) for more
1198 *** TTY redisplay is slow.
1200 XEmacs has fairly new TTY redisplay support (beginning from 19.12),
1201 which doesn't include some basic TTY optimizations -- like using
1202 scrolling regions to move around blocks of text. This is why
1203 redisplay on the traditional terminals, or over slow lines can be very
1206 If you are interested in fixing this, please let us know at
1207 <xemacs-beta@xemacs.org>.
1209 *** Screen is updated wrong, but only on one kind of terminal.
1211 This could mean that the termcap entry you are using for that terminal
1212 is wrong, or it could mean that Emacs has a bug handing the
1213 combination of features specified for that terminal.
1215 The first step in tracking this down is to record what characters
1216 Emacs is sending to the terminal. Execute the Lisp expression
1217 (open-termscript "./emacs-script") to make Emacs write all terminal
1218 output into the file ~/emacs-script as well; then do what makes the
1219 screen update wrong, and look at the file and decode the characters
1220 using the manual for the terminal. There are several possibilities:
1222 1) The characters sent are correct, according to the terminal manual.
1224 In this case, there is no obvious bug in Emacs, and most likely you
1225 need more padding, or possibly the terminal manual is wrong.
1227 2) The characters sent are incorrect, due to an obscure aspect of the
1228 terminal behavior not described in an obvious way by termcap.
1230 This case is hard. It will be necessary to think of a way for Emacs
1231 to distinguish between terminals with this kind of behavior and other
1232 terminals that behave subtly differently but are classified the same
1233 by termcap; or else find an algorithm for Emacs to use that avoids the
1234 difference. Such changes must be tested on many kinds of terminals.
1236 3) The termcap entry is wrong.
1238 See the file etc/TERMS for information on changes that are known to be
1239 needed in commonly used termcap entries for certain terminals.
1241 4) The characters sent are incorrect, and clearly cannot be right for
1242 any terminal with the termcap entry you were using.
1244 This is unambiguously an Emacs bug, and can probably be fixed in
1245 termcap.c, terminfo.c, tparam.c, cm.c, redisplay-tty.c,
1246 redisplay-output.c, or redisplay.c.
1248 *** My buffers are full of \000 characters or otherwise corrupt.
1250 Some compilers have trouble with gmalloc.c and ralloc.c; try recompiling
1251 without optimization. If that doesn't work, try recompiling with
1252 SYSTEM_MALLOC defined, and/or with REL_ALLOC undefined.
1254 *** A position you specified in .Xresources is ignored, using twm.
1256 twm normally ignores "program-specified" positions.
1257 You can tell it to obey them with this command in your `.twmrc' file:
1259 UsePPosition "on" #allow clents to request a position
1261 *** With M-x enable-flow-control, you need to type C-\ twice to do
1262 incremental search--a single C-\ gets no response.
1264 This has been traced to communicating with your machine via kermit,
1265 with C-\ as the kermit escape character. One solution is to use
1266 another escape character in kermit. One user did
1268 set escape-character 17
1270 in his .kermrc file, to make C-q the kermit escape character.
1272 *** The Motif version of Emacs paints the screen a solid color.
1274 This has been observed to result from the following X resource:
1276 Emacs*default.attributeFont: -*-courier-medium-r-*-*-*-140-*-*-*-*-iso8859-*
1278 That the resource has this effect indicates a bug in something, but we
1279 do not yet know what. If it is an Emacs bug, we hope someone can
1280 explain what the bug is so we can fix it. In the mean time, removing
1281 the resource prevents the problem.
1283 *** After running emacs once, subsequent invocations crash.
1285 Some versions of SVR4 have a serious bug in the implementation of the
1286 mmap () system call in the kernel; this causes emacs to run correctly
1287 the first time, and then crash when run a second time.
1289 Contact your vendor and ask for the mmap bug fix; in the mean time,
1290 you may be able to work around the problem by adding a line to your
1291 operating system description file (whose name is reported by the
1292 configure script) that reads:
1293 #define SYSTEM_MALLOC
1294 This makes Emacs use memory less efficiently, but seems to work around
1297 *** Inability to send an Alt-modified key, when Emacs is communicating
1298 directly with an X server.
1300 If you have tried to bind an Alt-modified key as a command, and it
1301 does not work to type the command, the first thing you should check is
1302 whether the key is getting through to Emacs. To do this, type C-h c
1303 followed by the Alt-modified key. C-h c should say what kind of event
1304 it read. If it says it read an Alt-modified key, then make sure you
1305 have made the key binding correctly.
1307 If C-h c reports an event that doesn't have the Alt modifier, it may
1308 be because your X server has no key for the Alt modifier. The X
1309 server that comes from MIT does not set up the Alt modifier by
1312 If your keyboard has keys named Alt, you can enable them as follows:
1314 xmodmap -e 'add mod2 = Alt_L'
1315 xmodmap -e 'add mod2 = Alt_R'
1317 If the keyboard has just one key named Alt, then only one of those
1318 commands is needed. The modifier `mod2' is a reasonable choice if you
1319 are using an unmodified MIT version of X. Otherwise, choose any
1320 modifier bit not otherwise used.
1322 If your keyboard does not have keys named Alt, you can use some other
1323 keys. Use the keysym command in xmodmap to turn a function key (or
1324 some other 'spare' key) into Alt_L or into Alt_R, and then use the
1325 commands show above to make them modifier keys.
1327 Note that if you have Alt keys but no Meta keys, Emacs translates Alt
1328 into Meta. This is because of the great importance of Meta in Emacs.
1330 *** In Shell mode, you get a ^M at the end of every line.
1332 This happens to people who use tcsh, because it is trying to be too
1333 smart. It sees that the Shell uses terminal type `unknown' and turns
1334 on the flag to output ^M at the end of each line. You can fix the
1335 problem by adding this to your .cshrc file:
1338 if ($EMACS == "t") then
1340 stty -icrnl -onlcr -echo susp ^Z
1344 *** An error message such as `X protocol error: BadMatch (invalid
1345 parameter attributes) on protocol request 93'.
1347 This comes from having an invalid X resource, such as
1349 (which is invalid because it specifies a color name for something
1350 that isn't a color.)
1352 The fix is to correct your X resources.
1354 *** Once you pull down a menu from the menubar, it won't go away.
1356 It has been claimed that this is caused by a bug in certain very old
1357 (1990?) versions of the twm window manager. It doesn't happen with
1358 recent vintages, or with other window managers.
1360 *** Emacs ignores the "help" key when running OLWM.
1362 OLWM grabs the help key, and retransmits it to the appropriate client
1363 using XSendEvent. Allowing emacs to react to synthetic events is a
1364 security hole, so this is turned off by default. You can enable it by
1365 setting the variable x-allow-sendevents to t. You can also cause fix
1366 this by telling OLWM to not grab the help key, with the null binding
1367 "OpenWindows.KeyboardCommand.Help:".
1369 *** Programs running under terminal emulator do not recognize `emacs'
1372 The cause of this is a shell startup file that sets the TERMCAP
1373 environment variable. The terminal emulator uses that variable to
1374 provide the information on the special terminal type that Emacs
1377 Rewrite your shell startup file so that it does not change TERMCAP
1378 in such a case. You could use the following conditional which sets
1379 it only if it is undefined.
1381 if ( ! ${?TERMCAP} ) setenv TERMCAP ~/my-termcap-file
1383 Or you could set TERMCAP only when you set TERM--which should not
1384 happen in a non-login shell.
1386 *** The popup menu appears at the bottom/right of my screen.
1388 You probably have something like the following in your ~/.Xresources
1390 Emacs.geometry: 81x56--9--1
1392 Use the following instead
1394 Emacs*EmacsFrame.geometry: 81x56--9--1
1396 *** When I try to use the PostgreSQL functions, I get a message about
1399 The only known case in which this happens is if you are using gcc, you
1400 configured with --error-checking=all and --with-modules, and you
1401 compiled with no optimization. If you encounter this problem in any
1402 other situation, please inform xemacs-beta@xemacs.org.
1404 This problem stems from a gcc bug. With no optimization, functions
1405 declared `extern inline' sometimes are not completely compiled away. An
1406 undefined symbol with the function's name is put into the resulting
1407 object file. In this case, when the postgresql module is loaded, the
1408 linker is unable to resolve that symbol, so the module load fails. The
1409 workaround is to recompile the module with optimization turned on. Any
1410 optimization level, including -Os, appears to work.
1412 *** C-z just refreshes the screen instead of suspending Emacs.
1414 You are probably using a shell that doesn't support job control, even
1415 though the system itself is capable of it. Try using a different
1419 *** XEmacs crashes on MacOS within font-lock, or when dealing
1420 with large compilation buffers, or in other regex applications.
1422 The default stack size under MacOS/X is rather small (512k as opposed
1423 to Solaris 8M), hosing the regexp code, which uses alloca()
1424 extensively, overflowing the stack when complex regexps are used.
1427 1) Increase your stack size, using `ulimit -s 8192' or a (t)csh
1430 2) Recompile regex.c with REGEX_MALLOC defined.
1433 *** Your Delete key sends a Backspace to the terminal, using an AIXterm.
1435 The solution is to include in your .Xresources the lines:
1437 *aixterm.Translations: #override <Key>BackSpace: string(0x7f)
1438 aixterm*ttyModes: erase ^?
1440 This makes your Backspace key send DEL (ASCII 127).
1442 *** On AIX 4, some programs fail when run in a Shell buffer
1443 with an error message like No terminfo entry for "unknown".
1445 On AIX, many terminal type definitions are not installed by default.
1446 `unknown' is one of them. Install the "Special Generic Terminal
1447 Definitions" to make them defined.
1449 *** On AIX, you get this message when running Emacs:
1451 Could not load program emacs
1452 Symbol smtcheckinit in csh is undefined
1453 Error was: Exec format error
1457 Could not load program .emacs
1458 Symbol _system_con in csh is undefined
1459 Symbol _fp_trapsta in csh is undefined
1460 Error was: Exec format error
1462 These can happen when you try to run on AIX 3.2.5 a program that was
1463 compiled with 3.2.4. The fix is to recompile.
1465 *** Trouble using ptys on AIX.
1467 People often install the pty devices on AIX incorrectly.
1468 Use `smit pty' to reinstall them properly.
1472 *** The Emacs window disappears when you type M-q.
1474 Some versions of the Open Look window manager interpret M-q as a quit
1475 command for whatever window you are typing at. If you want to use
1476 Emacs with that window manager, you should try to configure the window
1477 manager to use some other command. You can disable the
1478 shortcut keys entirely by adding this line to ~/.OWdefaults:
1480 OpenWindows.WindowMenuAccelerators: False
1482 *** When Emacs tries to ring the bell, you get an error like
1484 audio: sst_open: SETQSIZE" Invalid argument
1485 audio: sst_close: SETREG MMR2, Invalid argument
1487 you have probably compiled using an ANSI C compiler, but with non-ANSI
1488 include files. In particular, on Suns, the file
1489 /usr/include/sun/audioio.h uses the _IOW macro to define the constant
1490 AUDIOSETQSIZE. _IOW in turn uses a K&R preprocessor feature that is
1491 now explicitly forbidden in ANSI preprocessors, namely substitution
1492 inside character constants. All ANSI C compilers must provide a
1493 workaround for this problem. Lucid's C compiler is shipped with a new
1494 set of system include files. If you are using GCC, there is a script
1495 called fixincludes that creates new versions of some system include
1496 files that use this obsolete feature.
1498 *** On Solaris 2.6, XEmacs dumps core when exiting.
1500 This happens if you're XEmacs is running on the same machine as the X
1501 server, and the optimized memory transport has been turned on by
1502 setting the environment variable XSUNTRANSPORT. The crash occurs
1503 during the call to XCloseDisplay.
1505 If this describes your situation, you need to undefine the
1506 XSUNTRANSPORT environment variable.
1508 *** On Solaris, C-x doesn't get through to Emacs when you use the console.
1510 This is a Solaris feature (at least on Intel x86 cpus). Type C-r
1511 C-r C-t, to toggle whether C-x gets through to Emacs.
1513 *** On Solaris 2.4, Dired hangs and C-g does not work. Or Emacs hangs
1514 forever waiting for termination of a subprocess that is a zombie.
1516 casper@fwi.uva.nl says the problem is in X11R6. Rebuild libX11.so
1517 after changing the file xc/config/cf/sunLib.tmpl. Change the lines
1520 #define SharedX11Reqs -lthread
1525 #if OSMinorVersion < 4
1527 #define SharedX11Reqs -lthread
1531 Be sure also to edit x/config/cf/sun.cf so that OSMinorVersion is 4
1532 (as it should be for Solaris 2.4). The file has three definitions for
1533 OSMinorVersion: the first is for x86, the second for SPARC under
1534 Solaris, and the third for SunOS 4. Make sure to update the
1535 definition for your type of machine and system.
1537 Then do `make Everything' in the top directory of X11R6, to rebuild
1538 the makefiles and rebuild X. The X built this way work only on
1539 Solaris 2.4, not on 2.3.
1541 For multithreaded X to work it necessary to install patch
1542 101925-02 to fix problems in header files [2.4]. You need
1543 to reinstall gcc or re-run just-fixinc after installing that
1546 However, Frank Rust <frust@iti.cs.tu-bs.de> used a simpler solution:
1548 #define ThreadedX YES
1550 #define ThreadedX NO
1551 in sun.cf and did `make World' to rebuild X11R6. Removing all
1552 `-DXTHREAD*' flags and `-lthread' entries from lib/X11/Makefile and
1553 typing 'make install' in that directory also seemed to work.
1555 *** On SunOS 4.1.3, Emacs unpredictably crashes in _yp_dobind_soft.
1557 This happens if you configure Emacs specifying just `sparc-sun-sunos4'
1558 on a system that is version 4.1.3. You must specify the precise
1559 version number (or let configure figure out the configuration, which
1560 it can do perfectly well for SunOS).
1562 *** Mail is lost when sent to local aliases.
1564 Many emacs mail user agents (VM and rmail, for instance) use the
1565 sendmail.el library. This library can arrange for mail to be
1566 delivered by passing messages to the /usr/lib/sendmail (usually)
1567 program . In doing so, it passes the '-t' flag to sendmail, which
1568 means that the name of the recipient of the message is not on the
1569 command line and, therefore, that sendmail must parse the message to
1570 obtain the destination address.
1572 There is a bug in the SunOS4.1.1 and SunOS4.1.3 versions of sendmail.
1573 In short, when given the -t flag, the SunOS sendmail won't recognize
1574 non-local (i.e. NIS) aliases. It has been reported that the Solaris
1575 2.x versions of sendmail do not have this bug. For those using SunOS
1576 4.1, the best fix is to install sendmail V8 or IDA sendmail (which
1577 have other advantages over the regular sendmail as well). At the time
1578 of this writing, these official versions are available:
1580 Sendmail V8 on ftp.cs.berkeley.edu in /ucb/sendmail:
1581 sendmail.8.6.9.base.tar.Z (the base system source & documentation)
1582 sendmail.8.6.9.cf.tar.Z (configuration files)
1583 sendmail.8.6.9.misc.tar.Z (miscellaneous support programs)
1584 sendmail.8.6.9.xdoc.tar.Z (extended documentation, with postscript)
1586 IDA sendmail on vixen.cso.uiuc.edu in /pub:
1587 sendmail-5.67b+IDA-1.5.tar.gz
1589 *** Emacs fails to understand most Internet host names, even though
1590 the names work properly with other programs on the same system.
1591 Emacs won't work with X-windows if the value of DISPLAY is HOSTNAME:0.
1592 Gnus can't make contact with the specified host for nntp.
1594 This typically happens on Suns and other systems that use shared
1595 libraries. The cause is that the site has installed a version of the
1596 shared library which uses a name server--but has not installed a
1597 similar version of the unshared library which Emacs uses.
1599 The result is that most programs, using the shared library, work with
1600 the nameserver, but Emacs does not.
1602 The fix is to install an unshared library that corresponds to what you
1603 installed in the shared library, and then relink Emacs.
1605 On SunOS 4.1, simply define HAVE_RES_INIT.
1607 If you have already installed the name resolver in the file libresolv.a,
1608 then you need to compile Emacs to use that library. The easiest way to
1609 do this is to add to config.h a definition of LIBS_SYSTEM, LIBS_MACHINE
1610 or LIB_STANDARD which uses -lresolv. Watch out! If you redefine a macro
1611 that is already in use in your configuration to supply some other libraries,
1612 be careful not to lose the others.
1614 Thus, you could start by adding this to config.h:
1616 #define LIBS_SYSTEM -lresolv
1618 Then if this gives you an error for redefining a macro, and you see that
1619 the s- file defines LIBS_SYSTEM as -lfoo -lbar, you could change config.h
1622 #define LIBS_SYSTEM -lresolv -lfoo -lbar
1624 *** With process-connection-type set to t, each line of subprocess
1625 output is terminated with a ^M, making ange-ftp and GNUS not work.
1627 On SunOS systems, this problem has been seen to be a result of an
1628 incomplete installation of gcc 2.2 which allowed some non-ANSI
1629 compatible include files into the compilation. In particular this
1630 affected virtually all ioctl() calls.
1634 *** XEmacs crashes on startup, in make-frame.
1636 Typically the Lisp backtrace includes
1638 make-frame(nil #<x-device on ":0.0" 0x2558>)
1640 somewhere near the top. The problem is due to an improvement in GNU
1641 ld that sorts the ELF reloc sections in the executable, giving
1642 dramatic speedups in startup for large executables. It also confuses
1643 the traditional unexec code in XEmacs, leading to the core dump. The
1644 solution is to use the --pdump or --ldflags='-z nocombreloc' options
1645 to configure. Recent 21.4 and 12.5 autodetect this in configure.
1647 Red Hat and SuSE (at least) distributed a prerelease version of ld
1648 (versions around 2.11.90.x.y) where autodetection is impossible. The
1649 recommended procedure is to upgrade to binutils >= 2.12 and rerun
1650 configure. Otherwise you must apply the flags by hand. --pdump is
1653 *** I want XEmacs to use the Alt key, not the XXX key, for Meta commands
1655 For historical reasons, XEmacs looks for a Meta key, then an Alt key.
1656 It binds Meta commands to the X11 modifier bit attached to the first
1657 of these it finds. On PCs, the Windows key is often assigned the Meta
1658 bit, but many desktop environments go to great lengths to get all apps
1659 to use the Alt key, and reserve the Windows key to (sensibly enough)
1662 One correct way to implement this was suggested on comp.emacs.xemacs
1663 (by Kilian Foth and in more detail by Michael Piotrowski): unmap the
1664 Meta modifier using xmodmap or xkb, and then map the Meta/Windows key
1665 to the Super or Hyper keysym and an appropriate mod bit. XEmacs will
1666 not find the Meta keysym, and default to using the Alt key for Meta
1667 keybindings. Typically few applications use the (X11) Meta modifier;
1668 it is tedious but not too much so to teach the ones you need to use
1669 Super instead of Meta. There may be further useful hints in the
1670 discussion of keymapping on non-Linux platforms.
1672 *** The color-gcc wrapper
1674 This wrapper colorizes the error messages from gcc. By default XEmacs
1675 does not interpret the escape sequences used to generate colors,
1676 resulting in a cluttered, hard-to-read buffer. You can remove the
1677 wrapper, or defeat the wrapper colorization in Emacs process buffers
1678 by editing the "nocolor" attribute in /etc/colorgccrc:
1680 $ diff -u /etc/colorgccrc.old /etc/colorgccrc
1681 --- /etc/colorgccrc.old Tue Dec 26 02:17:46 2000
1682 +++ /etc/colorgccrc Tue Dec 26 02:15:48 2000
1685 +nocolor: dumb emacs
1687 If you want colorization in your Emacs buffers, you may get good
1688 results from the ansi-color.el library:
1690 http://www.geocities.com/kensanata/color-emacs.html#ansicolors
1692 This is written for the mainline GNU Emacs but the author has made
1693 efforts to adapt it to XEmacs. YMMV.
1695 *** Slow startup on Linux.
1697 People using systems based on the Linux kernel sometimes report that
1698 startup takes 10 to 15 seconds longer than `usual'. There are two
1699 problems, one older, one newer.
1701 **** Old problem: IPv4 host lookup
1703 On older systems, this is because Emacs looks up the host name when it
1704 starts. Normally, this takes negligible time; the extra delay is due
1705 to improper system configuration. (Recent Linux distros usually have
1706 this configuration correct "out of the box".) This problem can occur
1707 for both networked and non-networked machines.
1709 Here is how to fix the configuration. It requires being root.
1711 ***** Networked Case
1713 First, make sure the files `/etc/hosts' and `/etc/host.conf' both
1714 exist. The first line in the `/etc/hosts' file should look like this
1715 (replace HOSTNAME with your host name):
1717 127.0.0.1 localhost HOSTNAME
1719 Also make sure that the `/etc/host.conf' files contains the following
1725 Any changes, permanent and temporary, to the host name should be
1726 indicated in the `/etc/hosts' file, since it acts a limited local
1727 database of addresses and names (e.g., some SLIP connections
1728 dynamically allocate ip addresses).
1730 ***** Non-Networked Case
1732 The solution described in the networked case applies here as well.
1733 However, if you never intend to network your machine, you can use a
1734 simpler solution: create an empty `/etc/host.conf' file. The command
1735 `touch /etc/host.conf' suffices to create the file. The `/etc/hosts'
1736 file is not necessary with this approach.
1738 **** New problem: IPv6 CNAME lookup
1740 A newer problem is due to XEmacs changing to use the modern
1741 getaddrinfo() interface from the older gethostbyname() interface. The
1742 solution above is insufficient, because getaddrinfo() by default tries
1743 to get IPv6 information for localhost. This always involves a dns
1744 lookup to get the CNAME, and the strategies above don't work. It then
1745 falls back to IPv4 behavior. This is good[tm] according the people at
1746 WIDE who know about IPv6.
1748 ***** Robust network case
1750 Configure your network so that there are no nameservers configured
1751 until the network is actually running. getaddrinfo() will not try to
1752 access a nameserver that isn't configured.
1754 ***** Flaky network case
1756 If you have a flaky modem or DSL connection that can be relied on only
1757 to go down whenever you want to bring XEmacs up, you need to force
1758 IPv4 behavior. Explicitly setting DISPLAY=127.0.0.1:0.0 (or whatever
1759 is appropriate) works in most cases.
1761 If you cannot or do not want to do that, you can hard code IPv4
1762 behavior in src/process-unix.c. This is bad[tm], on your own head be
1763 it. Use the configure option `--with-ipv6-cname=no'.
1767 The Mandrake Linux distribution is attempting to comprehensively
1768 update the user interface, and make it consistent across
1769 applications. This is very difficult, and will occasionally cause
1770 conflicts with applications like Emacs with their own long-established
1771 interfaces. Known issues specific to Mandrake or especially common:
1773 Some versions of XEmacs (21.1.9 is known) distributed with Mandrake
1774 were patched to make the Meta and Alt keysyms synonymous. These
1775 normally work as expected in the Mandrake environment. However,
1776 custom-built XEmacsen (including all 21.2 betas) will "inexplicably"
1777 not respect the "Alt-invokes-Meta-commands" convention. See "I want
1778 XEmacs to use the Alt key" below.
1780 The color-gcc wrapper (see below) is in common use on the Mandrake
1783 *** You get crashes in a non-C locale with Linux GNU Libc 2.0.
1785 Internationalization was not the top priority for GNU Libc 2.0.
1786 As of this writing (1998-12-28) you may get crashes while running
1787 XEmacs in a non-C locale. For example, `LC_ALL=en_US xemacs' crashes
1788 while `LC_ALL=C xemacs' runs fine. This happens for example with GNU
1789 libc 2.0.7. Installing libintl.a and libintl.h built from gettext
1790 0.10.35 and re-building XEmacs solves the crashes. Presumably soon
1791 everyone will upgrade to GNU Libc 2.1 and this problem will go away.
1793 *** `C-z', or `M-x suspend-emacs' hangs instead of suspending.
1795 If you build with `gpm' support on Linux, you cannot suspend XEmacs
1796 because gpm installs a buggy SIGTSTP handler. Either compile with
1797 `--with-gpm=no', or don't suspend XEmacs on the Linux console until
1800 *** With certain fonts, when the cursor appears on a character, the
1801 character doesn't appear--you get a solid box instead.
1803 One user on a Linux system reported that this problem went away with
1804 installation of a new X server. The failing server was XFree86 3.1.1.
1805 XFree86 3.1.2 works.
1808 *** On Irix, I don't see the toolbar icons and I'm getting lots of
1809 entries in the warnings buffer.
1811 SGI ships a really old Xpm library in /usr/lib which does not work at
1812 all well with XEmacs. The solution is to install your own copy of the
1813 latest version of Xpm somewhere and then use the --site-includes and
1814 --site-libraries flags to tell configure where to find it.
1816 *** Trouble using ptys on IRIX, or running out of ptys.
1818 The program mkpts (which may be in `/usr/adm' or `/usr/sbin') needs to
1819 be set-UID to root, or non-root programs like Emacs will not be able
1820 to allocate ptys reliably.
1822 *** Beware of the default image & graphics library on Irix
1824 Richard Cognot <cognot@ensg.u-nancy.fr> writes:
1826 You *have* to compile your own jpeg lib. The one delivered with SGI
1827 systems is a C++ lib, which apparently XEmacs cannot cope with.
1830 ** Digital UNIX/OSF/VMS/Ultrix
1831 *** XEmacs crashes on Digital Unix within font-lock, or when dealing
1832 with large compilation buffers, or in other regex applications.
1834 The default stack size under Digital Unix is rather small (2M as
1835 opposed to Solaris 8M), hosing the regexp code, which uses alloca()
1836 extensively, overflowing the stack when complex regexps are used.
1839 1) Increase your stack size, using `ulimit -s 8192' or a (t)csh
1842 2) Recompile regex.c with REGEX_MALLOC defined.
1844 *** The `Alt' key doesn't behave as `Meta' when running DECwindows.
1846 The default DEC keyboard mapping has the Alt keys set up to generate the
1847 keysym `Multi_key', which has a meaning to xemacs which is distinct from that
1848 of the `Meta_L' and `Meta-R' keysyms. A second problem is that certain keys
1849 have the Mod2 modifier attached to them for no adequately explored reason.
1850 The correct fix is to pass this file to xmodmap upon starting X:
1853 keysym Multi_key = Alt_L
1857 *** The Compose key on a DEC keyboard does not work as Meta key.
1859 This shell command should fix it:
1861 xmodmap -e 'keycode 0xb1 = Meta_L'
1863 *** `expand-file-name' fails to work on any but the machine you dumped
1866 On Ultrix, if you use any of the functions which look up information
1867 in the passwd database before dumping Emacs (say, by using
1868 expand-file-name in site-init.el), then those functions will not work
1869 in the dumped Emacs on any host but the one Emacs was dumped on.
1871 The solution? Don't use expand-file-name in site-init.el, or in
1872 anything it loads. Yuck - some solution.
1874 I'm not sure why this happens; if you can find out exactly what is
1875 going on, and perhaps find a fix or a workaround, please let us know.
1876 Perhaps the YP functions cache some information, the cache is included
1877 in the dumped Emacs, and is then inaccurate on any other host.
1881 *** I get complaints about the mapping of my HP keyboard at startup,
1882 but I haven't changed anything.
1884 The default HP keymap is set up to have Mod1 assigned to two different keys:
1885 Meta_L and Mode_switch (even though there is not actually a Mode_switch key on
1886 the keyboard -- it uses an "imaginary" keycode.) There actually is a reason
1887 for this, but it's not a good one. The correct fix is to execute this command
1890 xmodmap -e 'remove mod1 = Mode_switch'
1892 *** On HP-UX, you get "poll: Interrupted system call" message in the
1893 window where XEmacs was launched.
1895 Richard Cognot <cognot@ensg.u-nancy.fr> writes:
1897 I get a very strange problem when linking libc.a dynamically: every
1898 event (mouse, keyboard, expose...) results in a "poll: Interrupted
1899 system call" message in the window where XEmacs was
1900 launched. Forcing a static link of libc.a alone by adding
1901 /usr/lib/libc.a at the end of the link line solves this. Note that
1902 my 9.07 build of 19.14b17 and my (old) build of 19.13 both exhibit
1903 the same behavior. I've tried various hpux patches to no avail. If
1904 this problem cannot be solved before the release date, binary kits
1905 for HP *must* be linked statically against libc, otherwise this
1906 problem will show up. (This is directed at whoever will volunteer
1907 for this kit, as I won't be available to do it, unless 19.14 gets
1908 delayed until mid-june ;-). I think this problem will be an FAQ soon
1909 after the release otherwise.
1911 Note: The above entry is probably not valid for XEmacs 21.0 and
1914 *** The right Alt key works wrong on German HP keyboards (and perhaps
1915 other non-English HP keyboards too).
1917 This is because HP-UX defines the modifiers wrong in X. Here is a
1918 shell script to fix the problem; be sure that it is run after VUE
1919 configures the X server.
1921 xmodmap 2> /dev/null - << EOF
1922 keysym Alt_L = Meta_L
1923 keysym Alt_R = Meta_R
1928 keysym Mode_switch = NoSymbol
1930 keysym Meta_R = Mode_switch
1931 add mod2 = Mode_switch
1935 *** XEmacs dumps core at startup when native audio is used. Native
1936 audio does not work with recent versions of HP-UX.
1938 Under HP-UX 10.20 and later (e.g., HP-UX 11.XX), with native audio
1939 enabled, the dumped XEmacs binary ("xemacs") core dumps at startup if
1940 recent versions of the libAlib.sl audio shared library is used. Note
1941 that "temacs" will run, but "xemacs" will dump core. This, of course,
1942 causes the XEmacs build to fail. If GNU malloc is enabled, a stack
1943 trace will show XEmacs to have crashed in the "first" call to malloc().
1945 This bug currently exists in all versions of XEmacs, when the undump
1946 mechanism is used. It is not known if using the experimental portable
1947 dumper will allow native audio to work.
1951 Recent versions of the HP-UX 10.20 (and later) audio shared library (in
1952 /opt/audio/lib), pulls in the libdce shared library, which pulls in a
1953 thread (libcma) library. This prevents the HP-UX undump() routine (in
1954 unexhp9k800.c) from properly working. What's happening is that some
1955 initialization routines are being called in the libcma library, *BEFORE*
1956 main() is called, and these initialization routines are calling
1957 malloc(). Unfortunately, in order for the undumper to work, XEmacs must
1958 adjust (move upwards) the sbrk() value *BEFORE* the first call to
1959 malloc(); if malloc() is called before XEmacs has properly adjusted sbrk
1960 (which is what is happening), dumped memory that is being used by
1961 XEmacs, is improperly re-allocated for use by malloc() and the dumped
1962 memory is corrupted. This causes XEmacs to die an horrible death.
1964 It is believed that versions of the audio library past December 1998
1965 will trigger this problem. Under HP-UX 10.20, you probably have to
1966 install audio library patches to encounter this. It's probable that
1967 recent "fresh, out-of-the-box" HP-UX 11.XX workstations also have this
1968 problem. For HP-UX 10.20, it's believed that audio patch PHSS_17121 (or
1969 a superceeding one, like PHSS_17554, PHSS_17971, PHSS_18777, PHSS_21481,
1970 or PHSS_21662, etc.) will trigger this.
1972 To check if your audio library will cause problems for XEmacs, run
1973 "chatr /opt/audio/lib/libAlib.sl". If "libdce" appears in the displayed
1974 shared library list, XEmacs will probably encounter problems if audio is
1979 Don't enable native audio. Re-run configure without native audio
1982 If your site supports it, try using NAS (Network Audio Support).
1984 Try using the experimental portable dumper. It may work, or it may
1988 *** `Pid xxx killed due to text modification or page I/O error'
1990 On HP-UX, you can get that error when the Emacs executable is on an NFS
1991 file system. HP-UX responds this way if it tries to swap in a page and
1992 does not get a response from the server within a timeout whose default
1993 value is just ten seconds.
1995 If this happens to you, extend the timeout period.
1997 *** Shell mode on HP-UX gives the message, "`tty`: Ambiguous".
1999 christos@theory.tn.cornell.edu says:
2001 The problem is that in your .cshrc you have something that tries to
2002 execute `tty`. If you are not running the shell on a real tty then tty
2003 will print "not a tty". Csh expects one word in some places, but tty
2004 is giving it back 3.
2006 The solution is to add a pair of quotes around `tty` to make it a
2009 if (`tty` == "/dev/console")
2011 should be changed to:
2013 if ("`tty`" == "/dev/console")
2015 Even better, move things that set up terminal sections out of .cshrc
2020 *** Regular expressions matching bugs on SCO systems.
2022 On SCO, there are problems in regexp matching when Emacs is compiled
2023 with the system compiler. The compiler version is "Microsoft C
2024 version 6", SCO 4.2.0h Dev Sys Maintenance Supplement 01/06/93; Quick
2025 C Compiler Version 1.00.46 (Beta). The solution is to compile with
2030 *** Conflicts with FSF NTEmacs
2032 Depending on how it is installed, FSF NTEmacs may setup various EMACS*
2033 variables in your environment. The presence of these variables may
2034 cause XEmacs to fail at startup, cause you to see corrupted
2035 doc-strings, or cause other random problems.
2037 You should remove these variables from your environment. These
2038 variables are not required to run FSF NTEmacs if you start it by
2041 *** XEmacs can't find my init file
2043 XEmacs looks for your init in your "home" directory -- either in
2044 `~/.xemacs/init.el' or `~/.emacs'. XEmacs decides that your "home"
2045 directory is, in order of preference:
2047 - The value of the HOME environment variable, if the variable exists.
2048 - The value of the registry entry SOFTWARE\XEmacs\XEmacs\HOME,
2050 - The value of the HOMEDRIVE and HOMEPATH environment variables, if
2051 these variables both exist.
2054 To determine what XEmacs thinks your home directory is, try opening
2055 a file in the `~' directory, and you should see its expansion in the
2056 modeline. If this doesn't work, type ESC : (user-home-directory).
2058 *** XEmacs can't find any packages
2060 XEmacs looks for your packages in subdirectories of a directory which
2061 is set at compile-time (see `config.inc'), and whose default is
2062 `C:\Program Files\XEmacs'. XEmacs also looks in `~/.xemacs', where
2063 `~' refers to your home directory (see previous entry). The variable
2064 `configure-package-path' holds the actual path that was compiled into
2065 your copy of XEmacs.
2067 The compile-time default location can be overridden by the EMACSPACKAGEPATH
2068 environment variable or by the SOFTWARE\XEmacs\XEmacs\EMACSPACKAGEPATH
2069 registry entry. You should check that these variables, if they exist,
2070 point to the actual location of your package tree.
2072 *** XEmacs doesn't die when shutting down Windows 95 or 98
2074 When shutting down Windows 95 or 98 you may see a dialog that says
2075 "xemacs / You must quit this program before you quit Windows".
2077 "Click OK to quit the program and Windows",
2078 but you won't be offered a chance to save any modified XEmacs buffers.
2082 The C-z, C-x, C-c, and C-v keystrokes have traditional uses in both
2083 emacs and Windows programs. XEmacs binds these keys to their
2084 traditional emacs uses, and provides Windows 3.x style bindings for
2085 the Cut, Copy and Paste functions.
2087 Function XEmacs binding
2088 -------- --------------
2094 You can rebind keys to make XEmacs more Windows-compatible; for
2095 example, to bind C-z to undo:
2097 (global-set-key [(control z)] 'undo)
2099 Rebindind C-x and C-c is trickier because by default these are prefix
2100 keys in XEmacs. See the "Key Bindings" node in the XEmacs manual.
2102 *** Behavior of selected regions
2104 Use the pending-del package to enable the standard Windows behavior of
2105 self-inserting deletes region.
2107 *** Limitations on the use of the AltGr key.
2109 In some locale and OS combinations you can't generate M-AltGr-key or
2110 C-M-AltGr-key sequences at all.
2112 To generate C-AltGr-key or C-M-AltGr-key sequences you must use the
2113 right-hand Control key and you must press it *after* AltGr.
2115 These limitations arise from fundamental problems in the way that the
2116 win32 API reports AltGr key events. There isn't anything that XEmacs
2117 can do to work round these problems that it isn't already doing.
2119 You may want to create alternative bindings if any of the standard
2120 XEmacs bindings require you to use some combination of Control or Meta
2123 *** Limited support for subprocesses under Windows 9x
2125 Attempting to use call-process to run a 16bit program gives a
2126 "Spawning child process: Exec format error". For example shell-command
2127 fails under Windows 95 and 98 if you use command.com or any other
2128 16bit program as your shell.
2130 XEmacs may incorrectly quote your call-process command if it contains
2131 double quotes, backslashes or spaces.
2133 start-process and functions that rely on it are supported under Windows 95,
2134 98 and NT. However, starting a 16bit program that requires keyboard input
2135 may cause XEmacs to hang or crash under Windows 95 and 98, and will leave
2136 the orphaned 16bit program consuming all available CPU time.
2138 Sending signals to subprocesses started by call-process or by
2139 start-process fails with a "Cannot send signal to process" error under
2140 Windows 95 and 98. As a side effect of this, quitting XEmacs while it
2141 is still running subprocesses causes it to crash under Windows 95 and
2146 *** Signal 11 when building or running a dumped XEmacs.
2148 See the section on Cygwin above, under building.
2150 *** XEmacs fails to start because cygXpm-noX4.dll was not found.
2152 Andy Piper <andy@xemacs.org> sez:
2154 cygXpm-noX4 is part of the cygwin distribution under libraries or
2155 graphics, but is not installed by default. You need to run the
2156 cygwin setup again and select this package.
2158 *** Subprocesses do not work.
2160 You do not have "tty" in your CYGWIN environment variable. This must
2161 be set in your autoexec.bat (win95) or the system properties (winnt)
2162 as it must be read before the cygwin DLL initializes.
2164 *** ^G does not work on hung subprocesses.
2166 This is a known problem. It can be remedied by defining BROKEN_SIGIO
2167 in src/s/cygwin.h, however this currently leads to instability in XEmacs.
2168 (#### is this still true?)
2170 *** Errors from make like `/c:not found' when running `M-x compile'.
2172 Make sure you set the environment variable MAKE_MODE to UNIX in your
2173 init file (.xemacs/init.el), Control Panel (Windows 2000/NT), or
2174 AUTOEXEC.BAT (Windows 98/95).
2176 *** There are no images in the toolbar buttons.
2178 You need version 4.71 of commctrl.dll which does not ship with windows
2179 95. You can get this by installing IE 4.0 or downloading it from the
2183 * Compatibility problems (with Emacs 18, GNU Emacs, or previous XEmacs/lemacs)
2184 ==============================================================================
2186 *** "Symbol's value as variable is void: unread-command-char".
2187 "Wrong type argument: arrayp, #<keymap 143 entries>"
2188 "Wrong type argument: stringp, [#<keypress-event return>]"
2190 There are a few incompatible changes in XEmacs, and these are the
2191 symptoms. Some of the emacs-lisp code you are running needs to be
2192 updated to be compatible with XEmacs.
2194 The code should not treat keymaps as arrays (use `define-key', etc.),
2195 should not use obsolete variables like `unread-command-char' (use
2196 `unread-command-events'). Many (most) of the new ways of doing things
2197 are compatible in GNU Emacs and XEmacs.
2199 Modern Emacs packages (Gnus, VM, W3, efs, etc) are written to support
2200 GNU Emacs and XEmacs. We have provided modified versions of several
2201 popular emacs packages (dired, etc) which are compatible with this
2202 version of emacs. Check to make sure you have not set your load-path
2203 so that your private copies of these packages are being found before
2204 the versions in the lisp directory.
2206 Make sure that your load-path and your $EMACSLOADPATH environment
2207 variable are not pointing at an Emacs18 lisp directory. This will
2210 ** Some packages that worked before now cause the error
2211 Wrong type argument: arrayp, #<face ... >
2213 Code which uses the `face' accessor functions must be recompiled with
2214 xemacs 19.9 or later. The functions whose callers must be recompiled
2215 are: face-font, face-foreground, face-background,
2216 face-background-pixmap, and face-underline-p. The .elc files
2217 generated by version 19.9 will work in 19.6 and 19.8, but older .elc
2218 files which contain calls to these functions will not work in 19.9.
2220 ** Signaling: (error "Byte code stack underflow (byte compiler bug), pc 38")
2222 This error is given when XEmacs 20 is compiled without MULE support
2223 but is attempting to load a .elc which requires MULE support. The fix
2224 is to rebytecompile the offending file.
2226 ** Signaling: (wrong-type-argument ...) when loading mail-abbrevs
2228 The is seen when installing the Insidious Big Brother Data Base (bbdb)
2229 which includes an outdated copy of mail-abbrevs.el. Remove the copy
2230 that comes with bbdb and use the one that comes with XEmacs.
2236 ** A reminder: XEmacs/Mule work does not currently receive *any*
2237 funding, and all work is done by volunteers. If you think you can
2238 help, please contact the XEmacs maintainers.
2240 ** XEmacs/Mule doesn't support TTY's satisfactorily.
2242 This is a major problem, which we plan to address in a future release
2243 of XEmacs. Basically, XEmacs should have primitives to be told
2244 whether the terminal can handle international output, and which
2245 locale. Also, it should be able to do approximations of characters to
2246 the nearest supported by the locale.
2248 ** Internationalized (Asian) Isearch doesn't work.
2250 Currently, Isearch doesn't directly support any of the input methods
2251 that are not XIM based (like egg, canna and quail) (and there are
2252 potential problems with XIM version too...). If you're using egg
2253 there is a workaround. Hitting <RET> right after C-s to invoke
2254 Isearch will put Isearch in string mode, where a complete string can
2255 be typed into the minibuffer and then processed by Isearch afterwards.
2256 Since egg is now supported in the minibuffer using string mode you can
2257 now use egg to input your Japanese, Korean or Chinese string, then hit
2258 return to send that to Isearch and then use standard Isearch commands
2261 ** Using egg and mousing around while in 'fence' mode screws up my
2264 Don't do this. The fence modes of egg and canna are currently very
2265 modal, and messing with where they expect point to be and what they
2266 think is the current buffer is just asking for trouble. If you're
2267 lucky they will realize that something is awry, and simply delete the
2268 fence, but worst case can trash other buffers too. We've tried to
2269 protect against this where we can, but there still are many ways to
2270 shoot yourself in the foot. So just finish what you are typing into
2271 the fence before reaching for the mouse.
2273 ** Not all languages in Quail are supported like Devanagari and Indian
2274 languages, Lao and Tibetan.
2276 Quail requires more work and testing. Although it has been ported to
2277 XEmacs, it works really well for Japanese and for the European
2280 ** Right-to-left mode is not yet implemented, so languages like
2281 Arabic, Hebrew and Thai don't work.
2283 Getting this right requires more work. It may be implemented in a
2284 future XEmacs version, but don't hold your breath. If you know
2285 someone who is ready to implement this, please let us know.
2287 ** We need more developers and native language testers. It's extremely
2288 difficult (and not particularly productive) to address languages that
2289 nobody is using and testing.
2291 ** The kWnn and cWnn support for Chinese and Korean needs developers
2292 and testers. It probably doesn't work.
2294 ** There are no `native XEmacs' TUTORIALs for any Asian languages,
2295 including Japanese. FSF Emacs and XEmacs tutorials are quite similar,
2296 so it should be sufficient to skim through the differences and apply
2297 them to the Japanese version.
2299 ** We only have localized menus translated for Japanese, and the
2300 Japanese menus are developing bitrot (the Mule menu appears in
2303 ** XIM is untested for any language other than Japanese.