* src/xterm.c (x_term_init): Only select for
RROutputPropertyNotify if GTK is in use. Emacs doesn't need it
itself, and output property changes will otherwise lead to 2
second freezes over connections with high network latency.
* src/xfns.c (x_set_use_frame_synchronization): Announce that we
do not want the compositor to unredirect the frame while
fullscreen.
* src/xterm.c (x_atom_refs): New atom
`_NET_WM_BYPASS_COMPOSITOR'.
* src/xterm.h (struct x_display_info): Likewise.
* src/xterm.c (x_detect_focus_change): Finally figure out what
the call to x_new_focus_frame does with the core focus, and do
the equivalent with the XInput 2 focus.
* src/comp.c (Fcomp_el_to_eln_filename): Update error message. (Bug#55861)
[ According to the Git metadata, this commit 8436e0bee9 was already
merged from `emacs-28`, yet the code says it was not. :-( ]
* src/xterm.c (xi_focus_handle_for_device): Clear implicit focus
along with FocusOut. (bug#57468)
(x_mouse_leave): Avoid invalid reads of
dpyinfo->x_focus_event_frame on input extension builds.
* src/xterm.c (handle_one_xevent): Ignore core crossing events
on input extension builds. The X server is not actually
supposed to deliver them to us, and it messes up MPX focus
tracking. (bug#57468)
* src/character.c (Fmax_char): Accept an optional argument
UNICODE, and, if non-nil, return the maximum codepoint defined by
Unicode.
* lisp/emacs-lisp/comp.el (comp-known-type-specifiers): Update the
signature of 'max-char'.
* etc/NEWS:
* doc/lispref/nonascii.texi (Character Codes): Update the
documentation of 'max-char'.
* src/xfaces.c (HANDLE_INVALID_NIL_VALUE): New macro, which displays
a warning for invalid uses of nil as a face attribute value.
(Finternal_set_lisp_face_attribute): Use the macro for the attributes
:foreground, :distant-foreground and :background.
* src/xfns.c (x_window, Fx_create_frame): Handle X errors while
trying to reparent the frame onto an explicitly specified
parent.
* src/xterm.c (handle_one_xevent): Handle unparenting embedded
windows correctly. This only works if the embedder is aware of
the fixes extension and has put the window into the right save
set.
(x_embed_frame): New function.
* src/xterm.h (FRAME_X_EMBEDDED_P): Fix coding style.
8807a4f532 Enable 256 colors in fbterm.
bbc23ca8ab One more fix for find-file.el
0cbcaf5119 ; Yet another doc fix for 'ff-other-file-alist' (bug#57325)
2d9a391727 Recommend using fbterm in the Linux console.
a5a92e577c ; * src/window.c (syms_of_window) <window-point-insertion-...
4d52fe6b96 * lisp/find-file.el (ff-other-file-alist): Doc fix. (Bug#...
d111b5b651 * lisp/info.el (Info-mode): Support the Linux console better.
cc945ec0ed ; Fix doc strings in latin1-disp.el
195fcc932e Improve the documentation of glyphless-character display
72ae02377e ; * lisp/jit-lock.el (jit-lock-chunk-size): Doc fix.
3fb69fad49 Fix documentation of 'glyphless-char-display'
# Conflicts:
# lisp/international/characters.el
# src/xdisp.c
* src/widget.c (update_wm_hints): Force a resource to change
every time this function is called. (bug#57475)
(EmacsFrameInitialize): Initialize new field.
* src/widgetprv.h (EmacsFramePart): New field `size_switch'.
* src/xterm.c (x_wm_set_size_hint): Don't change flags if flags
is 0.
* src/xselect.c (receive_incremental_selection): New arg
REAL_BYTES_RET. Set it to the actual size instead of using the
size of the array after it was grown by xpalloc.
(x_get_window_property_as_lisp_data): Adjust call to
receive_incremental_selection.
Reported by Tom Gillespie; error found by Greg Hendershott.
* src/lread.c (read_stack_reset): New.
(read0): Reset read stack when unwinding from an error.
* src/xterm.c (handle_one_xevent): Always round values as
recommended, instead of just truncating them. Also fixes
weird "jitter" during drag-and-drop on some clients.
* lisp/international/characters.el (glyphless-char-display-control):
* src/xdisp.c (syms_of_xdisp) <glyphless-char-display>: Mention
the 'glyphless-char' face in the doc string.
* doc/lispref/display.texi (Glyphless Chars): Index
'glyphless-char' face.
* lisp/simple.el (line-move, line-move-finish):
Use truncated-partial-width-window-p.
* lisp/window.el (count-screen-lines, scroll-command--goto-goal-column):
Use truncated-partial-width-window-p.
(truncated-partial-width-window-p): Replace window-width with
window-total-width.
* src/indent.c (scan_for_column): Bring the logic of using
truncated-partial-width-window-p closer to what the display engine does.
This restores the code we had in realize_gui_face before commit bf0d3f7.
The problem described in bug#17973, which led to that commit, only
happens if one uses a specific (misc-fixed) font family, not for
the usual default fonts used by Emacs, and I'm not sure what's
described there is a bug at all. At least for the purposes of
text-scale-adjust, it makes no sense to ignore the
foundry/family/adstyle of the original font, because we _want_ the
same (or very similar) font, just of a different size.
And likely in other use cases: if the :font attribute of a face
specifies some font properties, we want to keep them all, not
arbitrarily to ignore some of them. And definitely catering to an
obscure use case such as the one cited in bug#17973 is NOT a good
reason to make such radical changes in face-realization behavior.
So I think backing out that part of commit bf0d3f7 is TRT, and if
we decide that this causes bug#17973 in too many situations we care
about, I'd rather find a kludge for that specific case than do that
for every face realization.
* src/xfaces.c (realize_gui_face): Preserve face attributes when
text is scaled. This reverts part of the changes installed in
commit bf0d3f7. (Bug#37473)
* src/nsterm.m (ns_draw_glyphless_glyph_string_foreground):
* src/pgtkterm.c (pgtk_draw_glyphless_glyph_string_foreground):
* src/haikuterm.c (haiku_draw_glyphless_glyph_string_foreground):
* src/xterm.c (x_draw_glyphless_glyph_string_foreground):
* src/w32term.c (w32_draw_glyphless_glyph_string_foreground):
* src/xdisp.c (lookup_glyphless_char_display): Handle extra-slot
of 'glyphless-char-display' that is a cons cell.
(syms_of_xdisp) <glyphless-char-display>: Update doc string.
* etc/NEWS:
* doc/lispref/display.texi (Glyphless Chars): Document the new
feature.
* lisp/faces.el (glyphless-char): Make the face stand out on TTY
frames that don't support the underline attribute.
* src/xmenu.c (x_menu_translate_generic_event, x_menu_show):
Pass through more events, correctly.
* src/xterm.c (handle_one_xevent): Don't abort if must_free_data
and xi_event is NULL; this is an Xlib bug.
* src/xterm.c (x_sync_trigger_fence): Restore old code, since that
makes it easier to adapt to more than 2 sync fences (which might be
necessary in the future, if we ever decide to do frame handling
asynchronously), and is much more recognizable for people coming from
other programs.
Also, port to platforms lacking CLOCK_MONOTONIC and int64_t, and
use 0 more consistently to represent missing timestamps.
* src/xterm.h (struct x_display_info):
Omit server_time_monotonic_p and server_time_offset if
!HAVE_CLOCK_GETTIME since they are unused in that case.
* src/xterm.h (struct x_display_info, struct x_output):
* src/xterm.c (x_sync_get_monotonic_time)
(x_sync_current_monotonic_time, x_sync_note_frame_times):
Use int_fast64_t instead of int64_t as POSIX doesn't
guarantee the latter. Similarly for uint_fast64_t.
(x_sync_get_monotonic_time, x_sync_current_monotonic_time)
(x_sync_note_frame_times, x_display_set_last_user_time):
Check for integer overflow in time arithmetic.
(CLOCK_MONOTONIC): Define to CLOCK_REALTIME if absent.
(x_sync_current_monotonic_time): Check for clock_gettime failure
and fall back on CLOCK_REALTIME if CLOCK_MONOTONIC does not work,
which POSIX allows.
(x_sync_current_monotonic_time, x_sync_note_frame_times)
(x_display_set_last_user_time):
Use 0 more consistently to represent missing timestamps.
* src/dispextern.h (reset_mouse_highlight): Fix coding style.
* src/xterm.c (xi_position_changed): New functions.
(xi_report_motion_window_clear, handle_one_xevent): Don't report
motion events if the pixel position did not actually change.
* src/xterm.h (struct xi_device_t): New fields
`last_motion_window', `last_motion_x' and `last_motion_y'.
* src/xterm.c (handle_one_xevent): If a window manager sends us
FocusIn and FocusOut with XSendEvent (they do that), don't
detect focus changes when XInput 2 is enabled. The X server
will tell us the actual truth if the focus really did change.
* src/buffer.h (struct buffer_text): New field.
(BUF_CHARS_UNCHANGED_MODIFIED, CHARS_UNCHANGED_MODIFIED): New macros.
* src/buffer.c (Fget_buffer_create): Initialize the new field.
* src/xdisp.c (mark_window_display_accurate_1): Set the new field.
(redisplay_window): Use it, together with CHARS_MODIFF, instead of
MODIFF and UNCHANGED_MODIFIED to decide whether to check for long
lines.
* src/pdumper.c (dump_buffer): Dump the new field.
