9:58 p.m.
Misc fixes for things.
andy
1999-10-14 Andy Piper <andy(a)xemacs.org>
* event-msw.c (mswindows_key_to_emacs_keysym): add some more
keycodes.
* redisplay-x.c (x_output_shadows): fix dodgy maths for border
calculations.
* gutter.c (output_gutter): be more accurate about the area to be
cleared since X seems to manage to do the clear after drawing the
border.
* redisplay.h (RESET_CHANGED_FLAGS): new macro for setting
redisplay flags as a group.
(RESET_CHANGED_SET_FLAGS): ditto.
(CLASS_RESET_CHANGED_FLAGS): ditto.
(GLOBAL_RESET_CHANGED_FLAGS): ditto.
(REDISPLAY_FLAGS_CHANGEDP): new macro for testing redisplay flags
as a group.
(CLASS_REDISPLAY_FLAGS_CHANGEDP): ditto.
(GLOBAL_REDISPLAY_FLAGS_CHANGEDP): ditto.
* redisplay.c (redisplay_frame): use CLASS_RESET_CHANGED_FLAGS
instead of setting flags individually.
(redisplay_device): ditto CLASS_REDISPLAY_FLAGS_CHANGEDP.
(redisplay_device): ditto.
(redisplay_device): ditto CLASS_RESET_CHANGED_FLAGS.
(redisplay_without_hooks): ditto GLOBAL_REDISPLAY_FLAGS_CHANGEDP.
(redisplay_without_hooks): ditto CLASS_REDISPLAY_FLAGS_CHANGEDP.
(redisplay_without_hooks): ditto.
(redisplay_without_hooks): ditto GLOBAL_RESET_CHANGED_FLAGS.
* redisplay-x.c (x_output_string): unmap subwindows in the area.
(x_output_blank): ditto.
* redisplay-output.c (redisplay_output_display_block): don't unmap
subwindows in case layout optimization can avoid it.
(redisplay_output_subwindow): unmap subwindows in the area.
(redisplay_output_layout): optimize the output of layouts - only
output glyphs that have changed if nothing else of redisplay
significance has occurred.
(redisplay_output_pixmap): change args to
redisplay_clear_clipped_region.
(redisplay_clear_clipped_region): allow the clipped region to be
cleared of all subwindows except the one passed in.
* redisplay-msw.c (mswindows_output_blank): unmap subwindows in
the area.
(mswindows_output_cursor): ditto.
(mswindows_output_string): ditto.
* gutter.c (calculate_gutter_size): if the window buffer is nil
then don't continue.
(update_frame_gutters): be more lenient about when we actually
update the gutter. Layout optimization makes this
non-costly. Cache redisplay flags that we want to temporarily
ignore.
* glyphs.c (Fglyph_animated_timeout_handler): handle image
specific timeouts rather than iterating over the instance cache.
(glyph_animated_timeout_mapper): deleted.
(add_glyph_animated_timeout): use a weak list to hold onto the
image so that it can be GC'ed.
(disable_glyph_animated_timeout): disable a specific timeout.
(vars_of_glyphs): disable-animated-pixmaps is a new boolean for
controlling whether pixmaps are animated or not.
* glyphs-msw.c (mswindows_finalize_image_instance): make sure the
image timeout gets disabled when the image gets freed.
* glyphs-eimage.c (gif_instantiate): remove meaningless
comment. Cope with timeouts specified in the gif extension block.
* event-msw.c (vars_of_event_mswindows): new variable
mswindows-meta-activates-menu.
(mswindows_wnd_proc): only goto defproc with VK_MENU if the user
wants it.
* glyphs-x.c (x_finalize_image_instance): delete mask first so
that we can compare with image.
1999-10-13 Andy Piper <andy(a)xemacs.org>
* gutter-items.el (progress-abort-glyph): new glyph for showing
abort status.
(append-progress): dispatch-event rather than sit-for.
(abort-progress): new function. Show the abort glyph with an
appropriate message.
(raw-append-progress): dispatch-event rather than sit-for.
(display-progress): cope with aborts.
* gui.el (make-gui-button): Use native widgets for buttons
unconditionally.
(insert-gui-button): ditto.
(gui-button-p): ditto.
* xbm-button.el: remove from core.
* xpm-button.el: remove from core.
--------------------------------------------------------------
Dr Andy Piper
Senior Consultant Architect, BEA Systems Ltd
4:30 a.m.
A few comments:
First, a comment like this:
+ /* We bogusly don't take f->extents_changed and f->glyphs_changed
+ into account. This is because if we do we always redisplay the
+ entire layout. So far I have seen no ill effects to we'll see. */
should always have a #### next to it to indicate a potential problem.
Also, the "progress" functions are too generically-named. For namespace
cleanliness, you should do something like
s/lprogress/labelled-progress/
then
s/progress/progress-gauge/g
or even
s/progress/gutter-progress-gauge/g
because it's quite possible that "regular" progress gauges (i.e. displayed
in
their own dialog box) will also be implemented.
Also, expressions like
(vector ...)
and
(list ...)
can be made much clearer using
`[ ... ]
and
`( ... )
For example, instead of
(vector
'layout :orientation 'vertical :justify 'left
:items (list
progress-text-glyph
(make-glyph
(vector
'layout :pixel-height progress-glyph-height
:orientation 'horizontal
:items (list
progress-gauge-glyph
(vector
'button :pixel-height (- progress-glyph-height 8)
:descriptor " Stop "
:callback '(funcall progress-stop-callback)))))))))
use
`[layout
:orientation vertical
:justify left
:items (,progress-text-glyph
[layout
:pixel-height ,progress-glyph-height
:orientation horizontal
:items (,progress-gauge-glyph
[button
:pixel-height ,(- progress-glyph-height 8)
:descriptor " Stop "
:callback (funcall progress-stop-callback)])])]
where I omitted the make-glyph call inside because it's apparently not
necessary.
ben
Andy Piper wrote:
Misc fixes for things.
andy
1999-10-14 Andy Piper <andy(a)xemacs.org>
* event-msw.c (mswindows_key_to_emacs_keysym): add some more
keycodes.
* redisplay-x.c (x_output_shadows): fix dodgy maths for border
calculations.
* gutter.c (output_gutter): be more accurate about the area to be
cleared since X seems to manage to do the clear after drawing the
border.
* redisplay.h (RESET_CHANGED_FLAGS): new macro for setting
redisplay flags as a group.
(RESET_CHANGED_SET_FLAGS): ditto.
(CLASS_RESET_CHANGED_FLAGS): ditto.
(GLOBAL_RESET_CHANGED_FLAGS): ditto.
(REDISPLAY_FLAGS_CHANGEDP): new macro for testing redisplay flags
as a group.
(CLASS_REDISPLAY_FLAGS_CHANGEDP): ditto.
(GLOBAL_REDISPLAY_FLAGS_CHANGEDP): ditto.
* redisplay.c (redisplay_frame): use CLASS_RESET_CHANGED_FLAGS
instead of setting flags individually.
(redisplay_device): ditto CLASS_REDISPLAY_FLAGS_CHANGEDP.
(redisplay_device): ditto.
(redisplay_device): ditto CLASS_RESET_CHANGED_FLAGS.
(redisplay_without_hooks): ditto GLOBAL_REDISPLAY_FLAGS_CHANGEDP.
(redisplay_without_hooks): ditto CLASS_REDISPLAY_FLAGS_CHANGEDP.
(redisplay_without_hooks): ditto.
(redisplay_without_hooks): ditto GLOBAL_RESET_CHANGED_FLAGS.
* redisplay-x.c (x_output_string): unmap subwindows in the area.
(x_output_blank): ditto.
* redisplay-output.c (redisplay_output_display_block): don't unmap
subwindows in case layout optimization can avoid it.
(redisplay_output_subwindow): unmap subwindows in the area.
(redisplay_output_layout): optimize the output of layouts - only
output glyphs that have changed if nothing else of redisplay
significance has occurred.
(redisplay_output_pixmap): change args to
redisplay_clear_clipped_region.
(redisplay_clear_clipped_region): allow the clipped region to be
cleared of all subwindows except the one passed in.
* redisplay-msw.c (mswindows_output_blank): unmap subwindows in
the area.
(mswindows_output_cursor): ditto.
(mswindows_output_string): ditto.
* gutter.c (calculate_gutter_size): if the window buffer is nil
then don't continue.
(update_frame_gutters): be more lenient about when we actually
update the gutter. Layout optimization makes this
non-costly. Cache redisplay flags that we want to temporarily
ignore.
* glyphs.c (Fglyph_animated_timeout_handler): handle image
specific timeouts rather than iterating over the instance cache.
(glyph_animated_timeout_mapper): deleted.
(add_glyph_animated_timeout): use a weak list to hold onto the
image so that it can be GC'ed.
(disable_glyph_animated_timeout): disable a specific timeout.
(vars_of_glyphs): disable-animated-pixmaps is a new boolean for
controlling whether pixmaps are animated or not.
* glyphs-msw.c (mswindows_finalize_image_instance): make sure the
image timeout gets disabled when the image gets freed.
* glyphs-eimage.c (gif_instantiate): remove meaningless
comment. Cope with timeouts specified in the gif extension block.
* event-msw.c (vars_of_event_mswindows): new variable
mswindows-meta-activates-menu.
(mswindows_wnd_proc): only goto defproc with VK_MENU if the user
wants it.
* glyphs-x.c (x_finalize_image_instance): delete mask first so
that we can compare with image.
1999-10-13 Andy Piper <andy(a)xemacs.org>
* gutter-items.el (progress-abort-glyph): new glyph for showing
abort status.
(append-progress): dispatch-event rather than sit-for.
(abort-progress): new function. Show the abort glyph with an
appropriate message.
(raw-append-progress): dispatch-event rather than sit-for.
(display-progress): cope with aborts.
* gui.el (make-gui-button): Use native widgets for buttons
unconditionally.
(insert-gui-button): ditto.
(gui-button-p): ditto.
* xbm-button.el: remove from core.
* xpm-button.el: remove from core.
------------------------------------------------------------------------
Name: miscstuff.patch
miscstuff.patch Type: Plain Text (text/plain)
Encoding: quoted-printable
------------------------------------------------------------------------
--------------------------------------------------------------
Dr Andy Piper
Senior Consultant Architect, BEA Systems Ltd
7:02 p.m.
At 08:30 PM 10/15/99 -0700, Ben Wing wrote:
should always have a #### next to it to indicate a potential problem.
Ok.
Also, the "progress" functions are too generically-named.
For
That's the point. If there becomes a better way of doing it then I don't
want to have to change code. The tty default is to show progress in the
echo area.
`[ ... ]
Yes. I was never sure of the , syntax so just stuck to what I knew worked.
where I omitted the make-glyph call inside because it's apparently
not
necessary.
Correct. The image normalization now includes automatic creation of things
that are obviously glyphs.
andy
--------------------------------------------------------------
Dr Andy Piper
Senior Consultant Architect, BEA Systems Ltd
12:07 a.m.
Regardless, a new function named "progress" is too generic. Since your intent
is always to display progress gauges, regardless of implementation, you should
change "progress" to "progress-gauge". Also, there is no need to
completely
follow the "message" paradigm of function naming -- functions that perform
actions (as opposed to those that just return values) should always have a
verb in them, so a function just called "progress-gauge" is not what you want
-- something like "show-progress-gauge-fmt" if you want to distinguish
printf-like and non-printf-like functions.
Andy Piper wrote:
At 08:30 PM 10/15/99 -0700, Ben Wing wrote:
>Also, the "progress" functions are too generically-named. For
That's the point. If there becomes a better way of doing it then I don't
want to have to change code. The tty default is to show progress in the
echo area.
9172
days inactive
9174
days old
3 comments
2 participants
participants (2)
-
Andy Piper -
Ben Wing