* bug#73552: obsolete (and broken) face attribute :reverse-video
@ 2024-09-29 10:49 Mattias Engdegård
2024-09-29 11:31 ` Gerd Möllmann
2024-09-29 13:55 ` Stefan Kangas
0 siblings, 2 replies; 10+ messages in thread
From: Mattias Engdegård @ 2024-09-29 10:49 UTC (permalink / raw)
To: 73552; +Cc: Gerd Möllmann, Eli Zaretskii, Stefan Monnier
The face attribute keyword :reverse-video has been broken in defface for a long time, and it's not used very much, so I added a compiler warning and a notice to NEWS.
This keyword along with :bold and :italic have been obsolete for longer than that and aren't even mentioned in any documentation (on purpose) but in the internal face-and-display machinery they still partly work.
I'm not quite sure why :reverse-video doesn't work in defface. Demo, in *scratch*:
(defface myface '((t :reverse-video t)) "my face")
(insert (propertize "bon bon" 'font-lock-face 'myface) ?\n)
=> bon bon ; not inverted
(insert (propertize "non non" 'font-lock-face '(:reverse-video t)) ?\n)
=> non non ; inverted
I first blamed a simple 14 years old mistake in `custom-fix-face-spec` but fixing that doesn't help, not sure why. But it means that we probably don't need to fix it, and indeed there are precious few packages using :reverse-video. One is make-mode.el; outside Emacs, I only found one (yaml-mode).
While :bold and :italic are obsolete as well, they are used a lot more (and actually seem to work) so I'm not warning about them now.
To summarise:
- Should we fix `custom-fix-face-spec`? Maybe, or just stop pretending that it handles :reverse-video.
- Should we bother to fix what other mechanism preventing :reverse-video from working? Probably no.
- Should we remove :reverse-video from the display machinery? Yes, but maybe not now?
- Should we remove :bold and :italic? Yes, but definitely not now.
^ permalink raw reply [flat|nested] 10+ messages in thread
* bug#73552: obsolete (and broken) face attribute :reverse-video
2024-09-29 10:49 bug#73552: obsolete (and broken) face attribute :reverse-video Mattias Engdegård
@ 2024-09-29 11:31 ` Gerd Möllmann
2024-09-29 12:35 ` Mattias Engdegård
2024-09-29 13:55 ` Stefan Kangas
1 sibling, 1 reply; 10+ messages in thread
From: Gerd Möllmann @ 2024-09-29 11:31 UTC (permalink / raw)
To: Mattias Engdegård; +Cc: 73552, eliz, monnier
Mattias Engdegård <mattias.engdegard@gmail.com> writes:
> The face attribute keyword :reverse-video has been broken in defface
> for a long time, and it's not used very much, so I added a compiler
> warning and a notice to NEWS.
>
> This keyword along with :bold and :italic have been obsolete for
> longer than that and aren't even mentioned in any documentation (on
> purpose) but in the internal face-and-display machinery they still
> partly work.
>
> I'm not quite sure why :reverse-video doesn't work in defface. Demo, in *scratch*:
>
> (defface myface '((t :reverse-video t)) "my face")
>
> (insert (propertize "bon bon" 'font-lock-face 'myface) ?\n)
> => bon bon ; not inverted
>
> (insert (propertize "non non" 'font-lock-face '(:reverse-video t)) ?\n)
> => non non ; inverted
>
> I first blamed a simple 14 years old mistake in `custom-fix-face-spec`
> but fixing that doesn't help, not sure why. But it means that we
> probably don't need to fix it, and indeed there are precious few
> packages using :reverse-video. One is make-mode.el; outside Emacs, I
> only found one (yaml-mode).
>
> While :bold and :italic are obsolete as well, they are used a lot more
> (and actually seem to work) so I'm not warning about them now.
>
> To summarise:
> - Should we fix `custom-fix-face-spec`? Maybe, or just stop pretending that it handles :reverse-video.
> - Should we bother to fix what other mechanism preventing :reverse-video from working? Probably no.
> - Should we remove :reverse-video from the display machinery? Yes, but maybe not now?
> - Should we remove :bold and :italic? Yes, but definitely not now.
FWIW I'm pretty sure that I added :italic, :bold, :reverse-video to what
later became Emacs 21 only for backwards compatibility with the old
pre-21 face implementation. At least on the surface, because 21 faces
work completely differently. (I also seem to remeber that reverse-video
was a synonym for inverse-video in the old code, maybe even
undocumented. Too lazy to look that up ATM, though :-).)
^ permalink raw reply [flat|nested] 10+ messages in thread
* bug#73552: obsolete (and broken) face attribute :reverse-video
2024-09-29 11:31 ` Gerd Möllmann
@ 2024-09-29 12:35 ` Mattias Engdegård
2024-09-29 12:55 ` Eli Zaretskii
0 siblings, 1 reply; 10+ messages in thread
From: Mattias Engdegård @ 2024-09-29 12:35 UTC (permalink / raw)
To: Gerd Möllmann; +Cc: 73552, eliz, monnier
29 sep. 2024 kl. 13.31 skrev Gerd Möllmann <gerd.moellmann@gmail.com>:
>> I first blamed a simple 14 years old mistake in `custom-fix-face-spec`
>> but fixing that doesn't help, not sure why.
Having said that I naturally had to find the bug and it appears to be `face-spec-set-2` which only sets attributes that are present in `face-x-resources` for some reason.
> I'm pretty sure that I added :italic, :bold, :reverse-video to what
> later became Emacs 21 only for backwards compatibility with the old
> pre-21 face implementation. At least on the surface, because 21 faces
> work completely differently. (I also seem to remeber that reverse-video
> was a synonym for inverse-video in the old code, maybe even
> undocumented.
Thanks. Given Gerd's testimony, the lack of use and general brokenness, I propose we just give up on :reverse-video, perhaps removing it from xfaces.c.
^ permalink raw reply [flat|nested] 10+ messages in thread
* bug#73552: obsolete (and broken) face attribute :reverse-video
2024-09-29 12:35 ` Mattias Engdegård
@ 2024-09-29 12:55 ` Eli Zaretskii
2024-09-29 13:41 ` Mattias Engdegård
0 siblings, 1 reply; 10+ messages in thread
From: Eli Zaretskii @ 2024-09-29 12:55 UTC (permalink / raw)
To: Mattias Engdegård; +Cc: gerd.moellmann, 73552, monnier
> From: Mattias Engdegård <mattias.engdegard@gmail.com>
> Date: Sun, 29 Sep 2024 14:35:01 +0200
> Cc: Emacs Bug Report <bug-gnu-emacs@gnu.org>,
> Eli Zaretskii <eliz@gnu.org>,
> Stefan Monnier <monnier@iro.umontreal.ca>
>
> 29 sep. 2024 kl. 13.31 skrev Gerd Möllmann <gerd.moellmann@gmail.com>:
>
> >> I first blamed a simple 14 years old mistake in `custom-fix-face-spec`
> >> but fixing that doesn't help, not sure why.
>
> Having said that I naturally had to find the bug and it appears to be `face-spec-set-2` which only sets attributes that are present in `face-x-resources` for some reason.
>
> > I'm pretty sure that I added :italic, :bold, :reverse-video to what
> > later became Emacs 21 only for backwards compatibility with the old
> > pre-21 face implementation. At least on the surface, because 21 faces
> > work completely differently. (I also seem to remeber that reverse-video
> > was a synonym for inverse-video in the old code, maybe even
> > undocumented.
>
> Thanks. Given Gerd's testimony, the lack of use and general brokenness, I propose we just give up on :reverse-video, perhaps removing it from xfaces.c.
I'm not sure I understand the proposal: we remove :reverse-video, but
keep :inverse-video (which _is_ included in face-x-resources)? Or did
you mean to remove :inverse-video as well?
^ permalink raw reply [flat|nested] 10+ messages in thread
* bug#73552: obsolete (and broken) face attribute :reverse-video
2024-09-29 12:55 ` Eli Zaretskii
@ 2024-09-29 13:41 ` Mattias Engdegård
2024-09-29 13:49 ` Eli Zaretskii
0 siblings, 1 reply; 10+ messages in thread
From: Mattias Engdegård @ 2024-09-29 13:41 UTC (permalink / raw)
To: Eli Zaretskii; +Cc: gerd.moellmann, 73552, monnier
29 sep. 2024 kl. 14.55 skrev Eli Zaretskii <eliz@gnu.org>:
> I'm not sure I understand the proposal: we remove :reverse-video, but
> keep :inverse-video (which _is_ included in face-x-resources)? Or did
> you mean to remove :inverse-video as well?
Leave face-x-resources as it is. Keep :inverse-video.
Remove all references to :reverse-video everywhere (except the newly introduced warning).
^ permalink raw reply [flat|nested] 10+ messages in thread
* bug#73552: obsolete (and broken) face attribute :reverse-video
2024-09-29 13:41 ` Mattias Engdegård
@ 2024-09-29 13:49 ` Eli Zaretskii
0 siblings, 0 replies; 10+ messages in thread
From: Eli Zaretskii @ 2024-09-29 13:49 UTC (permalink / raw)
To: Mattias Engdegård; +Cc: gerd.moellmann, 73552, monnier
> From: Mattias Engdegård <mattias.engdegard@gmail.com>
> Date: Sun, 29 Sep 2024 15:41:53 +0200
> Cc: gerd.moellmann@gmail.com,
> bug-gnu-emacs@gnu.org,
> monnier@iro.umontreal.ca
>
> 29 sep. 2024 kl. 14.55 skrev Eli Zaretskii <eliz@gnu.org>:
>
> > I'm not sure I understand the proposal: we remove :reverse-video, but
> > keep :inverse-video (which _is_ included in face-x-resources)? Or did
> > you mean to remove :inverse-video as well?
>
> Leave face-x-resources as it is. Keep :inverse-video.
> Remove all references to :reverse-video everywhere (except the newly introduced warning).
I think the use of :reverse-video in make-mode.el should be changed to
:inverse-video. Otherwise, I'm okay with retiring this old and
half-supported alias.
^ permalink raw reply [flat|nested] 10+ messages in thread
* bug#73552: obsolete (and broken) face attribute :reverse-video
2024-09-29 10:49 bug#73552: obsolete (and broken) face attribute :reverse-video Mattias Engdegård
2024-09-29 11:31 ` Gerd Möllmann
@ 2024-09-29 13:55 ` Stefan Kangas
2024-09-29 16:22 ` Mattias Engdegård
1 sibling, 1 reply; 10+ messages in thread
From: Stefan Kangas @ 2024-09-29 13:55 UTC (permalink / raw)
To: Mattias Engdegård, 73552
Cc: Gerd Möllmann, Eli Zaretskii, Stefan Monnier
Mattias Engdegård <mattias.engdegard@gmail.com> writes:
> I first blamed a simple 14 years old mistake in `custom-fix-face-spec`
> but fixing that doesn't help, not sure why. But it means that we
> probably don't need to fix it, and indeed there are precious few
> packages using :reverse-video. One is make-mode.el; outside Emacs, I
> only found one (yaml-mode).
I suggest fixing this in make-mode.el to avoid the warning.
> While :bold and :italic are obsolete as well, they are used a lot more
> (and actually seem to work) so I'm not warning about them now.
We could start by doing s/:bold t/:weight bold/ and
s/:italic t/:slant italic/ in our tree.
> To summarise:
> - Should we fix `custom-fix-face-spec`? Maybe, or just stop pretending that it handles :reverse-video.
> - Should we bother to fix what other mechanism preventing :reverse-video from working? Probably no.
> - Should we remove :reverse-video from the display machinery? Yes, but maybe not now?
It doesn't seem worth fixing :reverse-video at this point, indeed.
I suggest removing it, given that it's been broken for so long and that
we now have a warning instead.
> - Should we remove :bold and :italic? Yes, but definitely not now.
I'd propose adding obsoletion warnings for :bold and :italic. I note
that they are not documented in the elisp manual.
^ permalink raw reply [flat|nested] 10+ messages in thread
* bug#73552: obsolete (and broken) face attribute :reverse-video
2024-09-29 13:55 ` Stefan Kangas
@ 2024-09-29 16:22 ` Mattias Engdegård
2024-09-29 16:42 ` Eli Zaretskii
0 siblings, 1 reply; 10+ messages in thread
From: Mattias Engdegård @ 2024-09-29 16:22 UTC (permalink / raw)
To: Stefan Kangas
Cc: Gerd Möllmann, Eli Zaretskii, Stefan Monnier, 73552-done
29 sep. 2024 kl. 15.49 skrev Eli Zaretskii <eliz@gnu.org>:
> I think the use of :reverse-video in make-mode.el should be changed to
> :inverse-video.
Oh, certainly (compiling that file now yields a warning).
The few extant uses of :reverse-video discovered so far seem to be fall-back cases for non-colour displays, decidedly a rarity these days.
> Otherwise, I'm okay with retiring this old and
> half-supported alias.
Thank you, now done.
29 sep. 2024 kl. 15.55 skrev Stefan Kangas <stefankangas@gmail.com>:
> We could start by doing s/:bold t/:weight bold/ and
> s/:italic t/:slant italic/ in our tree.
Yes, maybe later. At least these still work, and are widely used by external packages.
> I'd propose adding obsoletion warnings for :bold and :italic. I note
> that they are not documented in the elisp manual.
Wr probably should but the compilation warnings only catch constant arguments to `defface`. Other uses (themes, font-lock, computed text properties, non-constant arguments to defface, etc) aren't caught; that would require a run-time warning.
Anyway, we're done with :reverse-video, thanks everyone!
^ permalink raw reply [flat|nested] 10+ messages in thread
* bug#73552: obsolete (and broken) face attribute :reverse-video
2024-09-29 16:22 ` Mattias Engdegård
@ 2024-09-29 16:42 ` Eli Zaretskii
2024-09-29 17:25 ` Mattias Engdegård
0 siblings, 1 reply; 10+ messages in thread
From: Eli Zaretskii @ 2024-09-29 16:42 UTC (permalink / raw)
To: Mattias Engdegård; +Cc: gerd.moellmann, stefankangas, monnier, 73552
> From: Mattias Engdegård <mattias.engdegard@gmail.com>
> Date: Sun, 29 Sep 2024 18:22:42 +0200
> Cc: 73552-done@debbugs.gnu.org,
> Gerd Möllmann <gerd.moellmann@gmail.com>,
> Eli Zaretskii <eliz@gnu.org>,
> Stefan Monnier <monnier@iro.umontreal.ca>
>
> 29 sep. 2024 kl. 15.49 skrev Eli Zaretskii <eliz@gnu.org>:
>
> > Otherwise, I'm okay with retiring this old and
> > half-supported alias.
>
> Thank you, now done.
Thanks. Should the warning in bytecomp.el be reworded to the effect
that :reverse-video is no longer supported?
^ permalink raw reply [flat|nested] 10+ messages in thread
end of thread, other threads:[~2024-09-29 17:25 UTC | newest]
Thread overview: 10+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2024-09-29 10:49 bug#73552: obsolete (and broken) face attribute :reverse-video Mattias Engdegård
2024-09-29 11:31 ` Gerd Möllmann
2024-09-29 12:35 ` Mattias Engdegård
2024-09-29 12:55 ` Eli Zaretskii
2024-09-29 13:41 ` Mattias Engdegård
2024-09-29 13:49 ` Eli Zaretskii
2024-09-29 13:55 ` Stefan Kangas
2024-09-29 16:22 ` Mattias Engdegård
2024-09-29 16:42 ` Eli Zaretskii
2024-09-29 17:25 ` Mattias Engdegård
Code repositories for project(s) associated with this public inbox
https://git.savannah.gnu.org/cgit/emacs.git
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).