| Age | Commit message (Collapse) | Author |
|
The previous code failed to detect some HTML emails, perhaps because
the text/html part was nested inside a multipart/alternative?
Lucked onto debbugs-gnu.el by asking xref for occurrences of
gnus-article-mime-handles in the Emacs tree; not sure if M-? showing
results from ~/.emacs.d/elpa was intended, but it sure helped here.
|
|
gnus-article-prepare-hook is _not_ run in the article buffer. No idea
why things fall into place on my desktop; probably some ill-defined
buffer-switching plumbing.
Also clarify nearby comment.
|
|
Letting visual-line-mode take care of wrapping.
|
|
By tricking shr into treating tables like regular blocks.
|
|
Since the article buffer is re-used, visual-line-mode will keep its
previous state if not explicitly set. So before the patch
1. I display article 1, which should be wrapped: the hook sets
visual-line-mode to t,
2. I display article 2, which should not be wrapped: truncate-lines is
set to t, but visual-line-mode remains t (not a biggie, truncate-lines
"wins"),
3. I decide I want to wrap article 2, so I call the visual-line-mode
toggling command: visual-line-mode is set to nil; seemingly nothing
happens.
This patch dispels the confusion.
|
|
Set shr-fill-text to nil; this automatically causes eww to enable
visual-line-mode.
For Gnus, the dance is a bit more involved. Not 100% happy with where
I landed, but at least I gave myself an escape hatch by rebinding C-c
d v in summary buffers.
Also,
* lift size restriction when fontifying citations: it mostly only ever
causes fontification failures when people review big patches,
* tweak header order slightly,
* align shr heading faces with eighters-title-* faces.
|
|
|
|
This option defaults to 1, which is sub-optimal for groups such as:
nntp+news.gmane.io:gmane.linux.suse.opensuse.devel
nntp+news.gmane.io:gmane.org.misc.sourcehut.general
nntp+news.gmane.io:gmane.comp.gcc.devel
nntp+news.gmane.io:gmane.emacs.help
nntp+news.gmane.io:gmane.emacs.bugs
nntp+news.gmane.io:gmane.emacs.devel
|
|
|
|
|
|
Can't even remember what this was about; obviously positives when
grouping threads due to subject-matching being too lax? 🤷
|
|
|
|
Lots of commands in Gnus buffers try to reinstate the setup defined in
gnus-buffer-configuration, so when eg drafting a message with the
original article displayed on the side, and using P or N to move
around the thread, there ends up being some friction in window
management.
So stop using the Group buffer as a way to left-pad articles; it
doesn't even look good on half-width frames anyway.
|
|
Make it more outline-friendly.
|
|
- clarify package initialization comments
- load Modus theme on terminal too
- use emoji variation selector when appropriate
- open patches in read-only mode
- now that Modus has been updated, ditch obsolete option
- etc.
|
|
- add [SPAM UNSURE]
- make tailing " :" optional
|
|
No specific reason, except maybe pacifying the mode-line, and making
sure my code still works.
|
|
Also move it to a dedicated variable.
|
|
|
|
Otherwise "Re: Repeat-mode" turns into "Re: peat-mode" when replying.
|
|
|
|
Keep showing group buffer on the left; that should help shove the
article away from the edge of the screen.
Also add Scandinavian abbreviation for "re:" to patterns to ignore
when determining subject changes.
|
|
|
|
By recognizing more "reply-cruft".
From a cursory glance at Emacs sources, the capture groups in the
default value are not used, so I see no need to preserve them.
|
|
NB: gnus-thread-sort-by-most-recent-{date,number} sometimes fail
because of bug#42334.
|
|
Searching for "KLG" yields fewer false positives than "me".
|
|
- Add ibuffer section for VC buffers.
- Let ERC pick up my full name.
Also, bind switch-to-buffer-obey-display-actions to make C-x 4 4 and
C-x 5 5 more reliable.
|
|
- get rid of diamonds
- don't use sender name if I'm the sender
- fix off-by-one error when humanizing dates
|
|
Sometimes, when adding dummy roots, Gnus displays the thread subject
on the first message:
(emacs-devel, fetched from archive.lwn.net)
┌◈
10:31 ├► martin rudalics Re: emacs rendering comparisson between emacs23 and emacs26.3
14:07 │├► Dmitry Gutov
14:35 ││├► Dmitry Gutov
14:45 ││╰► martin rudalics
13:47 │╰► Alan Mackenzie
14:45 ┆ ╰► martin rudalics
And sometimes it doesn't:
(bug-gnu-emacs, fetched from news.gmane.io)
┌◈
02:08 ├► Juri Linkov
05:33 ╰► Stefan Monnier
Adding %S to the dummy line-format will cause the subject to appear
twice in the first situation (on the dummy root, then on the first
message); not great, but better than not knowing what the topic is.
|
|
Frustratingly, sometimes the previous layout gave things like this:
O Yesterday 03:59 ┌◈ Richard Stallman pull requests
O Yesterday 08:54 ├► Eli Zaretskii
03:46 │├► Richard Stallman
O Yesterday 14:00 │╰► Clément Pit-Claudel
O Yesterday 15:05 │ ├► Stefan Monnier
O Yesterday 14:30 │ ╰► Eli Zaretskii
O Yesterday 15:37 │ ╰► Clément Pit-Claudel
Yesterday 16:21 │ ╰► Eli Zaretskii
Yesterday 17:39 │ ├► Clément Pit-Claudel
Yesterday 20:21 │ │╰► Eli Zaretskii
Yesterday 16:41 │ ╰► Dmitry Gutov
Yesterday 20:16 │ ├► Eli Zaretskii
Yesterday 20:24 │ │├► Dmitry Gutov
Yesterday 20:34 │ │╰► 조성빈
Yesterday 20:28 │ ├► Eli Zaretskii
Yesterday 21:39 │ │╰► Dmitry Gutov
03:46 │ ╰► Richard Stallman
O Yesterday 04:49 ╰► Stefan Monnier
03:45 ╰► Richard Stallman
04:03 ╰► Stefan Monnier
O Thu 20:31 ┌◈ Alan Mackenzie Re: emacs rendering comparisson between emacs23 and emacs26.3
O Thu 21:25 ├► Stefan Monnier
O Thu 21:38 ╰► rrandresf@gmail.com
O Yesterday 08:00 ├► Eli Zaretskii
O Yesterday 19:36 │├► andrés ramírez
03:48 │╰► Richard Stallman
08:09 │ ╰► Eli Zaretskii
10:08 │ ╰► Eli Zaretskii
O 03:40 ╰► Richard Stallman
O 10:21 ├► Michael Albinus
10:58 ╰► Eli Zaretskii
11:09 ╰► Eli Zaretskii
12:20 ╰► Ulrich Mueller
I.e. for some reason Alan's response to the "emacs rendering
comparison thread" is not formatted as a false root, and things get
misaligned.
Trying to set gnus-summary-make-false-root to 'empty caused this:
O Thu 20:31 ┌◈ Alan Mackenzie Re: emacs rendering comparisson between emacs23 and emacs26.3
O Thu 21:25 │├► Stefan Monnier
O Thu 21:38 │╰► rrandresf@gmail.com
O 03:40 │ ├► Richard Stallman
O 10:21 │ │├► Michael Albinus
10:58 │ │╰► Eli Zaretskii
11:09 │ │ ╰► Eli Zaretskii
O Yesterday 08:00 │ ╰► Eli Zaretskii
O Yesterday 19:36 │ ├► andrés ramírez
03:48 │ ╰► Richard Stallman
08:09 │ ╰► Eli Zaretskii
10:08 │ ├► Eli Zaretskii
12:43 │ ╰► Alan Mackenzie
12:20 ◈ Ulrich Mueller
Still no false root, and Ulrich's answer is misaligned.
The new settings yield this layout:
┄┐
O Thu 20:31 ├► Alan Mackenzie Re: emacs rendering comparisson between emacs23 and emacs26.3
O Thu 21:25 │├► Stefan Monnier
O Thu 21:38 │╰► rrandresf@gmail.com
O 03:40 │ ├► Richard Stallman
O 10:21 │ │├► Michael Albinus
10:58 │ │╰► Eli Zaretskii
11:09 │ │ ╰► Eli Zaretskii
O Yesterday 08:00 │ ╰► Eli Zaretskii
O Yesterday 19:36 │ ├► andrés ramírez
03:48 │ ╰► Richard Stallman
08:09 │ ╰► Eli Zaretskii
10:08 │ ├► Eli Zaretskii
12:58 │ │╰► Alan Mackenzie
12:43 │ ╰► Alan Mackenzie
13:05 │ ╰► Eli Zaretskii
13:18 │ ├► Alan Mackenzie
13:28 │ │╰► Eli Zaretskii
13:43 │ ╰► Stefan Monnier
12:20 ╰► Ulrich Mueller
It'd be nice if the dummy root held the subject line, but oh well. At
least things are aligned correctly 🙌
|
|
This way subjects are aligned, as well as the names of authors of root
articles.
Use this specific variation of WHITE DIAMOND CONTAINING BLACK …
DIAMOND because it is covered by DejaVu Sans Mono; others require
variable-width fonts.
|
|
This way roots, false roots, first-level line drawings and single
message are aligned.
|
|
Remove nnimap-stream: the default value is good enough, which is a
good thing since I had misspelled it 🤦.
Also tweak Info reference so that I can evaluate it directly.
|
|
Add %* to prevent point from jumping sideways when changing articles.
|
|
Keep false root, otherwise I cannot mark threads with T k.
Apply crude fontification with gnus-face-0.
|
|
Use box drawing characters to show article tree. Don't "adopt"
rootless subthreads as that makes articles look like they answer
unrelated messages.
|
|
AFAICT, by default Gnus threads by *root article* number, oldest
above, newest below.
I'd like to keep the "newest below" logic (messages are sorted within
a thread anyhow), while making sure threads updated recently show
up *below* older threads.
Hence:
- put gnus-thread-sort-by-number first to break ties, following the
recommendation from (gnus) Sorting the Summary Buffer,
- invert gnus-thread-sort-by-most-recent-number, to have
recently-updated threads below older threads.
Not sure this implements exactly what I want (I think I saw some
surprising results here and there), but it seems to come close.
Also note down more useful key bindings.
|
|
So that I can use NNTP to reach bug-gnu-emacs and help-gnu-emacs.
Gmane has moved recently:
https://lars.ingebrigtsen.no/2020/01/15/news-gmane-org-is-now-news-gmane-io/
|
|
|
|
|
|
- something something Sunday week-*end* something
- been using these packages for a while now
- getting used to the "most recent below" convention; it's what every
interface shows within threads anyway
|
|
Since dates are variable-width and left-aligned, and the threaded view
pushes the next info (author name) to the right, without delimiters
it was hard to tell whether an article is the OP or not.
|
|
|
|
Don't know why I thought those gnus- variables deserved to be in their
own setq.
|
|
|
|
Get rid of
- the scoring indicator
- the number of lines
- the mouse face
|
|
|
|
- fix accidentally committed Emacs 27 customization
- add a useful Gnus shortcut when dealing with drafts
|
|
On Debian Jessie at least, this is necessary.
|
|
|