When looking for the most up-to-date link to grab a scihub url
we used to use wikiless but now simply directly scrape from
wikipedia.
Functionality is not changed but should work much faster and
more reliably.
The redirector can now be fed with either a list of targets to pick from
(as before) or a "farside" entry which points to a farside service
redirect or multiple.
The dict entry then looks like:
```python
"quora": {
"source": ["quora.com"],
"farside": ["quetre"],
}
```
It has the source service url as usual, but then instead of a "target"
entry contains a "farside" entry.
A redirect dict entry can take both "target" and "farside" entries, but
will then always give precedence to farside choices.
This should go a long way in helping keep OSS frontends up to date. It
now mostly depends on farside entries in turn being updated.
One potential concern is the new centralization in relying on farside as
the redirect authority.
These are the redirects that made the most trouble for me, and honestly I
do not think wikipedia really belongs in the same category of 'necessary
to redirect' surveillance pages like reddit,instagram or twitter.
Updated reddit redirects once again.
Will soon have to figure out a different way of
approaching the lib- redirections with so many
services blocking/throttling/shutting down the
FOSS frontends. For now, I will bear with it.
Simple wrapper for xdg-open functionality. Simply refers
to xdg-open except if there exists mimeo on the system
which it will refer to instead.
So, a simple preference modificator for mimeo over
xdg-open since that is my preference too.
Also gave it a short name so I can do open whenever I
want and don't have to tax my left hand with tying xdg.
Since we now use dotter we can simplify the dir structure for
qutebrowser a lot. Everything dot-filed earlier can now reside
in simple directories called config (for ~/.config/qutebrowser),
data (for ~/.local/share/qutebrowser), and scripts
(for ~/.local/bin) files.
Do not allow html5 canvas reading in qutebrowser. This shuts off one of the main
avenues of browser fingerprinting (aside from cookies) still currently
undertaken. It will possibly stop some obscure streaming sites from working but
I do not care, especially at the cost of fingerprinting information.
When not showing tabs (only displaying them on actually
switchting between tabs), so far the duration that the
tabs would be displayed was just 800ms, way too short
to actually provide legible info.
This commit changes it to 2000ms, which may still be
too short but is at least more digestible.
Main linking buttons (f/F) *can* be used for this but the resulting
userscript has to account for many more different scenarios than
opening a page.
If it does not, we can not 'click' on text entry fields, dropdowns,
sliders etc with the f-buttons anymore.
This commit switches the qute-gemini script to be invoked via ,g or ,G
thus with the local leader mapping instead.
Code grabbing has been slightly refactored to always check for the
existence of the qutebrowser fifo before sending anything.
Translate has been fixed up a little to use lingva for selection
translations, only falling back to google translate for complete page
translations since I do not think any other services support this for
now.
Now possibly a little less clunky (though still using soon-to-be
deprecated modules), and able to correctly surf gemini/non-gemini pages
as a userscript or as a stand-alone script (simply calling it from
command line).
Works as before for tabbed pages (the clunky rename/symlink way).
Can now also be used as the default way to surf to links so I switched
out my qutebrowser f/F link hinting for this script. Will simply open
links if staying on http but open gemini version as local file if moving
on to gemini.
Could probably be rewritten as an actual plugin to interject itself in
link opening to be a little more elegant (similar to the redirect code I
have running to move to open source web frontends).
Both bookmarkers can now either send the current page ("s for sending to
shaarli; "w to send to wallabag) or send any link on the current page
(;s for shaarli, ;w for wallabag).
Both have been adapted to the same format, both can be sent a link as an
argument as well (./wallabag_add.sh 'myawesome.blog/entry').
Moved the qutebrowser userscript to open recently downloaded files
to be accessible as a normal shell script as well
(`recently-downloaded`).
Mapped this to Super+Shift+D in riverwm to be easily able to open
the most recent downloads from anywhere.
Double the default lines shown to 20.
Blocked hosts reside in a single (giant) plaintext file, which we
now remove from the repository.
Especially with the new blocking sources we grew from around 700k
(which was already a lot at ~35000 lines) to around 5MB which is
just stupid to keep in the git repo if we don't have to.
Since qutebrowser automatically re-creates the file for us on
running `:adblock-update` there's no reason to waste space here.
Scribe links are often not redirected correctly if belonging to medium's
'global-identity' redirections. This is a first attempt at fixing those
by removing the superfluous string from the scribe URL.
Generalized enough to work as a 'post-processing' function for the
redirection plugin, which can be set up by pointing the 'postprocess'
key of a page entry to a callable object (most likely a function).
Fixed text not flowing to the external (nvim) editor and saved text not
being brought back into qutebrowser.
Same issue as here https://github.com/qutebrowser/qutebrowser/issues/6707
it essentially amounts to the terminal not having its own running
process id which qutebrowser uses to know when the application closes.
Thus, it thinks it closes immediately and deletes the temporary file. No
changes are brought back and the file is empty for the editor.
With the fix, this does not happen anymore.
I rewrote and extended the nsxiv-rifle script recently, and this change
now: removes the old script, adds the new one as a submodule into the
repository and links a working binary into the path.
The new script is essentially still just an image grabber-and-opener,
but supports more image viewers, is a little more advanced and can for
example replace arbitrary parts of urls (and is documented a little more
nicely).
Make qutebrowser use new script and remove leftover references to
nsxiv-rifle. Qutebrowser also gets a third option of viewing images, now
there are:
,i -> open selected link in imageviewer
,I -> open current page in imageviewer
,<C-i> -> open selected image in imageviewer
There was not much structure to the shortcuts versus commands so far.
This turns most userscripts into invokable commands, which in turn are
invoked by the respective shortcut mappings. Mappings themselves have
not changed.
New commands are:
- `readable` to invoke readability script
- `save-to-pdf` (from `save_to_pdf`) to save current page as pdf
- `recent-downloads` to show list of recently downloaded files
Added script which takes you to the corresponding sci-hub entry for any
DOI. DOIs can be passed in three ways:
- via hinted link (shortcut `;p` to start hinting)
- via selected text (select text then invoke `send-to-scihub` command
with `"p`)
- or from meta tags in current page (invoke `send-to-scihub` command
with `"p` when on article page)
It will grab the newest sci-hub link and attempt to bring you the
corresponding pdf file.
Added cookie-blocker script which attempts to remove the cookie banners
popping up on a variety of websites. Works well for some of the large
ones (e.g. reddit, google, stackoverflow) but less well on smaller
pages.
Can be invoked with `:cookie-block` command alias.
Now that I have time to enjoy a game every now and again once more,
pcgamingwiki is a wonderful resource to start looking into all sorts of
compatibility issues and troubleshooting, so it is added as a
qutebrowser search engine as `pcw`.
Changed the hinting style for image viewing from rapid hinting to
normal, single link hinting.
In the future I will have to think more about image viewing
shortcuts in general:
I want to be able to open the current page, a single link (default) and
multiple links in quick succession - how to accomplish without
overloading the lleader+i mapping terribly?
Plus, ideally I would want there to be a shortcut to open both links
(hint links) and images themselves (hint images) which may require even
further mappings.
Added two mappings: localleader+i and localleader+I (localleader being ,
in my case) to open single images or galleries in nsxiv.
The maps make use of nsxiv-rifle and mirror the setup for opening video
in mpv with +m/+M where capital letters open the current page and small
ones let you follow a link.
Removed alias to invoke sending a page to outline to make it more
readable. Two reasons: First, outline seems gone. The website is hogged
by advertisements. Second, when I really need a page in readability
mode, I can just send it to my personal wallabag instance and as a bonus
even get it archived automatically.
Changed default tab visibility option to only show when multiple tabs
are open.
Fixed default tab visibility to always take effect when qutebrowser is
started and always allow simultaneous toggling to tabs and status bar
(<leader>tt)
Added wikiless as automatic redirection for any wikipedia page.
Will only work for generall `wikipedia.org` links or specifically
english-language `en.wikipedia.org` links.
I have been using wayland for about three months now and the 'missing
things' roadmap has been shrinking steadily. There are still some
little things that are annoying, most concerning old xwayland or Gnome
applications such as when I have to make use of Zotero or Element -- all
GUI applications that I would prefer not to use anyway.
Otherwise it's a very pleasant experience and *extremely* solid. I don't
remember any crashes, it feels relatively light weight, multi-monitor
works without any hick-ups and I really can't complain.
I will for the foreseeable future stay with wayland, since it works so
well for me. My riverwm setup will still need some tinkering here and
there to make use of some of its nice features and not just pretend to
be using i3 again. Lastly, I will need to peruse my settings and find
left-over x11 things that will inevitably be hiding in this repository.
All doable things and that makes me happy.
Finally made the switch from alacritty to kitty. I have been thinking
about this for a while. Both, fundamentally, serve my purposes just
fine. Both are fast, customizable, gpu accelerated, and so on.
Kitty feels a little faster on the input, but this should not provide
major differences.
One big difference, however, is now very apparent and I can feel it:
Alacritty, on wayland, does not support any picture preview. It does not
support sixel, and things like w3mimg or ueberzug are based on and
require X11 to run.
Kitty brings its own graphics display library and it seems both pretty
stable and fast.
I have not done much more with it than use it in things like vifm image
previews but it should be much more stable than things like ueberzug,
much faster than things like sixel. Time will tell.
Switched other modules to make use of kitty instead of alacritty:
vifm uses kitty previews,
river spawns kitty instances,
systemd units use kitty instances,
waybar presents extra mouse-click interactions through kitty,
and styler contains a processor to style kitty permanently.
I would love to converge this all a bit more on the `$TERMINAL` env var,
but this is unfortunately difficult for things like systemd and waybar.
For waybar I currently see no real way except for a custom
`ideal-terminal` script which just goes down the list of terminal
emulators I want to run, depending on which is installed,
since it does not read env vars,
while for systemd it might be feasible to import user environment
variables,
but also connected to additional complexity and overhead which it does
not seem worth for the currently two simple service units it affects.
Also removed some obsolete sxhkd and sh settings from the move to
wayland.
