This completely removes the tabspaces option from the Alacritty
configuration, due to frequent misuse of it. Based on some research,
none of the terminal emulators support setting the value for tabspaces
or read the terminfo to determine init_tabs value at startup. The tested
terminal emulators were URxvt, XTerm, and Termite.
This resolves an issue with tabs not breaking across line boundaries,
instead the characters would just all get written to the last column and
thus be lost.
It also tweaks the behavior of what happens when the terminal resizes
with the default tabspaces changed, using something like the `tabs`
program. Previously all tabstops would be reset to the default on
resize, which is what URxvt does. Now the tabspaces are kept and the new
columns are filled with the default tabstops, which emulates Termite.
Previously we were rounding pattern's `pixelsize` before `fc_sort`, however we were using not rounded one in `get_glyph`, so bitmap fonts could look a bit smaller when used in a mix with scalable fonts.
The block selection will now only insert extra newline characters after
each line if the last line isn't already included. This resolves an
issue with duplicate newlines, since newlines are automatically appended
when the last column is part of a selection.
Fixes#3304.
This resolves an issue in the parser where it would stop as soon as the
first unknown value is encountered in private mode/sgr attribute
escapes.
Fixes#3339.
The merge strategy `union` should hopefully reduce the number of
conflicts when trying to rebase the configuration file, since most of
the changes are simple additions anyways.
This reverts commit 7f4dce2ee0.
Originally it was assumed that macOS always sends the \x7f on backspace
anyways, however this is not true. It seems like the character on
backspace can change even within the same terminal session, so we need
to have our own binding to reliably set the correct binding.
A solution for #1606 should be implemented in cooperation with winit.
Instead of trying to expand the start and end of a selection across
full-width glyphs, the selection should now only go from its origin to
the end without any kind of expansion.
Instead, the expansion is now done where the cells are actually checked
for their selection status, expanding across the entire full-width glyph
whenever any part of it is selected.
Fixes#3106.
There's no reason why we should ever manually set the terminal to not be
dirty, since this can lead to a lot of other logic being affected. This
also does not have any benefit and was likely added in the event loop
rework as a bug (probably should have been dirty = true).
Previous implementation was querying Fontconfig using `charset` in a pattern,
which was leading to unpredictable fallbacks in some cases, since Fontconfig
was picking the font with the most coverage for a given charset, regardless of
user configuration. Moreover all fallback was based on font_match which is
extremely slow for such performance sensitive task as a fallback, so alacritty
had a hard times on vtebench's unicode-random-write.
The new approach is to use some internal fallback list from font_sort
and iterate over it to get a proper fallback font, since it matches the
following example query from `fc-match`:
`fc-match -s "monospace:pixelsize=X:style=Y"
That being said it's more intuitive for users to setup their system Fontconfig
fallback, and also most applications are doing similar things. Moreover the new
implementation uses internal caches over Fontconfig API when possible and
performs font matches only once during load of requested font with font_sort,
which leads to dramatically improved performance on already mentioned
vtebench's unicode-random-write.
Fixes#3176.
Fixes#3134.
Fixes#2657.
Fixes#1560.
Fixes#965.
Fixes#511.
This cleans up the Alacritty scripts a bit by removing some of them
which are not recommended to be used anymore and switching from the
official FlameGraph tool to the more specialized Rust FlameGraph
implementation.
We currently log whenever we fall back to the default config because of
an error in the config itself. We also log when the config was
successfully loaded and where it was loaded from. The only scenario
where no config related message is logged is when there is no
configuration file present.
Logging this case should make it easier to debug issues like #3240,
without requiring any knowledge from maintainers about this edgecase.