When the DPI factor changed and the window hasn't been resized,
Alacritty incorrectly resized the window to the previous size without
scaling it by the new DPI factor.

This has been fixed by making sure that every resize event triggered by
a DPI change uses the DPI factor to scale the window.

The mesa workaround has lead to some issues with
rendering on Wayland.

To resolve this problem, the mesa workaround has been
restructured in a way which still allows clearing the screen
before rendering without killing performance with the mesa
driver. The performance is identical to the master brach
and there have been no recorded regressions.

This changes a lot of small details in the configuration file in an
attempt to unify its structure. These are some main guidelines used for
this refactoring:
    - Specify that changes require restart consistently
    - Unify the specification of available field values
    - Provide clear distinction between description title and body

Besides these guidelines used to unify minor details in the
configuration file, the section on key configuration has been completely
reworked in an attempt to reduce the amount of text used. This should
make it possible to understand what's going on without having to read
any text.

The notice that modifiers are not supported has been removed from the
mouse binding documentation.

The latest change to window decorations
(3b46859e) introduced a regression when
running Alacritty without the `decorations` field specified in the
configuration file. Since serde wasn't setup to fallback to the default,
the complete config deserialization would fail.

This resolves this issue by deserializing it to the default decorations
value "Full". To make this setting a little more forgiving, this also
introduces another change which ignores the case for the configuration
options. So both `full` and `FuLl` are now accepted.

The bracketed paste mode change isn't so much a fix as it is a change to
address a potential security issue.

The decorations config was changed from a bool to an enum.
`full` has taken the place of `true`, and `none`, has replaced `false`.

On macOS, there are now options for `transparent` and `buttonless`.
These options are explained in both the CHANGELOG and in the
configuration files.

* Fix Bracketed Paste Mode when input contains end sequence.
* Remove \x1b instead of just the paste end sequence.

Update snapcraft.yaml

The existing snapcraft.yaml is pointing to a file that has been renamed,
subsequently breaking the snap build. This change renames the .desktop
file to match what is in-tree.

Alacritty has some checks in place to make sure that unsafe
code would not fail because of invalid struct sizes. This managed
to successfully catch an incorrect unsafe block on 32-bit machines.

To make sure this block works on both 32-bit and 64-bit systems,
it has been altered to make use of the platform-dependent `usize`
type. This will always make use of correct sizes without having to
rely on conditional compilation.

The core-* dependencies have been updated and every breaking
change has been resolved.

These are the main changes which required adaption:
- font_path() returns a PathBuf now
- get_descriptors() returns an Option<CFArray>
- get_advances_for_glyphs and get_glyphs_for_characters are now unsafe.

All packages which did not have breaking updates have also been updated.

Alacritty made the assumption that every window started as focused and
because of that the hollow cursor wouldn't show up for windows which are
launched without focus.

Since even the initial focus should be reported as a focus event by
winit, this could be easily fixed just setting the default window state
to unfocused instead of focused.

This fixes #1563.

This adds support for the `hidden` escape sequence `\e[8m`, which will
render the text as invisible.

This has also raised a few questions about the rendering of foreground
and background colors and their interaction with the different escape
sequences. Previously, Alacritty has oriented itself after URxvt, which
has some strange and unexpected behavior.

The new implementation of color inversion is modeled after XTerm, which
has a consistent pattern of always inverting the foreground and
background colors. This should hopefully lead to less confusion for the
user and a more consistent behavior.

A full matrix showcasing the new way Alacritty inverses text can be
found here:
https://i.imgur.com/d1XavG7.png

This fixes #1454 and fixes #1455.

macOS 10.14 will bring a new system wide dark mode. To enable this, the
[official guides][] suggest to relink using the newest OS. This
approach, however, did not work for me as described in [an issue][] in
the glutin repository.

As a second option, the accompanying `Info.plist` file can also
overwrite the link-time check and enable dark mode rendering if the
system config is set by setting `NSRequiresAquaSystemAppearance` to
`YES`. This approach seems to work flawlessly no matter if a user opts
into dark mode or not.

I would appreciate it if someone can test this on macOS 10.13 as well,
but I suppose the key there is unused and would not break anything.

[official guides]: https://developer.apple.com/documentation/appkit/appkit_release_notes_for_macos_10.14_beta
[an issue]: https://github.com/tomaka/glutin/issues/1053#issuecomment-409315461

Changing tabspaces from the default (8) requires a corresponding
update to the `it` item in the terminfo entry used.

Some applications, like Emacs, rely on knowing the width of a tab, and
will experience unexpected behavior if the terminfo data does not
match the actual width used.

Since the scrollback history allocates all lines in memory, it is
possible to specify a scrollback history which is big enough to freeze
the computer due to OOM.

To resolve this issue, an upper limit of `100_000` has been set for the
scrollback history. Even though this might still cause some systems to
freeze, this should provide a good balance for most users.

The zsh completion name for the deb installer has been corrected.

The installation instructions for the zsh completions have also been
updated to make it possible to install them without root permissios.

Set `COLORTERM` to `truecolor` in order for applications to be able to
detect that alacritty supports 24-bit colors.

See https://gist.github.com/XVilka/8346728 for more details.

Closes #1526.

Fixes a regression on non-macOS platforms caused by the fix for
issue #1291. The fix is to follow platform norms for mouse click
behavior on unfocused terminals. On macOS, the first click (that
gives the window focus) is swallowed, and has no effect on the
terminal. On all other platforms, the first click is passed through
to the terminal.

Since Alacritty does not require the AUR anymore, its mention has been
removed from the README.

To make future changes simpler, the README has also been reworded so
less changes are required for adding new operating systems.

Scrollback

* Change deb installation from crates.io to git

There have been a number of issues an PRs opened since
the cargo-deb installation does not work with the latest
version from crates.io.

To help out users until the crates.io version is updated,
the installation instructions have been temporarily
changed to install `cargo-deb` through github.

* Revert cargo-deb install back to use crates.io

Since `cargo-deb` has been updated on crates.io it is now
possible to just install it from crates.io and build Alacritty's
deb without having to rely on github.

* Update dependencies

This fixes an `illegal hardware instruction (core dumped)`
error when building in release mode.

* Remove redundant copy when selecting font_key

* Bump version number to 0.2.0

Since the Scrollback branch introduces some major changes, this bumps
the version number from 0.1.0 to 0.2.0.

The versions of Alacritty have not been updated regularly to this point,
so the scrollback branch is a good point in time to start updating
Alacritty's version on a regular basis.

Further changes to the readme, like dropping the 'alpha' status and
updating it to 'beta' could also be introduced with this branch. This
way there will be a clean cut which updates everything as soon as
scrollback is merged.

Building versions is another thing which would be a good thing to start
reasonably quickly. However starting this on the main branch after
scrollback has been merged seems like a more reliable way to move
forward.

This fixes #1240.

* Add a CHANGELOG file

A CHANGELOG file has been added to offer a bit more transparency over
which features have been changed, added and potentially removed in
Alacritty.

There are various formats available for the CHANGELOG file but the most
common and sensible one seems to be the one defined by
https://keepachangelog.com/en/1.0.0. Following the template proposed by
this it should be possible to create a clear CHANGELOG which makes it
simple for new contributors to figure out exactly which formatting
should be used for it.

Since there have been quite a few changes to Alacritty already, not all
changes have been added to the changelog. However a few entries have
been ported just to give a bit of an example what the format should look
like. This also helps with the 0.2.0 version since it will not be
completely empty in the changelog.

This fixes #1534.

* Update CHANGELOG

This updates the CHANGELOG to include the changes introduced by
43882ade.

The IL escape sequence (CSI Ps L) allows inserting blank, uninitialized
lines. `Ps` is a placeholder for the number of lines that should be
inserted. Before this change Alacritty would crash when a large number
of lines was passed as `Ps` parameter.

The issue was caused whenever the current line of the cursor plus the
lines that should be inserted would leave the bottom of the terminal,
since this makes indexing impossible.

This patch makes sure that the biggest amount of lines inserted does
never exceed the end of the visible region minus the current line of the
curser, which fixes the underflow issue.

This fixes #1515.

The clearing the screen for the `ansi::ClearMode::Saved` enum value
has been implemented. This is used to clear all lines which are
currently outside of the visible region but still inside the scrollback
buffer.

The specifications of XTerm indicate that the clearing of saved lines
should only clear the saved lines and not the saved lines plus the
currently visible part of the grid. Applications like `clear` send both
the escape for clearing history plus the escape for clearing history
when requested, so all sources seem to agree here.

To allow both clearing the screen and the saved lines when a key is
pressed the `process_key_bindings` method has been altered so multiple
bindings can be specified. So it is now possible to execute both `^L`
and `ClearHistory` with just a single binding. The
`process_mouse_bindings` method has also been changed for consistency.

To make sure everything works properly a test has been added which
clears the history and then attempts to scroll. Since scrolling is the
only way for a user to check if scrollback is available, this seems like
a nice abstraction to check if there is a scrollback.

Since Alacritty never had any scrollback history, the behavior when the
window height was increased was to just keep the prompt on the same line
it has been before the resize. However the usual behavior of terminal
emulators is to keep the distance from the prompt to the bottom of the
screen consistent whenever possible.

This fixes this behavior by loading lines from the scrollback buffer
when the window height is increased. This is only done when scrollback
is available, so there are only N lines available, the maximum amount of
lines which will be loaded when growing the height is N. Since the
number of lines available in the alternate screen buffer is 0, it still
behaves the same way it did before this patch.

Different terminal emulators have different behaviors when this is done
in the alt screen buffer, XTerm for example loads history from the
normal screen buffer when growing the height of the window from the
alternate screen buffer. Since this seems wrong (the alt screen is not
supposed to have any scrollback), the behavior of Termite (VTE) has been
chosen instead.

In Termite the alt screen buffer never loads any scrollback history
itself, however when the terminal height is grown while the alternate
screen is active, the normal screen's scrollback history lines are
loaded.

This fixes #1502.

Some minor formating issues were still present in the current state of
the PR, these should all be resolved with this.

The comment about converting between logical and physical sizes was also
still present in the `src/display.rs` even though the conversion itself
has been moved to the `src/event.rs` file. To fix this the comment has
also been moved to the `src/event.rs` file.