There's currently a small paint glitch for vertical toolbars due to the
way StylePainter::paint_surface() draws a MidGray line at the bottom of
whatever a "surface" is supposed to be.
You can now press Ctrl+Shift+Up/Down in a GTextEditor and the currently
selected line(s) will all move together one step up/down.
If there is no selection, we move the line with the cursor on it. :^)
Since on_change handlers can alter the text document we're working on,
we have to make sure they've been run before we try looking at spans.
This fixes some flakiness when a paint happened before HackStudio had
a chance to re-highlight some C++ while editing it.
The design where clients of GTextEditor perform syntax highlighting in
the "arbitrary code execution" on_change callback is not very good.
We should find a way to move highlighting closer to the editor.
It should be possible for the CSS parser to fail, and we'll know it
failed if it returns nullptr. Returning RefPtr's makes it actually
possible to return nullptr. :^)
This simple helper escapes '<', '>' and '&' so they can be used in HTML
text without interfering with the parser.
Use this in IRCClient to prevent incoming messages from messing with
the DOM :^)
This function parses a partial DOM and returns it wrapped in a document
fragment node (DocumentFragment.)
There are now two entrances into the HTML parser, one for parsing full
documents, and one for parsing fragments. Internally the both wrap the
same parsing function.
Add LayoutPosition and LayoutRange classes. The layout tree root node
now has a selection() LayoutRange. It's essentially a start and end
LayoutPosition.
A LayoutPosition is a LayoutNode, and an optional index into that node.
The index is only relevant for text nodes, where it's the character
index into the rendered text.
HtmlView now updates the selection start/end of the LayoutDocument when
clicking and dragging with the left mouse button.
We don't paint the selection yet, and there's no way to copy what's
selected. It only exists as a LayoutRange.
When adding a widget to a parent, you don't always want to append it to
the set of existing children, but instead insert it before one of them.
This patch makes that possible by adding CObject::insert_child_before()
which also produces a ChildAdded event with an additional before_child
pointer. This pointer is then used by GWidget to make sure that any
layout present maintains the correct order. (Without doing that, newly
added children would still be appended into the layout order, despite
having a different widget sibling order.)
Renamed "Position" to "Elapsed". "channel/channels" automatically
changes now when more than one channel exist. The current file name
is now displayed in the window title.
m_loaded_samples was incremented with the value of the processed
buffer. This causes m_loaded_samples to be bigger at some point
than m_total_samples when downsampling, as the buffer would contain
more samples than actually loaded.
Files opened with O_DIRECT will now bypass the disk cache in read/write
operations (though metadata operations will still hit the disk cache.)
This will allow us to test actual disk performance instead of testing
disk *cache* performance, if that's what we want. :^)
There's room for improvment here, we're very aggressively flushing any
dirty cache entries for the specific block before reading/writing that
block. This is done by walking the entire cache, which may be slow.
LibAudio now supports pausing playback, clearing the buffer queue,
retrieving the played samples since the last clear and retrieving
the currently playing shared buffer id
`atof()` has now been implemented as part of the standard C library.
It supports scientific notation such as `1.2e-3` etc, ala the version
found as part of `glibc`.
It's a bit chunky, so there's probably room for optimisations here
and there, however, for now it works as intended (and allows Quake
to run).
This patch adds a limit of 200 unsent messages per client. If a client
does not handle its incoming messages and we manage to queue up 200
messages for it, we'll now disconnect that client. :^)
If an IPC client is giving us EAGAIN when trying to send him a message,
we now queue up the messages inside the CoreIPCServer::Connection and
will retry flushing them on next post/receive.
This prevents WindowServer from freezing up when one of its clients is
not taking care of its incoming messages.
The old implementation of sin() had a very unacceptable amount of
error that was causing a lot of texture perspective issues in Quake.
This has been remedied through the use of the hardware `fsin`
x87 instruction. As has been noted in #246, this instruction is both
very slow, and can become wildly inaccurate for more precise values,
however the current implementation made an incorrect assumption about
applications "will end up writing their own implemtnation anyways",
which is not the case for Quake in Software mode, which relies heavily
on a correct `sin()` and `cos()` function for `R_ScanEdges()`.
Realistically, we should be using something like the CORDIC algorithm
(https://en.wikipedia.org/wiki/CORDIC) or `Taylor Expansion`, however
for now these provides a somewhat accurate result that allows Quake to
run without any texture or geometry issues.
Ports/.port_include.sh, Toolchain/BuildIt.sh, Toolchain/UseIt.sh
have been left largely untouched due to use of Bash-exclusive
functions and variables such as $BASH_SOURCE, pushd and popd.
