Generally, how much space we have in squashfs, or tmpfs shouldn't
interest us. This becomes more relevant in distros like Ubuntu, where
snaps are a thing, and each snap mounts their own FS in a squashfs that
is always full, thus falsifying the output of sysinfo.
Partial fix for #2271
This isn't an exhaustive list, but it's everything I could find. The bug still exists in the parser though, this is just a workaround for the moment
* Make sure `help()` doesn't cause hexchat to hang
Replace `pydoc.help` with a copy of `pydoc.Helper` with an empty
`StringIO` instead of stdin
* Handle BytesIO vs StringIO on 2.7
This is mostly useful to avoid a newer gettext dependency
for translating the appdata file but it is also just useless
data for some distros without any app store.
Closes#2219
The switch to the meson build system broke plugins on macOS. GNU libtool
builds shared libraries with ".dylib" and shared modules (plugins) with
the extension ".so", but meson is using ".dylib" for both.
Although overriding the name_suffix for shared_module() in meson is
possible, this would be messy for other platforms as there is no way to
query the default. Therefore it seems like we have to go with ".dylib"
for now on macOS.
However, G_MODULE_SUFFIX is defined to ".so", because glib follows what
GNU libtool does. Therefore define a separate preprocessor macro that
has the correct extension.
See: https://github.com/mesonbuild/meson/issues/1160
With the switch to meson, the problem previously fixed in #1822 came
back. The build system might pick up the installed hexchat-config.h
instead of using the header in the source directory, as the compiler
arguments would be in the order of "-I${prefix}/include -I..".
It seems that the c_args in meson are always put to the front of the
compiler arguments, in order to be able to override any include paths
from dependencies. However, this was not the intention here, so perl
should also be modeled as a dependency. This ensures that the arguments
with local include directories come first.
