Linus Lewandowski

Note: Patch against pygi is not enough, I can't depend on the latest version of pygi as it's not on pypi.

Note: If I decide to auto-remove trailing \0 from ay when unpacking, I'll need to "repack" all GLib.Variants received from the user - to add trailing \0 (as changing the...

pydbus is a **Pythonic** dbus library. This means, that its primary goal is to be as intuitive, and standard-python-like, as possible. To hide as much of DBus's implementation details as...

Hm... Maybe we could create a new DBus annotation saying something like "treat all arguments of this method as \0-terminated bytestrings"? This wouldn't break anything, and high-level libraries will be...

I've looked at GDBus, and it looks like they're copying the values while converting between messages and GVariants. https://github.com/GNOME/glib/blob/master/gio/gdbusmessage.c#L1558 (g_variant_new_string copies the data, as it uses [g_bytes_new](https://developer.gnome.org/glib/stable/glib-Byte-Arrays.html#g-bytes-new) internally)

(That was for strings, for 'ay's ([gdbusmessage.c#L1687](https://github.com/GNOME/glib/blob/master/gio/gdbusmessage.c#L1687)) g_variant_new_fixed_array is used, which also [copies the data](https://github.com/GNOME/glib/blob/master/glib/gvariant.c#L1228))

AFAIR kdbus used GVariant's serialization, but it's dead now. Looks like we can safely change GVariant's in-memory representation to always include trailing \0 in "ay"s, and possibly create GVariant serialization...

I've never had any problems with checking why the test has failed. There is a traceback, and the line with `assert` is included. Of course `assert(False)`s aren't too readable, but...

And BTW, the GNOME Music test is not really useful for testing pydbus. I've made it when I was porting GNOME Music to pydbus (however, they've chosen to go straight...

I'm not dead, but I have other priorities. Maybe I'll find time to get back to it sometime in the future; in the past I was usually doing things in...