Hi On Mon, Sep 15, 2014 at 1:58 PM, Ran Benita <ran...@gmail.com> wrote: > On Mon, Sep 15, 2014 at 08:21:34AM +0200, David Herrmann wrote: >> Why don't we require a keymap in xkb_compose_state_new()? Should be >> easy to do and we can then track modifiers reliably. >> Ok, a keysym which is a modifier in one keymap should never be a >> normal key in another.. unlike keycodes. But I don't see why we need >> to rely on that when we can just take a keymap. > > I prefer not adding a dependency on a keymap just for this one little > thing. That would preclude people who don't use xkbcommon keymaps from > using xkbcommon-compose (not likely, but maybe), and complicate the API > slightly. For not much gain..
We could allow NULL and then use this fallback.. > Also, that's what the de facto specification here (Xlib) says, so it's not > a bug, it's a feature :) Fair enough. You might guess that I don't care much about Xlib, but this is just a nitpick so I'm fine either way. Thanks David _______________________________________________ wayland-devel mailing list wayland-devel@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/wayland-devel
