Re: No ibus modules for GTK4 with Sway

[Tomas Volf <~@wolfsden.cz>]

[-- Attachment #1: Type: text/plain, Size: 2044 bytes --]


Hi Ricardo,

Ricardo Wurmus <rekado@elephly.net> writes:

> Hi Guix,
>
> I've reconfigured my system with Sway and I noticed that GTK4
> applications now ignore my input method.  GTK3 applications and Emacs
> still work fine.
>
> GTK4 applications used to work with IBus under Gnome, so I'm guessing
> that some more configuration is required to make this work...  Does
> anyone have an idea what might be missing?
>
> FWIW, I find the configuration rather confusing.  Our Cookbook doesn't
> contain a working example for using Sway (which is surprising given its
> popularity among Guix and Nix users), and whatever the reference manual
> mentions when it comes to Sway appears to be unnecessary or overly
> complicated (such as the configuration of greetd, which required the
> removal of other services).
>
> I still haven't arrived at a comfortable system configuration (GDM still
> starts up even though I've removed it from my configuration; only
> elogind remains), but the input method problem with GTK4 applications is
> the most pressing issue at this time.
>
> I'd be happy for any hints!

I do not think I am using any GTK 4 applications (what would be some
light example I could test?), but I wonder whether you need to set
GTK_PATH (probably via GUIX_GTK4_PATH) to something to allow it to find
the im modules.  Would it be possible to check whether Gnome sets it?

I see for GTK 2 and 3 we have GUIX_GTK2_IM_MODULE_FILE and
GUIX_GTK3_IM_MODULE_FILE, there is none for GTK 4 though.  So maybe I am
wrong and the auto-discovery should just work for 4.

However it looks like we *are* building the im module in ibus:

--8<---------------cut here---------------start------------->8---
$ ls $(guix build ibus | tail -n1)/lib/gtk-4.0/4.0.0/immodules
libim-ibus.la  libim-ibus.so
--8<---------------cut here---------------end--------------->8---

Hope this helps,
Tomas

-- 
There are only two hard things in Computer Science:
cache invalidation, naming things and off-by-one errors.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 853 bytes --]