Comment 14 for bug 2006059

I could not reproduce with kinetic (see my steps in bug 2016193). I went back to a clean jammy install, reproduced, then upgraded just the gdm3 packages to kinetic's versions, and the problem went away, so it seems gdm3 in isolation is the relevant difference. I took a look at the differences between the jammy and kinetic gdm3 source, and there really isn't that much different. One piece that stood out is the following, which appears to be a typo:

diff -urpN gdm3-42.0/daemon/gdm-local-display-factory.c gdm3-43.0/daemon/gdm-local-display-factory.c
--- gdm3-42.0/daemon/gdm-local-display-factory.c 2023-04-17 15:07:42.000000000 -0600
+++ gdm3-43.0/daemon/gdm-local-display-factory.c 2022-09-20 09:22:21.000000000 -0600
@@ -402,7 +402,7 @@ gdm_local_display_factory_create_transie
                 display = gdm_local_display_new ();
                 g_object_set (G_OBJECT (display),
                               "session-type", session_types[0],
- "supported-session-tyes", session_types,
+ "supported-session-types", session_types,
                               NULL);
                 is_initial = TRUE;
         }

On a lark, I rebuilt the jammy version of gdm3 to use "supported-session-types" here, and that seems to have fixed the problem for me. I can reliable downgrade to the jammy version of gdm3 and reproduce, and reliably *not* reproduce the bug after installing my patched version.