A bug in GtkTree?

[Constantine Vetoshev <con consela com>]

Greetings all,

I apologize in advance if I am submitting a bug report to this list in
error; but the GTK+ FAQ and the GNOME bug submission web page instructions
conflict on this issue (the FAQ suggests bug reports go to the mailing
list and http://bugs.gnome.org/Reporting.html suggests they are filed
to submit@bugs.gnome.org), and I figured this is the fastest way to make
others aware of the problem. That said:

I am using GTK+ 1.2.3. I believe there is a bug in how removal of subtree
elements is handled in the GtkTree widget. Here is what happens: if the last
item in a subtree is removed via gtk_container_remove() (and possibly by
other methods, I have not tried this), the subtree ceases being a subtree.
The +/- note disappears from the window and it seems like the subtree
is being treated as a simple GtkTreeItem. The GTK_IS_TREE assertion fails
if it is called on the subtree (and thus obviously nothing can be added
to it).

I have kludged around this problem by adding a check in my application
that recreates the subtree if it was destroyed by this side effect.
I assume this is a bug.

Attached is the simplest program I could come up with that shows this
problem (50 lines of code).

Regards,
Constantine Vetoshev

PGP signature