Life & hacking

Some instructions here because the porting guide in the Gtk+ docs is quite short at the moment:

Convert the files

gtk_builder_convert <old_glade_file> <new_gtkbuilder_file>

Code changes

• Remove #include <glade/glade.h>
• GladeXML* => GtkBuilder*
• glade_xml_new (FILE, “first_widget”, NULL) becomes
  
GError* error = NULL;
GtkBuilder* builder = gtk_builder_new();
if (!gtk_builder_add_from_file (builder, FILE, &error)
{
g_warning ("Couldn't load builder file: %s", error->message);
g_error_free(error);
}

• glade_xml_get_widget (gxml, “widget_name”) becomes GTK_WIDGET (gtk_builder_get_object (builder, “widget_name”))
• glade_get_widget_name (widget) can be replaced by gtk_widget_get_name(widget)
• glade_xml_get_widget_prefix (gxml, “prefix”) can be emulated by gtk_builder_get_objects(builder) together with manual filtering. It returns a GSList* instead of a GList* though.

That’s it basically. Don’t forget to adjust your Makefile.am to install the correct files and update POTFILES.in!

After reading the 3.0 plan of the release-team I was curious what the end-users would say. It is rather difficult to find out because there is no average end-user. Anyway, I tried to summarize opinions from two popular german computer magazines:

• Heise-online: The biggest computer magazine in Germany (online and offline). They feature a rather big and detailed article on GNOME 3.0.
• Pro-Linux News: This is a non-profit Linux News site. Their article is smaller and the Nerd-factor of the readers is certainly a bit higher here.

Apart from some more off-topic posts (GNOME vs. KDE, MacOS is the only real thing, etc.) on the Pro-Linux board the reactions were quite similar with those main points:

1. Don’t make it a KDE 4.0! People rely on a stable desktop…
2. This looks interesting and cool! GNOME really needed some revamp…
3. Hmm, I cannot see why this is better than now – do we need it?

The rest of the comments are mostly based on misunderstandings (heise writes that “Zeitgeinst” will replace Nautilus which is simply untrue). Funny enough though, quite a lot of people comment that they liked our little steps in 2.x a lot which most sums up with point 1. These was much less critism than I expected actually and instead people applaud us for the work in the past years.

Of course this is no representative review about user comments, it is just a small limited summary (only German, people who are really interested in computing). In the end I think we seem to do well, but should mind the three points raised: Make it stable, make it cool, make it useful!