On Sun, 2025-09-28 at 18:15 +0200, sj wrote:
> > As far as I can see that re-definition of
> > set_continuous_typesetting*
> > is entirely spurious, so your hunk 3 is good.
> Thanks, I was starting to suspect pdfview.c might not have been built
> at 
> one point but it was still strange.
> 
> > What is not working? No Print View?
> 
> Right now Print View (or any typesetting-related window, as far as I
> can 
> see) doesn't even open,

this would mean that USE_EVINCE is not set as the Print View is created
(in core/view.c) via this:

#ifdef USE_EVINCE
  install_printpreview ();
#endif

I guess from this that the name USE_EVINCE was devised before the
libevince strayed from the original API, before we switched to libatril
to provide all the ev_* functions. So this will be down to the
configure step, for a normal user-facing build of Denemo I think you
would need USE_EVINCE true and something to make it use the libatril
library.

>  I also get these messages in the terminal:
> 
>     Denemo - MESSAGE : Set TogglePrintView to active
> 
>     Denemo - MESSAGE : Set TogglePrintView to inactive
> 
> randomly interspersed with a bunch of
> 
>     Gtk - CRITICAL: gtk_widget_queue_draw: assertion 'GTK_IS_WIDGET 
> (widget)' failed
>     Gtk - CRITICAL: gtk_widget_queue_draw: assertion 'GTK_IS_WIDGET 
> (widget)' failed
> 
> errors.

So those look like the showing/hiding of the Print View window is not
avoided when USE_EVINCE is false, and consequently it tries to draw a
non-existent widget. 

Richard


>  (Also got a [GLib-GObject - CRITICAL: value "-1" of type 'gint' 
> is invalid or out of range for property 'gtk-tooltip-timeout' of type
> 'gint'] error, though that seems like something else that might have 
> gotten itself stuck into my config, at least at first glance.) Or
> maybe 
> not, at a second glance, it's always just four errors.
> 
> The number of GTK_IS_WIDGET(widget) errors seems about on par with me
> spamming the Print View button, but sometimes a few less. (My
> immediate 
> goal was to package it for Void Linux, I think it had worked just
> fine 
> last year, so I unfortunately don't have a clean debug build lying 
> about, but I'll clone it tomorrow for real and check it out.)
> 


Reply via email to