Bernie Maier wrote:
> As I've mentioned in various earlier emails, the Linux port of the
> application I'm working on uses the basic GTK+ Mozilla embedding widget.
> It basically works, but I have a further bunch of possibly naive
> questions about it:
>
> 1. On many web pages, my debugging code shows that there's a whole
> bunch of net_state_all signals emitted for a NULL URI. What on
> earth are these?
Really? Can you get a backtrace on those so I can try and track down
why that's happening?
>
> 2. Is there a recommended way of determining when the document has
> finished loading? There seems to be a couple of different signals
> I can use. I'm currently using net_state_all and checking for the
> "stop" and "is document" flags, but signals still seem to fire
> for the page after that net_state_all arrives. Should I instead
> be keeping a count based on net_start (increment) and net_stop
> (decrement) signals, and assuming when the count falls back to zero
> that the document has finished loading?
That's the way it's supposed to work. Once again, a URL would be nice
where that's happening.
>
> 3. Is the API compatible across different versions? I'm compiling and
> linking against the 0.9.1 milestone but want users to not have to
> download my embedding package if they already have a (possibly later)
> version of Mozilla installed. I was hoping the answer is yes, but
> I suspect it is no because when I use milestone 0.9.4, calls to
> gtk_moz_embed_append_data() seg fault.
>
In general, yes it is. There have been several minor API additions over time but I
don't think that I've changed anything in a way that's not backwards compatible.
Also, with your specific problem can you open a bug please?
--Chris
--
------------
Christopher Blizzard
http://people.redhat.com/blizzard/
Mozilla.org - we're on a mission from God. Still.
------------
