Hi Boris,

> Those create new nsWebBrowsers, right?  

yes, absolutely. they get Init-ed, a parent URI content listener is
set etc. We simulate the firefox tabs with the concept of full-screen
"sessions", one new nsWebBrowser-derived object for each session that
gets created. Works like a charm.

>I'd think the parent URI content listeners for those would get called.  Do 
>they not?

only the OnStartURIOpen method gets called, twice, first time with URI
pointing to the PDF doc and second time with uri = about:blank. Both
times on the listener for the newly-created nsWebBrowser (uh, that's
strange... I expected an about:blank on the old one and the PDF URI on
the new one... maybe I have a bug lying somewhere?)

no other nsIURIContentListener interface methods are called.

the old "session" with the link gets swapped on the background and I
remain with a blank page on the foreground (ok, this is how our app is
designed... just to tell you the creation/swapping mechanism works as
expected)

... I noticed that a nsIEmbeddingSiteWindow::SetVisibility() call get
issued in the load, and I was implementing that with a "return
NS_ERROR_NOT_IMPLEMENTED" body. I tried to return NS_OK, and also I
completely removed the nsIEmbeddingSiteWindow interface  (which I
don't use) from my class, but nothing changed.. but I have some more
stuff "implemented" in that way. Maybe something gets called, goes
silently wrong for some reasons (because just a "return
NS_ERROR_NOT_IMPLEMENTED" is not acceptable) and breaks the underlying
flow of calls?

the nsIEmbeddingSiteWindow::SetVisibility() is NOT called if I click
on a PDF URI without target=_blank attribute.. and in that case
everything works.

I have read somewhere that implementation of nsIEmbeddingSiteWindow
interface that is mandatory . Do you confirm that? Maybe those "return
NS_ERROR_NOT_IMPLEMENTED" implementations on stuff like SetDimensions,
SetFocus, GetTitle, GetSiteWindow etc, or their complete lacking,  can
break something?

Thank you,
Aaron

_______________________________________________
dev-embedding mailing list
dev-embedding@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-embedding

Reply via email to