Hi All, > A GTK and Cairo free DirectFB backend would be the best :)
I agree with dok. A "pure" DirectFB backend would be lighter than a gtk-based backend. For porting Gecko, we must port the classes contained in two directories in its sourcecode: gfx and widget. The "gfx" dir provides drawing functions and the "widget" dir provides basical widget implementations. Perhaps Gecko has mostly all the code for drawing the widgets it needs. So we don't need all the power a complex toolkit like Gtk provides. Yes, maybe using Gtk makes the porting process more easy, but it doens't is the ideal. Maybe we could use some LiTE in the Gecko port, as I've already said, as it should provide all that we need. The company I work will sponsor me in the project of porting Gecko to DirectFB, as they want to run Gecko under really poor hardware. It must run as lighter as possible, so "pure" DirectFB should be the best. Then, we could make terminals with Linux kernel + uclibc + DirectFB + Gecko. It seems to be very light... > Have you also looked at WebKit? Anyway, hmm, it would be tempting to get > Gecko running as a DirectFB/Web OS solution. I think the future > embedded Os is a Web OS. Yes, it's tempting to me too =) And yes, I've looked at WebKit. It's the Apple's khtml-based engine. I've even tested a port of it to gtk, named gtk-webcore. It seems very promising, but Gecko is far more madure. And Gecko not only renders HTML, it provides XUL, so a whole complex interface can be made on top of it. All the Mozilla suite is written in XUL: Firefox, Thunderbird, Seamonkey, and so on. It's managed and drawed by the Gecko core. Yes, if we port Gecko, we can imediattely run the Mozilla suite and any other XUL-based software under pure DirectFB. It is wonderful, isn't it? ;) Regards, Paulo Matias _______________________________________________ directfb-dev mailing list [email protected] http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-dev
