On Wed, 25 Aug 2004, Teemu Ikonen wrote:

Down at the bottom of the roadmap (http://www.inkscape.org/roadmap.php) there's a note about splitting the SVG canvas out of Inkscape.

I'm wondering how far this goal is in terms of development effort. The canvas in Inkscape is by far the best looking and most functional I've seen in any free software project. I'm sure there are many projects which would benefit from a canvas with nice antialiased output to screen, good print quality by SVG->PS conversion and the ability to manipulate the figure by pointing and clicking in some way.

Could someone familiar with the Inkscape code estimate if it would make sense at the moment to try to use the canvas this way in, say, another GTKmm-application? Or is the code just too much coupled with GUI right now?

If you'd like to give it a shot, I don't see any reason why it couldn't be attempted right now, even doing it just as an experiment to start with. A number of people have asked about it, so there's definitely interest in it.

What I think we need to figure out, to start with are some questions:

* What functions would the SVG Canvas need to support?

* Does the SVG Canvas need to be a Gtk object or could it be done abstractly? Perhaps we'd have an underlying SVG library with a gtkmm wrapper atop it? See Gtk::DrawingArea

* How does this relate to our Cairo plans?

Those questions aside, the process of splitting the canvas out would probably involve creating a copy of the codebase and then just start stripping out everything not needed for a canvas. Then create a canvas class with the functions it should support and start testing it out. :-)

Bryce