Re: Heated agreement? (was) Re: Canvas shortcomings

[Mark <jamess1 wwnet net>]

On 21 Jun 2001, Lauris Kaplinski wrote:

> 
> On 20 Jun 2001 06:15:08 +1000, Martin Sevior wrote:
> > It seems to me that we're in heated agreement that there should be a set
> > of virtual primitive functions that call arbitary backend graphics
> > contexts. Right?
> > 
> > To start the ball rolling I could post the set implemented by AbiWord if
> > that would help in establishing what set of virtual functions should be
> > defined.
> 
> Well, I try to word my initial problem more clearly.
> 
> We are talking about TWO generic graphic API-s
> 
> 1. Drawing API (like libart or Gdk drawing calls)
> 
> 2. Object-building API (like current canvas API)
> 
> Whehter we need (1)? Whether we need (2)? If we need both, should they
> be connected? If yes, then how?
> 
> Now, is the question relevant at all?
> IMHO it is. People are constantly talking about adding ::print to 
> canvas, or migrating all graphic code to single API. One cannot talk
> about those issues meaningfully, without answering abovementioned
> questions first.
> 

What I think would be nice is a vector based API at the level of
Postscript in gdk. This would only do drawing, wouldn't have lightweight
shape classes or anything of the sort. This would be (1). And then for
(2), the canvas API with objects, events and things of that sort would be
nice. These wouldn't really be related, only that (2) of course depends on
(1). In addition (1) would have the device abstraction. So you can draw to
the printer, screen, file, ect. Therefore printing the canvas would be
trivial under this setup.

Mark