[Cairo] Re: [xsvg] cairo_text_extents ?

[Carl Worth]

On Dec 4, Bill Spitzak wrote:
 > Okay I'll accept that, if changing the matrix does not cause the current 
 > point to change.

Correct. As for adding another matrix slot into the gstate, we can
look into that again if there's ever a demonstrable reason to do this
for performance.

 > It appears Postscript (or Ghostscript) works like Cairo.

Good. I was implementing to match my reading of the PostScript
specification. It's good to get some assurance from another
implementation.

 > I want to allow a platform-specific "configuration" program that says "the 
 > default font is this". As designed I CANNOT do this unless every program uses 
 > the platform-specific interface.

I don't think I've understood your comment here. Is it that you would
like an interface that would enumerate the acceptable family names
that could be passed to cairo_select_font? We could probably add that
without going too far over the slippery slope.

Otherwise, I'm afraid I'll have to ask you to restate what you see
missing from the current API.

-Carl