>> Looking at the 0.3.5 sources I'd originally though that the output of >> metapost was used by pdftex, but I see that you parse the PS and write >> your own pdf fragments. Since SVG uses exactly same rendering model as >> pdf it should be easy enough to adapt this to produce SVG fragments >> also. With these, replacing pdftex for the job of stitching the >> fragments together into a map should be feasible.
Yes, SVG is planned to be implemented in about a month. The PostScript parsing needs to be enhanced a bit (some simplifications were takan because of similarity of PS and PDF). The idea is to get only map in SVG (and other formats), leaving the Atlas for PDF. >> Also, do you have a CVS/svn repository for the code? If so, would it be >> possible for that to be exposed read-only? I hate developing against >> point releases when I know that a project is still undergoing such >> active development. We don't use online CVS repository -- I would have no possibility to access it. We used to put the development tarball on the web page every week or so, but that led only to confusion. M.
