Currently when we render EMF+ images, they are first rendered to a bitmap (using a set size/resolution). This causes fonts to render poorly and pdf export to contain the bitmap, instead of the text.
This is documented clearly in vcl/README. To fix this, I believe that it should be possible to copy cppcanvas/source/mtrenderer/emfplus.cxx into vcl/source/filter/wmf and adapt it to use "class WinMtfOutput" So, vcl draws the EMF+ in the same way as the current EMF. The vcl/README says this is a lot of work. I am not sure exactly why this is. It appears that it is merely a cut and paste exercise What I would find helpful is somebody to comment on the accuracy of vcl/README, since this section was updated a couple of years ago, the comments may no longer be valid. There was a comment in http://nabble.documentfoundation.org/Query-about-the-direction-of-EMF-EMF-pr ocessing-td4132203.html#a4132247 by Thorsten that suggests there are further issues. Thanks. _______________________________________________ LibreOffice mailing list LibreOffice@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/libreoffice
