Hi, Todays update has a fix for the compact mode slant interference (as explained by Wolfgang a few days ago). Hopefully that works out well. We (MS and HH) also managed to make curve detection more accurate (the decision routine that is responsible for straight lines in the MP output) so that extreme paths with control points way beyond reasonable still produce the right output (before we could get a wrong bbox due to clipping the curve in the backend). One can actually also force "always curve" with "withcurvature 1" (more variants might follow) but that is more or us to have a test option. Curvature detection is mostly a pdf efficiency feature and normally users should rely on the default to work okay. Other new mp features are relative and absolute points which makes it possible to define paths in a turtle graphics way (a bit like hlineto and vline in postscript), a native "--" operator (equivalent but somewehat more efficient on huge generated paths). The luametafun manual has a chapter on paths that describes some of this. Yet another new feature is "firstcontrol" and "secondcontrol" that is an addendum to controls that makes both control points the same when only one is given. Finally there is a fix in mp where some directions, think { dir 120 }, gave the wrong result due to sign swapping (so it was seen as { dir 200} or { dir -60 }) as side effect of negative zeros fed into tan2. This became obvious when MS reprocessed a graphic, as last year we switched from scsled (no bug) to double (long time bug exposed). (The bug is also present in LuaTeX double/binary mode but it will be fixed there too.) The rest is not (yet) worth mentioning, Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl -----------------------------------------------------------------