Hi all, A quick note from me. I did not do anything useful on three out of the past seven days, thanks to a short-lived but nasty virus. As a result, I am a little bit behind on my schedule. == opentype font writing == The current code can actually already produce a syntactically correct PDF with wholesale embedded OpenType font, but any kind of metrics or ToUnicode information is still totally absent. As a result of that each character is exactly 1em wide, and cut&paste produces absolute rubbish. This is too ugly to be usable for anything at all, so you have to wait a little bit longer. As soon as I can get the Metrics for OpenType fonts correctly into the PDF, there will be a new snapshot (probably on sunday, this is not rocket science). Subsetting and ToUnicode is something to worry a little later, it is a bit less urgent than having valid output one can look at. == Opentype font reading == What is still missing from the otf-to-lua parser is any kind of Apple support. The code does not handle Mac names properly yet, not does it understand anything about AAT or Apple MM fonts. I would be happy if someone knowledgeable could borrow me an OpenType font with AAT features so I could implement and test support for that next week. I actually own an OS X Tiger license, so it would probably even be legal ;-) == Aleph bidi support == I have started work on supporting Aleph's bidi typesetting in PDF mode as well, simple text seems to be largely ok now. More complicated documents will no doubt still show some bugs, but fixing those will be easy (once they are discovered). Exactly what happens with specials is not yet investigated. Omega (16-bit wide) virtual fonts now work OK in PDF mode. == Other known bugs == The following problems are still open and will need attention next week: * Word-boundary ligatures are corrupt. This is a result of differences in the internal data structure. * Font expansion is current disabled, also because of data structure issues * The experimental \letterspacefont is also non-functional, but I am very tempted to leave that alone for the moment. * The font-character arrays can eat up quite a lot of memory now that encodings can be 21bits wide, I will have to re-implement the storage to be more memory-efficient. So far, Taco