David Kastrup wrote:
(in fact, it is a **** nuisance that [dimexpr] integer division rounds instead of truncating).
I agree, but that is a completely different problem.
concerning latex and old docs ... there will always be pdftex vs < 2, so with luatex we have no intention to be 100% downward compatible at all (some internals may also be adapted / extended / opened up, some good old tex code may disappear (lig rebuilding already has), etc
Interesting. What do you do instead of ligature rebuilding?
What has happened already is that ligature replacement and kerning are totally separated inside luatex: at font loading time, the ligkern information from the tfm metrics is split into ligatures and kernings. This means there are changes inside main_control() and reconstitute(). The reconstitution process was not 'perfect' before, and that now makes it hard to ascertain wether the new code is 100% identical. I think so, but I am not willing to bet on it. What will happen soon is that alternative implementation will become used for the hyphenation algorithm: one that does not limit words to 64 characters arbitrarily, and that allows pattern loading and augmenting at run-time. And there will likely be more changes along these lines. Taco