On 6/12/2020 4:35 AM, Jairo A. del Rio wrote:

It's actually nice to find those bugs when one makes heavy use of math environments. For comparison, LuaTeX prints the correct output. Well, thanks for finding them! Normnally those bugs are easy to fix. Concerning math, there are no real fundamental changes under the hood, but in the process of adding some more control there can be side effects. For instance:

-- In order to make a bit nicer token interface (from the lua end) there has been a bit more abstraction and some internal quantities sit in better defined slots with zero based indices not, but in order to do that I need to adapt some references and might overlook some ... no big deal as there are not that many. -- Some math rendering is quite hard codes wrt style and already much has been made configurable (either controlled by keywords or by additional parameters, think of \mathfoomode like ones). But it's easy to overlook something there. This is why user testing helps. -- Some new features (like prescripts that are handy for proper chemistry as well as (i still have to adapt some code) mathml) but those will go unnoticed / not impact regular math. -- New (often already old by now) trickery like local math parameters, which at some point wil be better interfaced. -- Actually plenty more is possible but not yet used / done. It will happen when I run into the code / have a reason, but the mechanisms are there. In general, not all that is there is yet used. I often do that stepwise (just try it out someplace), also because sometimes we need to split mkiv and lmtx sources then. Stepwise changes are easier because all kind of things can interfere. Wrt non-math i.e. general new features in the engine, these can for instance involve more robust macro coding, full expansion, less horrific tracing, future possibilities. Users are welcome to suggest additional low level context macros that they're missing. We're also wondering if some low level macros can go away or maybe some 'raw' ones can be aliased to regular ones as performance issues matter less nowadays. There are for instance a few new tricks with macros but I'm not sure how many users really define there own, and if so they never asked for those specific things. I'm anyway more or less through my list of engine related todo's (that accumulated over years). All that said, I really appreciate the patience that users have with testing, 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 -----------------------------------------------------------------