On Wed, 28 Aug 2024 at 12:03, Shiv Shankar Dayal wrote:
I have two specific interests. I want to make asymptote work with ConTeXt but I can live with including generated images and use PDFs.
This is a much better specified question. It would ideally require some work both on the Asymptote side, as well as some work on the ConTeXt side. Asymptote is in fact written in C++, so it might be a lot closer to your particular expertise. I would say that what's most needed is someone with a great passion to get this complete, and some clear communication pathway between the two projects. Asymptote is very LaTeX-centric and should ideally be ConTeXt-ilized to simplify some operations. I would start by playing with a super simple minimum example written in plain asymptote. Figure out how to compile Asymptote on your computer and try to process the asy file into a ConTeXt file. This conversion then needs to be tweaked in order to produce two different results: - A standalone file that could be processed directly by ConTeXt. - A minimalistic file that could be included from another ConTeXt source to generate the final document. Check the file that you get right now, figure out what goes wrong, what could be improved ... If you have a simple file that no longer needs asymptote processing, you can check on the ConTeXt list what could still be improved, and then you can go and start fiddling with Asymptote source code (with help of Asy developers) until the files get progressively better. Iterate between the Asymptote and ConTeXt developers until the results are fully satisfactory. It has been a while since I touched any asymptote, but I suspect that what you want to achieve is something like \starttext \startasymptote some graphics \stopasymptote \placefigure[label]{really nice 3D model}{% \startasymptote another 3D graphics \stopasymptote} \stoptext where ConTeXt would process this file, prepare some input to asymptote, and then asymptote would have to be called just once to generate all the required graphics at once, and ConTeXt would eventually include them, properly scaled etc. Start with graphics, continue with simple text labels ... progress towards proper interactive 3D models ...
The second interest is in Aditya Mahajan's syntax highlighting module. It is very nice but it invokes VI making the entire thing slow. I understand the advantage of using VI is that we do not have to implement anything when a new language comes. VI will have syntax highlighting and it will be automatically done for us.
ConTeXt does support native syntax highlighting. You can try to create some new grammars to extend language coverage (Hans only made sure that TeX, lua, metapost etc. work correctly). Or you can try to start brainstorming whether some existing OpenSource grammars could be integrated to do syntax highlighting using Lua. Vi was a shortcut to get the syntax highlighting done in some way at all. In either case start a separate thread covering exactly that one topic. Mojca