Hi Reuben,
Seeing that we haven't chatted for quite some time, I wondered if you had time to fill me in on where you are at in respect of Zile, and any roadmap details you want to share?
Nothing has changed (and in fact, I've simply been working on other projects, rather than overwhelmed). I have one more, trivial, patch to port from C (adding a constant empty estr), and then it's time to rewrite the buffer code using alien for direct byte array manipulation.
My only other different thought is that I am not sure ctypesgen/mork is going to be a robust solution for automatic interfacing to libraries on a wide range of systems: it's just asking too much of an ad hoc C parser to parse arbitrary system headers. I was thinking instead of using gobjectintrospection. There are only hints of suitable curses bindings:
but the other functions that we need for file management &c. should be findable in glib. Using glib should have other advantages, and though it will naturally reduce portability slightly, I think it's in line with trying to reduce the amount of work we do by relying on other GNU packages; in any case, glib is pretty widely ported!
On my side, I noticed last week that C Zile has the same interminable slowness as Lua Zile if you paste a large block of text into it... which is actually good news, because it means the slowness is algorithmic and not intrinsic to Lua :-) I might see if I can find a way to speed that up, so that I can get back to adding syntax highlighting to Zi... unless you have a plan in mind for addressing that already?
I don't. I hope that profiling will show where the bottleneck is; I have found and fixed one or two performance bugs very simply in the past, and take it as a good sign that the code is getting simpler on the rare occasions where (re-)optimisation is necessary! Feel free to have a go.