15
« on: January 08, 2023, 03:51:22 pm »
Having the libcn in a lesser state than EMX is... well, as a developer I understand that there's so much to do and so little time, so applying YAGNI principle, if these calls were not required when the libs were leverage for a given port (say firefox), it's understandable they aren't at the same quality.
However, when it's time to port, what should be the target now? It should be libcn. But... I understand it can't be used for console apps? Nonsense.
Perhaps the priority of the development community (not putting the burden on one dev, but more in a collegial way - if we want to conquer the hill, we need a strategy) should be to give some love to that runtime while moving forward porting stuff. But where to start? Is what's missing documented at all, or is it the knowledge of a few devs?
Back to vim. I'm using the OW version and it's pretty good. I was using it as a porting exercise, knowing that if I'd checkout the commit before the os2 support was removed, it _should_ build. It did. But as you confirmed, console appears broken in many ways. I wanted to port console apps, so unless the libcn is fixed I can't... and no problem to debug and eventually contribute to libcn, yet back to my original comments, the setup time is somewhat time-consuming and error-prone.
Still, I want to give credits to the folks at BWW who documented a lot. Now it needs to reach next level. One toolset for porting, one toolset for legacy (PM and DOS or win32s ports) and reduced friction setup for both. I'm probably oversimplifying, still I think a hobbyist purchasing ArcaOS should be able to code in less than an hour, that's not really the case right now.