Zini wrote: ↑19 Jun 2018, 16:15
Well, I am not happy with this turn of events. I elaborated extensively why I think it is a bad idea. But trying to force my own opinion though is unlikely to result in anything good in this situation, so newscript and most likely Lua it is.
Still, don't give up now, though. You're in a position where you can be fiercely skeptical of any step being taken. If this team was led by someone who was rushing headlong into what seemed to them like scripting paradise, that would pretty much surely result in a bad product. You're critical, though, and that makes me feel safe for this part of the project.
(I admit to not being selfless here - I find the discussions on this subject fascinating. )
Okulo wrote: ↑19 Jun 2018, 19:26
Still, don't give up now, though. You're in a position where you can be fiercely skeptical of any step being taken. If this team was led by someone who was rushing headlong into what seemed to them like scripting paradise, that would pretty much surely result in a bad product. You're critical, though, and that makes me feel safe for this part of the project.
Seconded, the reason that any considerations can be made towards any improvements to oldscript is because of the excellent leadership of Zini, without which openmw would most certainly not be in as excellent shape as it is today. The critical look on the project has been essential in steering the project in the right direction without which many would have been working on superficial features rather than essential cornerstones of the engine, and without which we may not had even the skeleton of an editor.
I mean, making a scripting language from scratch isn't actually hard, it just requires the technical know-how. The reason Lua is popular is because it's premade and very fast, and straightforward to integrate into existing C/C++ projects. It's not actually a "good language" or anything. It would probably be somewhat annoying to dehardcore OpenMW in Lua; it would be better as an OpenCS plugin language or something.
Glad to see the project move in this direction. My offer still stands. I want to see OpenMW be less of a headache for modders, and I think if we worked together to make a compatible API it would go a long way. I guess I'll see if that's something you guys are interested in once it comes out of private committee.
You are welcome to join in on the design team, once we are ready to move forward. Just want to point out one thing: Even an attempt at compatibility would make MWSE at most a subset OpenMW. There are improvements on the post 1.0 roadmap that just won't be possible without full source access to the actual engine.
We can always shoot for feature-parity... so that what OpenMW supports could also be supported by MWSE, I wouldn't put it past NullCascade to try support things even it's just a stub API so that mods could work on both OpenMW and Morrowind+MWSE. By work, I mean at least doesn't crash or refuse to run. I would suspect that MWSE would warn users at run-time that some extensions are OpenMW specific and wouldn't apply to Morrowind.
Am I over reaching here? Or am I thinking about this the wrong way?
It's also what NullCascade keeps pointing out that there are still mods being written and that they are increasingly using MWSE. It isn't a 1.0 goal, but perhaps a post 1.0 goal can be MWSE scripting support.
Then we can begin closing the gap between the two.
For example a compatible API for MWSE-Lua and OpenMW-Lua?