Page 1 of 1

How much of a problem is Irrlicht right now?

Posted: Fri Oct 09, 2020 07:58
by Zughy
Hey there,
I was wondering how much of a problem is Irrlicht right now, as I know the engine is abandoned and MT has a personal fork in its own repo. Also, I've been hearing people complaining about it since I've joined here; so a few questions:
  • isn't a problem to work on MT AND on the other engine? Like, MTG has been dropped, but if then another project to maintain pops up, things don't seem that different
  • would it make more sense to detach from Irrlicht entirely (and using, idk, Chromium Embedded Framework for GUIs) or to adopt a solution like IrrlichtBAW which is an updated fork aiming at being threadable and relying on Vulkan?
  • most importantly: have the core devs made up their mind yet?

Re: How much of a problem is Irrlicht right now?

Posted: Sat Oct 10, 2020 20:04
by v-rob
In my corner of the engine, GUIs, I want to abandon it. Irrlicht's GUI can be worked with (and we have been working with it for quite a while) but it's got a general lack of features and quite a few bugs here and there. Specifically, I want to replace it with a complete CSM Lua-side solution which is the most extensible solution, but failing that, I'll probably look around for other solutions than Irrlicht's existing GUI system. It's not like changing the GUI system is very much of a problem anyway because basically all of the GUI code is going to have to be rewritten anyway.

But I do have some other problems with Irrlicht. For instance, it has been brought to my attention that formspecs on Android... stink. Some are completely unusable. A big part of the problem is Irrlicht's lack of support for touchscreen input events. It has been added on Android which (as far as I can tell) takes touchscreen events and converts them to mouse events, but it lacks support for things like double clicking, breaking some formspec elements and just bugging out others.

So, to answer the questions directly:
  • Yes, it is a pain to work on two separate things because half of the work is done by the Irrlicht elements, so the other half (other features Irrlicht doesn't have but we want) have to be done outside the element.
  • I don't know anything about replacing Irrlicht in other areas of the engine, but replacing most of Irrlicht in the GUI sector seems quite feasible. I think CEF is way too hefty for GUIs; I would prefer a lighter weight, more extensible solution.
  • If possible and feasible, I will replace it in my sector.

Re: How much of a problem is Irrlicht right now?

Posted: Sat Oct 10, 2020 22:16
by Zughy
Thanks v-rob for the answer. About the more lightweight solution, how about Servo?