I think versioning of Unity should be organized differently.
Now there are 2017LTS, 2019.1, 2019.2 and future 2018.4 LTS, 2019.3, 2020.1.
Now, one issue with this is that maintainig and bugfixing needs a lot work and therefore postpones future releases. Features which were planned for 2018.3 and 2019.1 where pushed back for 2020.1.
And even then they will be in preview till 2020.3.
I would suggest the roadmap to be structured in a way so that these version-forks are less scattered.
The goal is to have stable and robust fewer versions of Unity so that developers can focus making it better with less frequency of new bugs and most importantly with future compatibility of projects in mind.
I was thinking that there could be only 1 LTS release and only 1 future development release which could happen twice a year.
So, let’s say in 2020, there would be only one 2019LTS and 2020 1st half and 2nd half release.
What do you think about it?
Is it possible to maintain it that way to get fewer but more stable engines?
They messed up 2018.3 by introducing things like the new prefab workflow, among other tiny things that never got changed until 2019.1…thus for me 2018.3 and 2018.4lts will both be completely pointless. They need to introduce big changes in the first half of the year not the freaking last!!! Otherwise this whole LTS thing is just not going to work for many, as you have no choice but to move onto the next years version just to see the feedback suggestion improvements actually implemented, as back porting rarely happens.
No. You misunderstand the purpose of a roadmap. A roadmap is merely a communication tool that provides users of the software with a rough idea of where the product is heading. It’s not actually a tool for improving development. Check the link below for a more thorough explanation of what a roadmap is and isn’t.
Interestingly enough the visual scripting stuff got delayed to 2020.1 …which should mean it will be more finished and won’t end up unfinished/bit broken in an LTS version of 2019.4 …so I think they’ve already got the idea of not putting in really big and new things into the end of the year releases.
Still as far as 3 feature releases in a year or 2, I don’t really mind, and it certainly doesn’t make a difference to Unity added things being supported by the community any quicker either. I mean I look at stuff like LWRP which has been out for a while now and yet 99.9% of all asset store examples still aren’t supporting it, doesn’t really thrill you with wanting to give it more of a go when your project and the asset examples used in it are broken.