Search Unity

  1. Unity Asset Manager is now available in public beta. Try it out now and join the conversation here in the forums.
    Dismiss Notice

New UI in 2019.3 / Fundamental feature releases

Discussion in '2019.3 Beta' started by tweedie, Oct 10, 2019.

  1. tweedie

    tweedie

    Joined:
    Apr 24, 2013
    Posts:
    311
    This isn't feedback on 2019.3 per se so apologies if it's not exactly the right place to ask it, but I'm curious as to why the new Editor UI was selected for 19.3? This is perhaps more a commentary on the new 'TECH' cycle.

    I appreciate the point of the new release cycle is so sparkly new features aren't held back for the next "big" release (annual), but I think you have to be selective about that rule, and new Editor UI is quite a fundamental shift.

    I'm honestly not a fan of the new UI, but that aside, it seems against the philosophy of the release cycle to launch it in 19.3. Overhauling the editor in this way has the potential to break the UI for any custom in-house tool, and it's releasing immediately before 2019 goes into LTS. Even if the new UI doesn't break any tools functionally, it's going to undoubtedly cause a myriad of hidden quirks and inconsistencies across all areas of UX.

    Why wasn't this chosen for 2020.1 so that it could have a full release cycle to reach maturity / allow people to smooth out the bumps with their own tooling, only then going into LTS? Not to mention Asset Store developers now can't easily support 2019 across the board without special treatment for 19.3+.

    I had similar thoughts about the new prefab system dropping in 2018.3 - and that was a really rocky time from a stability POV. I suppose there's a good argument this is precisely in-line with the cycle's philosophy - new features at any cost until x.4 / LTS, but given most users expect to be able to transition a project from x.1 through to x.4, I'm not sure it makes a lot of sense to make such a breaking, fundamental workflow change in the last issue of a release, before it must be supported for 2 years.

    I am of the mindset that taking a project from 18.4 -> 19.1 might be risky, but that I should be able to transition from 19.1 -> 19.4 with little headache, however this wasn't the case for 2018 (in fact, quite the opposite), nor does it appear it will be for 2019.

    I hope this doesn't come off as confrontational, I'm just curious as to whether Unity ever consider delaying features despite the TECH cycle, when they're as fundamental as the editor or prefabs. It's strange to me that 18.2-> 18.3 was a harder upgrade than 18.3 -> 19.1, and wonder if other people have thought the same. :)
     
    JoNax97 and M_R like this.
  2. brunocoimbra

    brunocoimbra

    Joined:
    Sep 2, 2015
    Posts:
    679
    At first I had the same feeling during the 18.2 to 18.3 transition, but after being able to use the new prefab workflows without being afraid of instabilities from other stuff on the 19.1 I changed my mind.

    Now I just think that it was never supposed to be able to easier to update from one version to another - Unity itself doesn't recommend to update the Unity version mid-development, they always warn about making a backup before updating - just that after every 3 tech (new features) releases there will be a more stable one full of patches for 2 years, so now I am glad that such big futures such as the prefab workflow and the new UI are coming sooner than later so that we can immediately have a stable version to check those things out on new projects.
     
    tweedie likes this.
  3. tweedie

    tweedie

    Joined:
    Apr 24, 2013
    Posts:
    311

    Yeah that's a totally fair way of looking at it. I think I have a (perhaps wrong / legacy) way of looking at the annual releases as a single "version" of Unity, when I should be considering them as very discrete things. This is probably due to using the year in the name, which makes each version in that chain feel part of a natural progression. Not to mention, the fact that the last version in that chain is the only one that receives continued support, makes it feel like a natural and intended version to head towards and then settle on for the rest of development.

    This does make x.1 - x.3 versions feel a bit more like extended betas, to test new features, and I think Unity would have to move away from using the year number for me to drop this intuition.

    All this said, getting new features quickly has indeed been fantastic, it's just in production situations where you may be waiting on a bug fix that's due in the next discrete version, but have to suffer the problems introduced by these fundamental changes, that things get tricky.
     
    brunocoimbra likes this.