Hey. I noticed in the roadmap that Unity 2017.3 seems a bit lacking in features compared to previous versions of Unity 2017 and I was wondering about that. Is it meant to be a version that is focused more on stabalization? Are there any more features coming? Is the beta period going to be shorter than 2017.1/2 ?
Not that I do not appreciate the hard work that the Unity team put in this engine, but it does raise my curiosity to why there are a lot less major features in comparison.
Firstly the release needs to be finished by the end of the year, hence two months work, Christmas period be quiet. Second I think everyone wishes for bug fixed over features… There lot of stuff missing from road-map such as stability for new iOS11/Android 8.0 and of course High Sierra… I also feel there been a lot of internal changes within Unity, but Im waiting for some solid replies…
2 Likes
There’s a bunch of things to take into consideration:
- The roadmap tends to get updated as the beta progresses. In the early 2017.2 betas, its list was much smaller than it is now
- From the looks of it, 2017.2 was a particularly tough release. There were lots of bugs and things that changed drastically. Unity is probably playing it safe with 2017.3, and making sure everything is stable
- There seems to be a lot of hidden things going on internally. For instance, 2017.3 includes a new PhysX broadphase algorithm, as well as kinematic rigidbody contact pairs. This is nowhere to be seen on the roadmap, but it’s a HUGE feature. Also, you can see changes being made for the C# Job System and the HDPipeline here and there. These are absent from the roadmap too
3 Likes
c# job system is likely taking much of the resources, and rightly so as it will literally change everything. Total performance game changer.
Since we moved to subscriptions we’re no longer have to hold features back for specific release it is just natural that some releases have more new features in them compared to previous ones. It just means that teams working on whatever features didn’t feel like their work was ready for this release and then aims for the next one which will be just 3 months down the line for you guys.
2 Likes
I’m really sorry to say but feeling of entitlement for new features on every release is very dangerous from Unity community. I would instead really hope there is strong emphasis on stability. We have been trying to keep up with upgrading to latest Unity version and over the recent 2 years it has been impossible due to regressions.
7 Likes
I’ll second this. Keeping up with API and system changes when maintaining even a small asset on the store has been a real nightmare throughout the 5x lifecycle. I can’t even imagine how the bigger assets providers pull it off.
And with the release of 2017, the editor itself has become clunky, cumbersome, and even downright broken. I’d really like to go back to the days where I don’t have to stare at Unity for a cumulative total of maybe an hour out of every day just waiting for it to react to button clicks or when switching to other windows.
2 Likes
I think the situation is really difficult to judge.
Things like the C# Job System, HDRenderer, Mono/.Net upgrades, Playables API, and expanded PhysX integration are all worth going through a few rough patches for.
2 Likes
Fair enough. I’ll admit that when the introduced multithreading to the backend of the UI it really wrecked a lot of stuff for me but now that I’m through it, its not bad at all. Early in 5x they pulled a lot of API rugs out from under me so I’m still a little bitter I guess 
It seems aggravating to suffer through several broken versions of a new feature before it becomes even remotely usable but I suppose in the end it will pay off eventually. Lately the API has greatly improved and the runtime seems much better now.
However, these days the editor itself it my biggest cause of pain. I really don’t exaggerate when I say a good sixth of my day is just waiting for the editor to register something.
2 Likes
Yes, occasionally it is perfectly fine as long as we don’t need to pay subscription for system that cannot be used to its full potential. Instead of saying “No guys, your latest Unity version is garbage, I stick with the previous major release until you fix your shit” … Voting with the money. Now we are “forced” to keep paying subscriptions even if we have to keep using old version without accountability to have a stable development environment. I prefer subscription system for sure, even if it removed some accountability from Unity devs, very happy to pay for the service.
I am playing devils advocate here just to make a point. I fucking LOVE Unity and the good work the team is doing. Yet I feel obliged to raise an issue that it is not all rose pedals and soap bubbles. It is a serious issue if new stuff keeps coming and old stability and regression issues are piled on the development debt. It will be devastating few years down the line and we, as community, Unity and us, will have hard times ahead.
Not all is doom and gloom, as said I am very happy despite these inconveniences with the genuine hope they get resolved soonest.
Enforcing artificial release schedules so you get your features more regularly will cause the company to slide backward to the old days that Unity has tried so hard to escape from.
Accepting an irregular cadence is fine, you can’t game subscriptions anyway and the sum of all updates over a couple of years is worth any kind of subscription, even with big fat zero new features.
Assuming everyone here actually pays for subs, that is.
2 Likes