Official Notice on DOTS compatibility with Unity 2021.1

When we talk about Unity changing the way they handle package, are we talking policy or technical?

 

Making such an announcement and then let more than a month pass without any life sign is just sad. Please either say something with substance, or nothing at all.

 

The thing that is upsetting is that they say "please, ask questions, we will answer." then just disappear for months on end. When they do come back, you basically get "Hey, for real, we're here! Will answer in just a second!" then months more of literally nothing. As a business, It's disrespectful to your customers and feels like a slap in the face. As if the community that helped build Unity through countless hours of detailed bug reports and repro project submissions means nothing to them anymore.

It's one thing to keep quiet about things that are upcoming and in current development and not ready for release yet, it's another to encourage people to ask questions, say multiple times you are going to answer them, and then do nothing of the sort. When you repeatedly fail to keep your word, even for something as simple as a few forum updates, it becomes hard to trust your word on anything. It's getting to the point where no matter what is said, I just think "right, I will believe it when I see it."

Something as simple as "Hey, we know we promised you guys updates, and we know we have not delivered on that yet, we apologize. We have not forgotten, we just can't do it quite yet" would be huge and go a long way, but you literally get nothing.

 

I think, there were no pressing issues in Martin's opinion. And also there was no new information he could provide to the old questions. Simple as that. Also, I believe that constant complaints will fix nothing.
To be honest I'm also disappointed there was no new information at GDC in July. No wonder, such prolonged silence builds up frustration in the community, but I want to believe there is a reason for this. I think we will be positively surprised when the update comes. Like it could mean integration with more (all in the best scenario) of an old API just like existing Raycast or Mesh API. It's of course just my wishful thinking. But from all information provided next update probably will land and the end of the year when UT can finally sync the DOTS version with Unitys.

 

Both, policy normally comes first. I believe it's mostly changes on the ui layer, and some pipeline changes with what makes it to public. I believe they cannot really speak to the public as much due to the policy changes perhaps.
While we are talking about the 'suprise update' from unity that comes in the future, I'm hoping we can push all things from parallel jobs, like updating meshes there, audio, textures, etc. Updating GPU data there too (compute buffers) would be cool. Unity's engineers did mention in various places that is what they hoped for. I'm also hoping for better dots tools, like others have mentioned, the old tiny editor. And of course, I'm hoping for WebGL support, us indie developers really need web builds!

 

If there is no news about DOTS in the next two weeks i'm moving to Unreal!!!!!!!!!!!!!!
(just a joke i want to die please kill me)

 

I get asked by my publisher if I plan to support DLSS and things of the like (that are bound to HDRP 12 / Unity 2021.2) and it feels weird that the answer has been I am not sure for six months now. We're kind of stuck on 2020 LTS and have no idea what to expect next.

Will 2020 LTS DOTS projects be able to migrate rather smoothly to 2021 LTS? If there is no answer to that at the moment, is there a vague ETA to expect some information on the future of DOTS? In a week? A month?

 

I think you are safe assuming 2021 LTS will not be supported. Maybe 2022 LTS who knows you cannot run a business based on wishful thinking.

 

The best case scenario is the end of the year / early 2022, but there is no definitive answer if that will be the case.

"'Yes! Why is that, Captain Ron?'
'Nobody knows.'"

 

Well, what feels weird to me is that you are using an experimental/preview tech on a project that you have a publisher for .
We can't complain about Unity for providing access to experimental tech, clearly stated as "not production ready". We are responsible for what we do with such access.

 

To be fair, a lot of us got excited about DOTS several years ago at the 2018 keynote, and further invigorated by the 2019 keynote. I can't blame them for starting a game on experimental features with the hopes that the tech would be production ready 3 years later when the game is closer to release. If you can't depend on highly promoted features being ready 3 years in the future, then that's a real bummer.

 

Actually... unity suggest us to use dots netcode in 2021Q2 release games...

https://blog.unity.com/games/navigating-unitys-multiplayer-netcode-transition
they updated this 2019 blog in 2020 April...and said :

• we’re targeting a production-quality release in early 2021 for DOTS-Netcode and UTP to be sufficiently stable and full featured. By this time, developers should have a fairly stable API surface, tech stack, and better documentation.

(and this already broken image)

i have some trust issue in unity now, when they release something, i'd like to told myself "oh, that's cool, i'll check it 5 years later"

 

Yes, yes. i do have trust issue in unity now. Burst? DOTS? Radicular slow mesh pipeline. Radicular slow culling. Radicular slow srp batcher compare with ue. Yes, yes. You have to DIY.

 

This thread is cursed and should be deleted at this point.

 

Taking risks in business might feel weird to you but we've been doing that for 10 years now and are doing really well. For our first IP, Kona, we started on Unity 4 and upgraded to 5 for PBR shading and Enlighten even if all these things were experimental / not suitable for production. This made the game look way better in the end and helped Unity stabilize the engine as we caught and logged issues in the process.

I am aware of the risks of using DOTS. By the way, switching back to Mono isn't that big of a trouble either. The point here is that experimental shouldn't mean "not giving any news for months" to developers. A roadmap is crucial as a lot of businesses rely on it to operate. Even if it is tagged as experimental, DOTS was meant to be the backend of all the new stuff in Unity from animation to world building. DOTS 0.17 is stable, works on all platforms with no issue, is performant by default, so intuitively we were thinking it would just get better [as it always has been with the new stuff in Unity].

 

Sure, go from not replying to simply sweeping it under the rug and pretend it never happened, that will surely win back the trust of our users...

 

Speaking of cursed, perhaps Unity devs are trapped in an alternate dimension and can't communicate with us properly? Do they need assistance? Are they frantically trying to post but it doesn't appear in our dimension?

 

Being a publicly traded company, the lack of giving out timelines and feature promises isn't necessarily illegitimate. There are certain risks with making promises that they may not be able to keep. Imagine if they were to say "DOTS 1.0 will come out on X/Y/2022" and X/Y/2022 comes and goes without DOTS 1.0. Or imagine that the promised performance metrics are not at all close when it does arrive. The result be that investors are annoyed and the stock value drops (regulators might take notice too).

All of that said, I certainly wouldn't mind a "hey we're still" post once in a while.

 

Several folks have said similar, as though concern for publicly listed companies and 'statements' about direction, objectives and schedules for software is a thing. It's not, certainly not for stuff listed as alpha, preview, experimental etc.

What's usually far more concerning to the stock market, when it comes to software tool making companies, is staleness of product, users becoming discontented, dissatisfied and disappointed by problems with the software AND a rival showing great potential.

That's where we are. Unreal 5 is showing enormous potential with two game changing additions to their pipeline.

The stock market knows that digital users of a product that has digital alternatives are able to switch, and that sentiment can swing quickly and strongly from one tooling to another.

Perhaps Unity is desperate for DOTS to provide a counter to the potential of Nanite and Lumen and Unreal's increasingly easier to use (out of the box) features and are waiting till they can make some bold claims about it.

Otherwise the silence is ominous, in and of itself.

But what's more worrying is that no other team within Unity, making anyone of the various systems that can and should benefit from DOTS, seems to know anything about its progress either, and are simply continuing their slow marches forward, whilst others (like Audio DSP and Tiny) look like ghost towns.

 

I feel this was a pretty clear position by Unity3D on the problem.

They've obviously hit a fork in the road when it comes to DOTS backwards compatability. Those who are in the space where the backwards compatibility will not yield future tech stream updates. Sucks.. either refactor or hold until they think of a way to bring compatibility back.

Those who aren't, play on.

Are there risks on both sides, sure. Does that stop you from shipping a game right now. No.

Tantrums aren't going to win, as a product manager you immediately see threads like this and leave them alone. As to answer on here is breaking into jail (easy to break into, hard to break out).

 

I can understand the strategy, but Pandora's box is already opened, they can easily shut up everyone here for as long as they need by giving a hint of progress, direction, update, or timeline.
If Unity's intention is avoid tantrums, they wouldn't say they were looking to provide more info and then not, with no reason why they haven't come back. Naturally the community is going to speculate(sometimes wildly) if it appears that you have been suddenly silenced. No Mans Sky devs used this strategy correctly, after the initial announcement of an update to the community, fully committed to silence, the fans eventually turned to conspiracy anyway, but it could have been far worse if devs would occasionally break the silence but nothing came of it.

Martin said they would go through questions, one month and not a word, the damage could easily be undone by doing giving one of the above, or saying "Sorry guys, I jumped the shark, we can't answer yet". If they can't talk about compatibility because they don't want to over promise there, where is .18? It was said to be being worked on in the OP, it's been 8 months in a schedule that was getting updates almost every month. Was it de-prioritized for compatibility? What's wrong saying that? Even if that is the case, by communicating(even now), we would be relieved and all would be 100% forgiven(I promise). Best case, we could even do their job for them and tell others what to expect instead of "I dunno, maybe we'll know in a year or three, pretty weird huh?".

If I were interested in packages/libraries for a project, this is one of the first things I might look at. Is the package complete? If not, when was the last update? Why is there a big gap? Do the devs have a roadmap? "Oh man, the community is pissed off, devs suddenly can't communicate, and there is a lot of FUD here, I'm staying far away and probably never using this." For a company that relies so much on it's community, this just isn't good and is worrying.

 

So once they say nothing you all lose your minds and start drumming up conspiracy theories around whats going on based purely on speculation. Thereby killing off any attempt confidence around adoption of the thing you hold so dearly - DOTS - and then wonder why its now a self-fulling prophecy filled with negative emotional charged adoption stalemate.

The answer is then "its his fault, he didn't finish the conversation".

I dunno. I'd not take the conversation either.

 

The first real complaint was over a month after the comment was made, so... yes, that is strange? A month is a long time to not converse while the conversation is positive. If he's that busy elsewhere, perhaps he shouldn't bother? It seems natural that people would complain or get worried at that point if it's something they're excited for, or might have expected to be some kind of PR.

 

I feel the answer to that response is on the front page of DOTS

As we’re rebuilding the core of Unity with DOTS, we’re continuously adding new features. Here’s an overview of the essential DOTS packages we’re working on at the moment

You can't hold someone accountable for a conversation that you yourself aren't lisening to is my point there. I feel the messaging up to this point is been pretty blunt and clear.

With the release of Unity 2021.1, DOTS is at a crossroads in terms of compatibility with the currently available Unity versions.

Again. I'm not misguided that they aren't at a place to make any firm bets. When they do, they will deliver a message one way or another. Putting timelines on something you can't commit to is far more dangerous than saying nothing at all.

If you have nothing concrete to offer up, silence is actually the defacto standard for situations like this.

At Microsoft we made it our backbone strategy for tech shifts, and it still is being deployed daily.

 

Quite a few of the questions asked weren't necessarily timeline based. But I'm sure most would agree that silence overall might have been better if they can't answer any questions at this point. Unfortunately for the sanity of this thread, the tease and vanish has happened before and naturally causes wild speculation and frustration until the air is cleared.

 

I don't mean this to be rude, so please don't take it that way.

In a gaming engine space, wherein the majority of the output is creative entertainment and enlightenment for our fellow humans, there are very few reasons to hold Microsoft up as a good/positive example of anything, ever.

In this case, even less, because they're a functional monopoly where they're not a literal monopoly. Not merely a tool maker and not existent within a highly competitive market, nor the underdog in that market attempting to rewrite the core of their tech, and stumbling along for years.

Because of why and how they've been successful, their media and narrative control, and what they do with success, we should be even more skeptical of Microsoft, and everything they do, and how and why they do it.

Google, Facebook and Amazon might be the only potentially worse examples of software organisations. Not coincidentally, they're all monopolies, too.

 

QUOTE="april_4_short, post: 7376543, member: 7785587"]I don't mean this to be rude, so please don't take it that way.[/QUOTE]
I have very thick skin, be rude if that helps you express. I don't mind

Why? you're using the fruit of our labour.... mono... c# ...

Not sure if you were aware of the time Silverlight, which was a product i also managed. We were up against a monopoly in the market - Adobe Flash, we were competing for the same space - even the name spelled it out "another word for a flash, is a silver light" ..

Why is that important - well we paid Novell at the time work on Moonlight which was the precursor to Mono which is now what you have in your hands with Unity3D. It was the first time Microsoft enabled .NET CLR to be deployed on everything but Microsoft technology.

Being the underdog we still followed the same playbook Unity3d is effectively playing now. We also were the foundation of Unity3D as you see it today given we also were all haggling over at the time Indie Game Developers hearts/minds. Remembering Unity3D's core DNA did derive from Flash Development as well.

hopefully i've demonstrated that even "monopolies" don't actually win. Silverlight lost the fight, we had more money, more developers and more resources but we still lost. Internet Explorer had the same, now its a meme to be using Microsoft Edge over Chrome.

The tech community aren't dumb, they are more self aware than we give credit for. If a company can't offer the needs of the day, someone eventually emerges to solve the problem. As the demand is great. You can't media buy your way anymore into owning and dominating that conversation.

If that were the case we'd all be using Unreal Engine

 

Mono has an interesting history, you should look into that before claiming it's the fruit of Microsoft.

C# is Java for Windows, please. And that's almost a verbatim parody of the original request/instruction that borne it.

 

I'm aware Mono existed prior to Moonlight what the implication was that when Microsoft opened its code base and resource up to fully support Mono by way of Moonlight this in turn gave a lot of momentum behind the project. So ease up on the venom there bud

Not sure your point on C# and Java history lesson here or more to the point who honestly cares? Its what we use in Unity3d and if you don't like it ..well.. tough?

 

You've missed the point, and the joke. And gone onto thinking it was venom, instead.

Look at it the way it was intended:

Mono isn't fruit. In a flora metaphor, at best, it's a weed.

 

Please stay sane and don't derail the thread.

Use 2020 LTS if you're using DOTS. Otherwise - wait for news / release of newer version that is compatible.
Don't expect it to land soon™ and you'd fine with both planning / business.

 

I have used Firefox for years, I can't even stand Chrome anymore.

If it makes you feel any better, my work PC's are all Windows and when I use them, I actually prefer Edge in its current form over Chrome.

 

So are we supposed to use this system at all or not ? Because cant limit development in Unity 2020 for sure, upgrade might be needed, so is a vast risk to use it for anything, making it totally useless to me.

Is there any kind of guarantee that will later work in 2021 and wont need radical changes (so wont be like remaking everything from start and totally differently, like it happened between first and later releases of this system)

Otherwise i suppose was a good call to remove the system from package manager, so noone uses it as it is not for general use and might not be for several years, if ever come to that.

 

Not really, it just makes it harder to install for those that use it. Protecting the users from themselves this way is akin to removing the shifter of a car because the driver might intentionally shift into reverse on the freeway. I don't understand why it matters if someone installs it when it's not ready, because when the dev tries to use it they will need to read the documentation and will come to a conclusion whether or not to proceed.

Just mark it as beta/experimental and be done. I have several packages installed that I was interested in but may or may not want to use when the time comes. The danger of using experimental packages is implied (imo).

 

I think the issue is one of false advertisement more than anything else, Unity had done all this huge talk on those systems that were apparently 10 years away from any final version to use concretely and spoke for them as they were near final systems to use, with big focus and showcase, than a simple "we also work on some ideas that may never come to fruition to use"

Same as with pipelines i guess. The issue is that some may try use these systems and invest in them, only to find out they wasted time and energy for something that no longer works and may never do in next versions.

 

Did you get around to trying DOTS/ECS in any of its forms for your Particle System?

Particularly interested in your opinion, because I'm of the opinion that nothing benefits from an ECS approach more than particle systems, because they're basically ECS by default and nature.

 

I have only worked on it with skinned mesh so can enable animations on massive scale units, for particles already have optimizer that is probably faster than this system, so not have plans for that yet.

 

WOW.

Do you think you could port your optimizer to the other systems of Unity, and save us all?

No pressure

 

Even though a Particle system can benefit from ECS, the nature that the sheer amount of data has to be uploaded to GPU will always be the bottleneck. Therefore, particle systems are not fit for ECS.

 

Ya that you should only use stable releases for a production game development is mostly a hobbyist meme. Capable studios are very often on the bleeding edges, because they are capable of dealing with the risks, and of making smart decisions on what they take risks on.

Back to the original issue brought up by @alexandre-fiset of how to align with HDRP. I would just give the publisher more information. Unity pushed out SRP's as stable long before they were that, and now it's causing havoc with alignment to DOTS. A re alignment had to happen here, reality was going to impose itself at the end of the day. We just didn't know the exact shape it would take. If your investors had some idea of how early and raw HDRP really was, they should understand. Either HDRP or DOTS had to take a hit here there is just no way around that, the piper has to be paid.

 

I've started learning DOTS about a year ago and gave up on it due to missing features.
I still very much would like to try it again and experiment, but I'm afraid of finding out later on that most of the stuff are deprecated now more than ever due to the silence that essentially started more than 6 months ago.

The only thing I want to know, is if DOTS is being restructured, updated(features), or if it's on hiatus until they figure things out.
I'll honestly be fine with any of those answers, as long as I would know what is going on before even bothering with it.

 

I understand why everyone is upset. Communication on this issue is dreadful.

I have 2 cents...
DOTS is pre-alpha, you should expect it to be buggy, incomplete and subject to wild change. For the devs, the burden of compatibility promises is just not feasible for DOTS at it's current stage.
If you haven't noticed, Unity 2021 is growing rapidly, probably why everyone is upset about the missing features. But this is the exact reason DOTS cant promise compatibility.
I think that's where people need to stop the speculation. It's just compatibility promises that cant be fulfilled.

For those that have it in the back of their head that DOTS is dying or on hold or whatever, that is just not true. It's in very active development and in its current state is even feasible to make games with if you put in the work as others have proven.

At my day job I work with eclipse RCP. All the docs are from 2008. But everything still works and there are plenty enough features to do what I need to.
Unity 2020 is only one year old and has active bug fixes... it's not the end of the world to use it. Plus you get the entire DOTS feature set, to me it's an easy choice to make.

 

Hi @MartinGram Sadly to tell u most probably Entities package needs to compatible with Unity 2021.2.x and later as soon as possible due to the fix not able to backport to Unity 2020 LTS. Meaning that if I stuck at Unity 2020 LTS I will not able to ship my title.

 

Where is this from?

I have been experiencing a fine resolution issue with deltatime for a couple of years, but don't have the skills to articulate or quantify it in a way that anyone thinks it's real. They seem to think I'm imagining things -- I could be.

Insomnia is a hell of a drug.

 

https://forum.unity.com/threads/cas...led-to-wait-for-fence-2.1133677/#post-7394930

 

Wow, after some time away from dots wanted to start a new pet project on it. Sad to see Unity ECS in such a state

 

2018 - ECS Early version Came Out, yeah! I just Wait
2019 - Demo is Wow... Amazing(MegaCity), I test then I just Wait
2020 - I thought ECS 1.0 will out, GDC Say it... but too many feature not done yet and upgrade is best is best, I just Wait
2021 - After Change DOTS 0.17 for packages detail purpose, nothing Happen, I just Wait
2022 - Unknow
2023 - Unknow
2024 - Unknow
2025 - Unknow

I hope that next time it will be a one-time big update, otherwise it will be like 2077 from 2013 to 2020... but no all same...
The more important is feature must be done -
Animation, Physic, Terrain , network, graphic, audio, visual scripting DOTS
Less than one hard to start..

2018 UNITY: We Hear You, Pls Try it, and Feedback
2019 UNITY: We Hear You, Pls Try it, and Feedback
2020 UNITY: We Hear You, Pls Try it, and Feedback
2021 UNITY: We Hear You, Pls Try it, and Feedback
2022 UNITY: Unknow
2023 UNITY: Unknow
2024 UNITY: Unknow
2025 UNITY: Unknow

 

https://forum.unity.com/threads/production-quality-netcode-moved-to-2021-q2.893017/
Here's the old one:

Imagine production quality DOTS netcode in Q3 2020

 

Almost two months ago, @MartinGram promised to get answers to the questions asked in this forum. That seems a long time to go without an update. It's time to hear from him again, even if it's only to say 'I'm still working on it.'