The Future of Unity Feedback

“Unity already knows what everyone wants” is not based on anything objective, is it? They can say “I think people want feature A”, but how can they be completely sure? Based on different people seeing every forum post and somehow using their combined subjective interpretation of what people are saying? Maybe votes are not perfect, but at least they can give an objective measurement of what people want.

5 Likes

Or I’m using my long-time experience with Unity and its various community features. Some of them aren’t even big issues, like polymorphic type serialization. Some of them are more complex, like a way to keep infinite loops from crashing the editor (though even an asset exists to help with this). Some of the things marked complete, like the improved terrain features but even those barely exist because the 2019 terrain improvements are kind of a joke. Other things have been under review since before I started using Unity, like the ability to resize the text in the editor, a basic accessibility feature.

I have an honest to god hard time thinking that the Feedback sections votes mean anything at all.

4 Likes

I think it would be nice if somebody from Unity could clarify whether the feedback votes are actually useless, specially if they are putting a similar feature in place in the near future.

1 Like

https://feedback.unity3d.com/suggestions/goat hope this survives.

And what does that even mean?

It means feedback was just ultimately a crapshoot with poor dialogue, subject to ridiculous demands with healthy support for said ridiculous demands, and had zero back and forth dialogue which IMHO is essential when trying to work with ideas.

As you can understand, to me it says nothing to see a single joke entry, added by yourself, and with Unity members going along. If you can honestly say there’s hard data indicating the signal-to-noise ratio of Feedback is bad, I’ll take your word for it, but then more questions come to mind (questions that Unity should also have to ask to themselves):

  • Is there a reason for the issue? A problem with the backend system, Unity neglecting the section, other? Does it have anything to do with the core idea of Feedback?

  • Is there evidence that having the feedback in the forum would improve the issue? After all, the members on both places are basically the same. If yes, again, what are the reasons that would explain that?

  • Back-and-forth dialogue is definitely important, but there’s a section of comments for each feedback entry. Is people not using it? Why?

Surely it would be a good idea to improve some parts of Feedback, like the comments section, being able to edit your entries, etc., but I still don’t see how using the forums instead would improve the signal-to-noise ratio, and it seems to me that several nice features would be lost or worsen, like providing Unity with objective data about the desirability of features (if they actually don’t use that then shrug and sad), search filters better tailored to feddback, etc.

3 Likes

Well it’s more of a case that unity developers are saying the forums give them really good data. It’s proven right there. While feedback gave them nothing at all except a vague notion people like goats, linux support and features that are already built in or easily done with a few lines of code.

On forums, particularly over the last year, there’s been an insane amount of feedback, 1:1 with unity devs. Try it yourself for proof: https://forum.unity.com/categories/betas-experimental-features.86/

Where there wasn’t any proof that feedback did anything useful, after several years. The voting points system was laughably bad, because you couldn’t put your support behind one thing and not another in any numerically-meaningful way.

Sure, everyone can fix or not fix or make 1000 bot accounts but ultimately it’s probably just better to run deep analytics through the forum, and get a bigger idea of what people are really asking for, and that’s without rigged votes. Just don’t change the forum into something like lithium with terrible usability for developers.

We are missing something though, and I’m sure @Buhlaine will be giving more details how newer feedback takes place closer to the time, so I’m going to be patient.

1 Like

The central things are:

  • Votes are a good idea. Limiting votes is also a good idea. Both of those things means that people prioritize important things. Having just 10 votes ever is a horrible idea, because that means that people won’t engage anymore when the 10 votes are spent. Having to remove votes in order to re-vote was poor fix, and I suspect that it probably didn’t happen much.
    I think getting more votes over time is a better course of action, if a voting system is implemented. Also consider only giving votes to accounts that are somewhat old? That’d both reduce spam, and increase the chance that the people who vote has used Unity at least a couple of times.

  • There needs to be a direct integration into the forums. If a dedicated feedback system should exist, it should be a subforum with special features. The forum has a bunch of things - like notifications for replies, proper quotes, etc. - that’s convenient for suggestions. It was also a bit of a disaster to have to log in again to get to suggestions*.

  • You need to handle issues that’s got a large amount of people interested, but you can’t do. Feedback’s most voted is littered with issues that’s got thousands of votes, that hung out there for years, with no Unity response. This has a bunch of bad effects, but the primary bad effect will be that people who are heavily engaged in those issues will lose faith in the feedback system.

  • In the end, I think it’s better to not have a feedback system (and stick with just dedicated threads and other, more manual feedback mechanisms), than to have a feedback system that doesn’t work or works poorly, because a bad feedback system just breeds resentment.

*By the way, why do we have to log in again everywhere? I’m logged in on the forums. I’m not logged in on Answers. If I log in on Answers, I’m not logged in on Feedback. If I log in on Feedback, I’m not logged in on the issue tracker. If I log in there, I’m still not logged in on the front page (https://unity3d.com/). I don’t have any specific cookie blocking going on.

6 Likes

I don’t think votes even make sense. In a democracy, they don’t make sense. They’re flawed, broken and you’ll get people with very little intelligence or knowledge voting for things that seriously don’t make sense.

In the real world, it’s the best we have.

In our development world we can have a benign dictatorship that cares about us. this is better because we’ve got AAA developers at Unity making decisions about what we want for making games. These guys have had years in the trenches with real AAA titles we’ve all played.

Having unity still have votes means all that think-tank brainpower at unity is dumb and not used properly. Why would any engine in 2019 have votes for core capability? If voting was meaningful, DOTS and SRP would flat out not exist

Masses will all vote for the thing that is inefficient but easy as possible, and that’s why the world had problems today. Problems software development really doesn’t need or want.

But votes IMHO do have a place - just for supporting materials like standard assets, quality of life enhancements and so on. For this kind of vote even forum polls suffice.

3 Likes

I don’t know, man, Unity does sometimes create products that nobody needs, and that nobody would’ve asked for.

If there had been somewhere we could vote on “what products would we spend money on or buy pro over”, I’m pretty sure “a custom source control system” wouldn’t be very popular - yet Unity must have spent a ton of money and man-hours on Collaborate. If there was a vote on “how could we improve the community”, nobody would’ve gone for “replace the successful jobs forum with an ugly, mismanaged LinkedIn clone”, yet we have Connect.

But I totally see what you’re saying about core engine features. Core features like DOTS are kind of hard to imagine. In addition, figuring out what can be done and what’s not viable depends largely on knowledge that Unity has that most of the user-base doesn’t.
I don’t think that means that these things shouldn’t be put to a vote, just that the voting needs to be done differently.

A big problem with the current Feedback page’s votes is that it’s just flat votes to spend on everything. So a simple feature that’d take a day gets votes from the same pool as things that would take years. Categorization of issues would be great. A list of things that Unity allows us to pick between would be really good, as that’d allow both focusing the user-base on a few issues, and give real side-by-side choices.

I think a good process for getting feedback on core capabilities would look something like this:

  1. Unity have schedule space ready for a team to work on a new, large feature in the near vicinity.
    Example: The scene visibility tools is about to be done in a month or so, so Unity’s project managers needs to figure out what that team should do next.

  2. Unity checks internally and with key partners what features they think would be a good idea to work on, that matches the workforce available.
    Example: A bunch of people internally at Unity want to overhaul Additive Scene Loading. Some others wants to work on the “Save changes in play mode” feature. A key partner that Unity is talking to is asking for upgrades to the Navmesh system.

  3. Unity puts out a call for response for what they should work on next in some field - Essentially “we need to decide what the next big project for some team should be”. A list of features Unity thinks is good ideas is included. The community are asked to say what they think about those features, and if there’s something else within that field that they’re interested in.
    Example: Unity makes a blog post announcing that they want feedback on what to work on, and lists the Additive loading, Play mode saves, and Navmesh improvements.

  4. The discussion gets to live for a while. Popular community suggestions are added, suggestions that the community don’t really care for are dropped.
    Example: It becomes clear that a lot of users really wants some new built-in components for the UI system. Nobody really talks about the play mode saves. So the UI system changes are added to the votes, and the play mode saves are dropped.

  5. A vote is held to see which of the suggestions are popular. A call to action is spread over all of the channels for people to come vote for them. Unity uses the vote to inform themselves on what features the user base thinks would create value for them.
    Example: Most people vote for the additive scene loading feature. Unity takes that into account.

I think a process like this would help Unity make informed decisions, and improve confidence with the user base, without sacrificing the ability to make smart decisions about what direction to move Unity in.
Of course, it would be really critical to make sure throughout the process that this isn’t a promise to do anything else than strongly consider the community opinion.

Note that the examples here make wild assumptions about what teams could work on what features - I realize that it’s probably not the same people who would work on navmesh and play mode changes and UI. I just had to come up with something.

8 Likes

I’m rather skeptic that analytics on the forum would give accurate results. In my opinion, votes are, let’s say, the least flawed approach. I think @Baste 's ideas regarding votes are quite interesting to improve them. But as I said earlier, I don’t see votes implying an obligation to work on a particular task, but as hints that give Unity more information to decide on what to spend more resources from the pool of possibilities that align with their vision. That wouldn’t hinder work on interesting new core features.

And they will have great feature ideas, no doubt, but I don’t think that alone is better than the combined wisdom of their knowledge plus the knowledge of <high_number> Unity developers out there; plus their real-world needs, because Unity may make a few demos and games, but not enough to cover all real-world scenarios. And also, when you are working all day making something, sometimes its flaws end up being invisible to you.

Yes, please, some feedback from Unity would be great, even if only to acknowledge they have read the suggestion.

1 Like

Alrigh, for now I have moved my suggestion from the Unity Feedback website, to the General Discussion forum.

1 Like

Until we have clear guidelines I suggest feedback be limited to:

  • Do post feedback in experimental forums in relation to features only available in those builds (read the notes for a specific feature/build/forum there)
  • Don’t post feedback in experimental if it’s not related to an experimental build or alpha/beta feature set.

Posting in General discussion is a last resort if you can’t find a home for it. It’s better though, to find what unity is talking about already in one of the experimental topics, and add to that. It’s right in front of the developer then and very related.

Prefix new posts with [Feedback]

Brand-new feature requests are actually quite rare if you look at the depth and scope of the experimental forum.

Finally, there’s the meta forum for chatting about forum itself.

2 Likes

My request I think is not related to any experimental features, that is why I posted it in general as I do not know where else it best fits.

https://discussions.unity.com/t/733627

I have renamed and prefixed this existing topic in general discussion. Ideally this topic would be in a launcher thread in experimental. I didn’t move it so people can see those possibilities and where best to reach Unity staff.

I’m just adding structure and value while keeping things open (none of these are hard rules btw) - just continuing dialogue with everyone until the community managers at Unity figure all this out which is kinda ongoing.

For now, finding the most similar discussion and adding your own value to that, is a common thing we all already do so just enhancing that might help.

Please feel free to post where you want though as it’s actual feedback and not just random chitchat.

General is not somewhere you want to go for posting suggestions. It’s mostly stupidity and rage, with a liberal sprinkling of the one-eyed leading the blind on legal issues.

3 Likes

Yes that’s true! which is why I was a bit leery of it!

I’ve always considered the Unity feedback portal to be a black hole, and I don’t think it’s possible to kill off a black hole, maybe ignore it.

It’s interesting that you’re going to be salvaging feedback that I would have considered vanished from the moment it was sent into the black hole… it boggles the mind :smile:

yeah but but but

2 Likes

Can it be moved to the Scripting forum? I tried to do it myself but I think I can not.