Anyone think it would be cool to have a stack exchange for feature requests?

General Forums Blender and CG Discussions
21

the problem with that will be:

  1. over time the amount of ideas listed, you will loose oversight

  2. the problem with posting duplicated ideas, because of the amount of ideas posted

I am pretty engaged with Fusion 360 since it was in Beta 4 years now and observed the issue with their idea station.
Having a pretty close connection to the forum and product managers there who communicate user wishes to the dev team I know that they internally sort all ideas, edit ideas, and place them on a tree to visualize what is possible, what fits into the developing resources, what might come later, and what is out of the way or not what they want.

That is why I would rather promote something that functions like a database and not threads.

Fun fact:
Last Dec in Las Vegas at the Autodesk University I got an Autodesk Fusion 360 Community Award for having flooded the idea station the most. They also have a Claas filter when I post so everybody will get altered lol.

22

I only say that if I had somewhere to post feature requests, I would bombard with requests here and there. Now imagine everyone doing it!

23

For such a thing to work (here or anywhere), there are some very specific things that I think would be absolutely necessary:

  • Heavy moderation - There would need to be at least one person (likely more) with the primary focus on reading posts/proposals and vetting them to make sure that they’re not support requests in disguise (e.g. “you can already do that, you just didn’t know how”), rants without substance, or overly vague (e.g. “make it better”).

  • No simple negativity - This would go for proposals and for responses. As mentioned, proposals can’t just be rants about something that’s allegedly broken or unusable. They need to be clear and well-structured. At the same time, feedback on proposals can’t just be “it won’t work” or “it’s a bad idea” without explaining how and why and (most importantly) how the proposal may be improved.

  • No expectation of follow-through - It doesn’t matter if the feature request is the most highly voted or not. If there’s not developer interest, or if the implementation doesn’t match Blender’s purpose or approach, the notion of “sour grapes” can not be acceptable. Feature request sites/lists are not priority lists and they’re not contracts with developers. They can’t be, and for a bevy of reasons. The easiest example: often what appears to be a simple request may require a series of other changes first… and those changes may spawn/facilitate other ideas of their own. We cannot expect there to be any requirement on developers to address the list… and if they do, we should also expect that the implementation may not be exactly as described in the proposal. Of course, communication is a two-way street, but it’s imperative that we remember that developers are their own people… they don’t exist to robotically do our bidding.

  • Developer involvement - We can’t expect moderators to email developers when a “request is ready.” Developers need to be active participants in the discussion of a feature request. Without that, you’re essentially what BlenderStorm turned into (and that site was originally written by a Blender developer).

There’s another one, but I didn’t have a short way to phrase it. I’ll try now, though. It would be ill-advised to expect any feature request discussion to “settle”. From the simple history of requests on this forum, it’s more common for discussions to peter out without any agreement whatsoever. At that point, I suppose it’s up to a developer to decide whether to address it at all.

Anyhow… those are my thoughts on how making something like that might work. It will be quite a bit of work for whomever takes this project on. It’s possible… but it would take someone with a rare combination of dedication, understanding of Blender, decent communication skills, and available free time.

24

Well the bugtracker is a database, so it would be simple to filter on a categeory.
The category future request, doesnt need to be visible on the coders side, until a pannel of “enlighted users” gives the aproval.
This team are not the developers, but are experienced users, who know how blender works, keeping the “dont know how to do it so i fill a request, or i fill a weird wish” out of it. After aproval by this team, the request becomes a candidate that can still be ignored or refused, with reasons such as, “our code isnt developing in that direction”, “to much work”, “no time”, ea… or be aproved “great idea”.

Also it might have some synergy to do this on the same system, it will bring people closer to development.
Once subscribed to a thing, people gets email updates about its development updates, or status updates.

Putting it on sperate sites, creates only more fuzz, more sites, more noise.

Lots of code-managing software keeps it all under a single hat, because its parts of the software cycle (you dev people should know).

25

@Fweeb: one of the longest and most convoluted no-s I have ever heard. Advice: get into politics, you have a natural talent for it.

As far as I am concerned, I file this community feature request proposal in the same box which contains NURBS and colored wireframes and get out.

26

Hahaha. Please read it again. It’s not a no. It’s an outline on how I would do it. And frankly, I actually have an interest in trying. I’d just need to figure out what other projects of mine I’d need to put aside in order to make it happen. Of course, if anyone else wants to do it before I can create time to do it myself, then I’m all for that, too. :stuck_out_tongue:

27

I fully agree on this, which is why I like the way how AD does it.

You post an idea, the moderators analyse it, then this is brought to the development manager and then developers are contacted.

If this should be productive both sides need to participate well, users and developers in an appropriate manner.
I understand that the devs have their own time constraints and not everything is possible for them to pick up, plus there is the
general development path for what already needs to be done to Blender by Ton overlooking that.

  • Developer involvement - We can’t expect moderators to email developers when a “request is ready.” Developers need to be active participants in the discussion of a feature request. Without that, you’re essentially what BlenderStorm turned into (and that site was originally written by a Blender developer).
28

Maybe we could do this in 2 process

  • 1 : Peoples propose ideas and talk about those ideas

Moderators helps to concentrate the ideas, delete doublons or merge ideas

  • 2 : One people or a group make a proposa (with video, picture etc)

Dev’s and the group talk about the proposal, what is possible, what is not and then agrred or not to propose it to the BF for starting the job.

29

I was going to say something in reply but it’s all been said already! Love the amount of interest on the idea. People talking about how it can be done is SOOOO much better than a Jaded user replying that it is impossible. :slight_smile:

Just for experimental reasons i set up an account on ASKBOT and created a free domain trial so that we can play around with the idea of having such a system. My feeling is that with a basic example such as this, people can get an idea of the strengths of such a system and see what works and what doesn’t. At some point I might dress it up a little but for now I have to get back to my work, but i’d like to give it a similar feel to the official blender website if i can.

http://blenderfeedback.askbot.com/

Feel free to go there and post something in the mean time.

It’s very similar to stack exchange but it’s easier to set up a site.

However… it’s not a solution because the best membership is limited to 700 users which does not really encompass all blender users!!

30

in UPBGE they use Git issues Bug: and Request:

blender developers don’t want to hear about game engine requests though, no matter how well thought out they are,

UPBGE has been awesome! (thanks youle and tristan!)

A sensor that has been missing from blender game forever was added tonight!

the most basic of things

if moved---------(decide what animation to play on how you move)

compiledBlenderUPBGE.png1360×768 133 KB

Now I am building blender and testing !
hopefully I will be a blenderDev soon and a UPBGE dev!

side note -> stack exchange is a good idea, but a really good idea is one where people can fund the development of features right there (like kickstart for good ideas)

31

It is always interesting seeing this discussion reappear here on BA. Concerning how to organise user feedback & requests.
Personally I think it is a important thing that is missing in the Blender development model - and something that is holding back the amount of donations to Blender Foundation.

Here is a collection of links with suggestions I made in the past, and some observations from other users :

 [A suggestion to make a user survey](http://www.blenderartists.org/forum/showthread.php?385183-No-more-wireframez!!1!!1&p=2963023&viewfull=1#post2963023)before a Blender 2.8 crowdsourcing campaign.

Could be a way for the developers to test the value of quantified user feedback.

A responseto a post by Campell. About using BF ressources to facilitate user/developer communication.
With quotes from other BA users.

A suggestion on how to make a system with voting on user requests, and voting on the development roadmap.
To get quantified feedback from the users, to mix with the insight from the developers.

And - as I mention in one of these posts - the ‘hub’ for user feedback have to be on Blender.org, and with direct involvement/commitment from BF.
Since BF developers are the ones to say OK to the final code.

32

In a way we do use our tracker like this - we have design tasks, where people who intend to work on a project can collaborate on planning.

The difference is, all the items in the tracker have a clear outcome. - They are errors to be fixed, tasks to completed, designs to be agreed on and developed or rejected. They’re assigned to people and eventually closed.
This means the tasks can be kept manageable by developers and user-module owners.

Feature requests on the other hand can very open ended and don’t have a clear outcome, there can be 1000’s of open requests and likely we wouldn’t be able to manage them (respond, de-duplicate … etc).

I’m not saying this can’t work how your suggesting for some projects, only that I don’t think it would work for Blender at the moment.

If we, for example we take all the suggestions on blenderstorm, and add them to developer.blender.org’s issue tracker… then what? We just have many random ideas and at the end up the day someone needs to go though them and respond. As far as I can see this just moves a problem we didn’t manage to solve in one system, to another system where it will continue to remain unsolved.

33

I completely concur with Razorblade’s suggestion to add a ‘Wishlist’ or ‘Feature Request’ severity to the issue tracker. There is a reason that every open-source project I can name (and all of the commercial development I’ve been involved with) tracks user requests this way. I’ve made the argument before, and continue to believe, that one of Blender’s biggest weaknesses as a project is its inability to effectively collect general user feedback.

The problem isn’t that you’re not getting enough requests it’s that you can’t keep track of them. How many wiki pages, personal TODO lists, and archived emails are the requests you’ve received over the years scattered across?

Wouldn’t it be better to have those requests and TODOs in a database rather than scattered across dozens of pages in the wiki like they are now? Let’s start with the two top level pages linked from the Development page:

http://wiki.blender.org/index.php/Dev:Ref/Requests
http://wiki.blender.org/index.php/Dev:Source/Development/Todo

How many duplicates are there in those lists? Actually, forget that, how about an easier question. How many items are on those lists?

Keeping requests in the tracker also solves concrete problems like the ones you outlined in a post on another recent thread:

These problems disappear if you have a ‘Wishlist’ severity. Instead of closing these ‘tricky’ bugs just assign them to Wishlist. The reports that represent real problems will attract additional comments over time and can eventually be addressed. The rest remain as documentation of priorities and decisions that have been made.

The problem of collecting and utilizing user feedback is not beyond the capacity of the Blender organization.

34

Yeah, The dev page would really be the best in my opinion. Comparing the wiki to the main dev hub i notice a difference in usability, not to mention that requests just don’t seem to belong on a wiki. And really requests are quite comparable to addons, only someone is not able to accomplish it themselves, which makes up a huge amount of blender users. Also EVERYONE who can actually HELP bring a request to realization is there, as it’s an useful hub. And since it’s a place for addons as well, a feature request could turn into an addon and the proof is then in the pudding. Change would be needed though.

One little problem, if this site is only optimized for a certain amount of activity, it’d need to be prepared for the masses. I think with that level of activity, allowing reputable users to have some low level moderation abilities is called for. How you might ask? I think stack exchange had that answer. The more a person’s actions on a public place are accepted by the public, the more trust worthy they are, and reward system can quantify that. Nobody’s perfect, and a moderator may be needed to settle something a high rep user cannot, but this would effectively eliminate micro management, and allow moderators to focus on higher level tasks. And of course if ratings were introduced, then the best requests could be sorted by user interest to help devs decided whether they think it’s worthwhile, that in itself is another level of organisation.

But… baby steps right? Rome wasn’t built in a day.

But ideally, these are the things i’d like to see if requests become a part of the dev site. And I’d be interested in knowing which of these things are most do-able in the near feature! I’ll just try to sort them in order of most realistic changes.

-‘Submit Request’ button. Justified to the far left, with a little explanation for the purpose, a nice little ‘rules’ link to prevent confusion, and a link to all requests.
-A rating system to gauge interest on tickets (at least those submitted as requests) on the preview of the task, i’d show this as a green/red ratio bar to save UI space and match the priority bar, then show the actual numbers when opening the task.
-Rating based sort mode to allow devs and users alike to get some perspective on the most wanted changes.
-point system to encourage users to think critically and post helpful suggestions, comments edits, addons etc.
In particular, people should be rewarded for addons because of how tremendously useful they are.
-basic level moderator skills assigned to respected/reputable users based on their point count, this will cut micro management and provide a motivation for users to be helpful, diligent and active, thus improving blender as a whole. (At least for moderating requests. Sometimes people are just not good asking for what they want and need help improving a question)

My feeling is that the mods are not going to be likely to be able to take on more work, so thinking of them, why not allow trusted users to assist in providing a better experience?

What do you think? I’m thinking of creating a mockup for how it’d look if you like the idea enough.

@pitiwazou

Maybe we could do this in 2 process

  • 1 : Peoples propose ideas and talk about those ideas

Moderators helps to concentrate the ideas, delete doublons or merge ideas

  • 2 : One people or a group make a proposa (with video, picture etc)

Dev’s and the group talk about the proposal, what is possible, what is not and then agrred or not to propose it to the BF for starting the job.

On getting this feedback proposal to the team CERTAINLY! I think this has enough interest that we’ll come up with a neat plan before long.

EDIT:

Pardon my ignorance but i overlooked the token system that is in place and also the subscriptions Obviously they already show the amount of interest on a topic. I was under the impression that most proposals/requests were little sub headings under a hierarchy of obscure submenus.

what made me realise this was this post here: https://developer.blender.org/T45734

But I’m sitting here trying to figure out how it’s come about. It’s not user made is it? I’ll look into it more. I have some ideas but i’m still very much ignorant about the current system.

35

I think therefor a group of people, who are not developers, but can imagine what is reasonable or not.
I mean often people think of complete GUI changes, thats not a short time god idea, but might be put to a gui selection.
But maybe someone has an idea with a bit less impact, and less huge coding impact, this could be something for python or c++
Maybe its something small for the game engine, or a filter, or texturing option etc.
Maybe something that isnt bug, but better behave differently in better way.
Then the ideas endup as a good idea list.

And thats just what it is, a list then of good ideas. Its not something that has to be done, but something that can be used by some coders for inspiration or hints of what might be good areas to improve based upon the user base of the program.

Somtimes such ideas, stay years under the table… i myself in 1993 told microsoft to integrate a web brouwser in their windows 95 startmenu, and then be able to install from there. Well it took them till windows 98, and then they had to remove it, until Windows 8/ and windows 10, they finally got a bit of that idea to create apps for touch screens, etc well at this time my orginal idea got lost.
But in a sense my ideas now got worked out in app stores… (and as a side result a new industry cloud computing went mainstream)
but well …we could have had that 15 years ago under windows 95, if MS had noted my Email about it (i was working as a MS partner then)

Thats why idea lists are usefull, and people should carefully look at such lists, and once so ever pick the good ideas out of it.
And if you endup as a coder with more ideas at the end of the day then code, also then fill the idea list.
Because people forget good ideas, and next day start thinking on other subjects.

I’m not sugesting that each idea should be done, should be coded, but better not waste another 15 years …
Then when a Blender v7.x or 8.x is ready and you want to go to x+1, you take the idea list and look what might be something to go for.
And hopefully those are a lot of smaller better do-able ideas, (that’s also ideally for new starting coders in python and c++ )

36

I’m not a big fan of using developer.blender.org for this either. The Phabricator system isn’t really built for managing user requests, but much more for organizing of general software development. You could for sure use it for this purpose too (as mentioned already, we already do to some extend), but in the end there are still alternatives specifically designed for managing user requests/feedback. Such a specialized system would simply be more efficient in day-to-day use compared to the generic Phabricator system.

For me the most important reason why I’m against this idea though, is that we’d mix two different ‘worlds’ to a point where it would become a burden for at least one of them. I think for us developers, it is important that we have some rather isolated (but obviously still open) place in which we can focus on the Blender development from a coding perspective. A place where we can manage the development of our code base without too much noise from outside.
This is already pretty hard since we have the bug tracker running on the same platform, adding a feature request tracker would destroy this isolation completely.
I’m all for bringing artists and developers together, but we should respect that both need some room to unfold.

How about a different approach: If we were to create an external site for feature requests and it proved its worth, then we could move the bug tracker to this site too. It’d create one place for bug reports, feature requests and feedback, and another for the coding work.

37

@JulianSeverin, moving bug tracking out of developer.blender.org means we’d loose integration (auto-closing, linking to reports from commit logs). The existing bug tracker seems to work quite well so don’t see any reason to move away from it.

@Razorblade, I don’t think anyone is saying not to store suggestions at all.

Having different Blender related sites can be seen as a total mess… or using the right tool for the job.

  • Mailing list works well for more official discussions.
  • Bug tracker works well for issue tracking.
  • Chat rooms are nice for informal discussions.
  • Stack Exchange works for Q&A.
  • Forums work well for general discussion.

If you’re a web developer - and you propose to merge some of these of this into one place… ok,
It’s a big project to undertake, however you approach it.

A place to track suggestions (if someone is prepared to run it) seems reasonable.
So far I only saw evidence that the user-community isn’t ready to run a site like this yet.
On the other hand - its something to experiment with for those who like to try.

@Simon h, the changes you’re suggesting for Phabricator are not trivial,
even taking baby steps someone would have to put a lot of work into this sooner or later. And its doubtful this would be accepted by the Phabricator project - meaning we would have to fork a 420k line project, currently maintained by facebook (over 200 contributors), so its unlikely we would be able to keep up with their developments.

38

The team that develops Phabricator uses it to track feature requests so they must think it works OK:

https://secure-.phabricator.com/book/phabcontrib/article/feature_requests/

Agreed, developing all of the features Simon is suggesting on top of Phabricator is probably not the best use of Blender’s resources. On the other hand, a default Phabricator install includes the ability to track ‘Wishlist’ priority tasks. By choosing to remove that ability from developer.blender.org the decision has already been made to run a custom version of the software.

I’m still struggling to understand what makes Blender so different from all of the other open source projects that operate just fine with a single issue tracker mixing bugs and requests. Commercial developers that want to crowd-source feedback are forced to run multiple systems because they don’t want to make their issue trackers public. They need to accept the additional overhead of running multiple systems, but that’s overhead an open source project can easily avoid.

39

I love blender, specifically a obscure corner of blender, (BGE) and when I wanted help building blender, people pointed at a webpage and then fell silent, over a year later…

Youle helped me build it, and UPBGE has a GIT for requests and bugs, and when people propose useful features…

They actually code them!

Imagine!

How silly is that?

40

@ideasman42, was also thinking about auto-closing, but guess it’s not the most complicated feature to code :wink: I agree that current tracker works quite well though, but I also see why people apparently think that it’s better to have trackers for such overlapping topics in one place. Anyway, was of course talking purely hypothetical, just to throw in a possible compromise.

@jedfrechette, I know but the point I’m trying to make is that there are systems which are better suited, just like ideasman mentioned too.