[solved] Hurdles to implementing Merge modifier?

[Bobo_The_Imp]

However i am slightly worried that giving the loudest users their pet requests will set a rather unsustainable precedent for the next time someone wants something NOW

I want to second this. I think we’ve gathered that there is interest at the very least from the devs to implement this later during modelling nodes.

we might be better to hold off porting edit-mesh tools to the modifier stack and move to constructive modeling when we have a nodal modifier system that better supports it.

This quote shows that to me. So to me this means “just wait”

Ok then, How many people have to “scream the loudest” before it would make sense to consider implementing a modifier solution?

This is not “waiting”, it’s impatience in the face of an answer from the devs saying it will probably come later. Screaming is also not a becoming verb that I see being used a lot here.

I feel blender is actually pretty great at answering community feedback. Blender isn’t a huge institution, and yet it gets so much work done so quickly, at least as of late. That’s because the blender staff is extremely talented, and not to mention there is such passion here from the community and the devs. The important thing to remember here is blenders staff isn’t huge. They get a lot done, but they only have so much bandwidth to get things done, and we HAVE to respect that on our end.

So things like:

There is no need to take users into account

Is not true. Blender is very much treated like production software, especially in recent years. Blender fixes bugs pretty fast, adds things we ask for frequently and is constantly thinking ahead of how to better the software. If a change is going drastically change things for users, they have to seriously think about how to approach that issue. And they do! They just do so on what their bandwidth allows. Please think of 2.8, yes that took a LONG time to make, but it kept building up and building up, adding more to-do’s. In the end, though, it successfully came out and finished those to-do’s and checked out a lot of RCS requests, I might add.

I know it’s easy to get angry after many years asking for the same request, but, inevitably the devs are clearly aware of this:

it’s come up from time to time that we could make more edit-mesh operations into modifiers.

This means not only just a merge modifier, but all other kinds of constructive modelling modifiers. The devs have the man power to make this happen, but it hasn’t, and this is due to the very same issues and concerns that Campbell has mention here. They are worried it’s not a good fit right now. However, once more, with modelling nodes coming from Jacques Lucke in the future, the devs think these problems will become more manage-able. That is the answer for now.

I must say that right now, I understand the frustration, but I feel that people are worried about “the devs not listening”, where I feel that could not be farther from the truth. It seems like WE are the ones not listening. They have had this come up time and time again, and they still haven’t for well thought out reasons. Those are concerns to help protect us and blender, specifically not entering in the constructive modelling modifiers barrel of worms with a system that doesn’t necessarily support it. That is blender considering the users.

Let’s wait for modelling nodes then see what happens guys. Let’s trust the devs and give them our patience.

[chippwalters]

Screaming is also not a becoming verb that I see being used a lot here.

That was a quote from another poster and not attributable to me. That was the reason I put it in quotes-- as is customary when quoting another. I’m sorry you missed that.

[Bobo_The_Imp]

Oh, I’m really sorry about that! I did miss that. I don’t want to attribute that to you then!

[jesterKing]

Hi all, before y’all continue some points:

• merge modifier will happen, it was discussed this week by devs.
• big parts of the discussion clearly show many aspects of the feature request process need to be improved
  • the developers and Blender Foundation are committed to do that
  • work on that started recently
  • changes have been happening for a while, but it’ll take time to get everything done (there is a lot to do)

I don’t mean to end the discussion on process improvement, but I would like to ask everybody to take a breather. The discussion has not gone unnoted. Many of the raised issues are being actively worked on, but I hope you can appreciate this will still take time to mature.

Anyone is always free to point out issues in how anything is handled, but rather having threads with tons of replies maybe just contact me directly via PM on this forum. I have been specifically asked to help improve this kind of things.

Public discussion is naturally also fine, but threads that quickly explode are something that we should try to avoid - it makes it hard to keep track and on topic (But it does nicely highlight the need for attention on the issues ).

/Nathan

[Wazou]

merge modifier will happen, it was discussed this week by devs.

Hi,

To be clear, by that you mean:

1 - As a modifier as we are requesting?

Or

2 - For eveything nodes later in the furure?

[erickblender]

Dude, be patient! This is how Blender has been developed for many years, things happen but slowly. Since many devs have answered your request, it will be better to slow it a bit down. They know about the problem and they are planning to do something about about, we wait and see. Trying to ask them to do it the way you want it, IMHO it is a bit extreme.

[LudvikKoutny]

I object to i said it slightly worried me the idea of giving things to people that scream the loudest, since it’s just gonna lead to more screaming people all with their own needs, which is not a sustainable model.

People scream the loudest for the features that will help them the most in their day to day work. Those that save them the most time or boost their workflow efficiency the most. Correlation in the user feedback is by far the best source of information to decide the development direction. It will result in the software people actually want to use, and that’s what’s most important.

Yes, it does need some mediator person who will ensure that the features are not just slapped in, but implement in a way which won’t clutter the UI, won’t increase the learning curve and won’t be unintuitive to use, but in most cases of features requested by large masses of people at once, implementations are pretty much straightforward. Well, at least they would be, but with Blender, there’s always been this weird fetish of first trying to execute it some exotic “Blender way”, then waiting few years, then realizing that’s not really working out, and only then biting the bullet and actually doing it the common sense, industry standard way.

[Wazou]

It’s extreme to ask for clarification on something? Seriously?

[jesterKing]

Please, no discussion on how to discuss things, there is is no need for that.

To be clear, by that you mean:

1 - As a modifier as we are requesting?

Or

2 - For eveything nodes later in the furure?

My understanding is that 1) is currently being discussed and worked on, but I can’t give you any time estimates on that.

[erickblender]

Yes i think it is bossy, and this is what i believe we should avoid in Blender. Software development is a very complicated process. You should know that by now.

[Wazou]

Ok, thank you for the clarification

I hope I’m not too extreme or I’m not pressuring you, I should be ashamed to even have the nerve to talk to you or any dev in the world, I should kill myself and burry my own bones in the sand! ^^

[mathieu]

hi,
my patch is submittable. I teach in my course how to do it. I’m ok if any of my student propose it. But I’m pretty sure my version will be rejected as I have seen patches using similar techniques being rejected already.

[jesterKing]

Don’t worry about that.

This thread ballooned a bit, but it is fine. It does illustrate some pain points to look at, especially process wise. But that is what the Blender Foundation also wants.

I just hope that for a specific thread we don’t end up meandering too much

I have been contemplating to add a new category on this forum that would allow us to discuss such matters without clogging the other parts of the forum too much.

As said, and maybe it hasn’t been too clear, I’m here to help with these type of issues, and try to solve them. (I leave the tech/implementation side to the big brains of the developers).

[jesterKing]

Oh, and before you get too hyped up - the modifier will be for 2.82 rather than 2.81. We’re already trying to stabilize and fix bugs for 2.81.

[Wazou]

The thing is that people are against the petition and how I handled the request more than the proposal itself.
They are all for the merge modifier.

It’s like devs are sacred and we should stick to what exist to talk to you (RCS or nothing).

As a user, I would love to have the ability to make proposal and to work with the devs to improve blender.
For example, on modifiers, there are some little work that could improve them a lot.
Some lack of features (vgroup missing, revert vgroup missing, tools who don’t work, lack of some modifiers, etc)

Make something like some post there, talk about them, make proposals and have them added to blender by the BF.

Let the users make a part of the work to improve blender.

And it could be the same about other important parts of blender that need to be improved for years.
The basics parts of blender are the ones we need the most to be improved, updated etc.
New tools are great but, basics are the most important IMO.

So, if the BF is willing to work with us, I thing blender could gain a lot and if they are willing to listen to pro users, it’s even better.
Sadly we are not like Pablo who make tools, improve them and code, but we can make a part of what he is doing.

It’s up to the BF!

[Wazou]

2.82 is better than 2-4 years ^^

[Spatz]

Thank you jesterking!

[Ape]

jesterKing for… king! Jokes aside, this is great news.

[BD3D]

i don’t get why this specific feature get such a public outrage as we have many other big problems and important feature missing all over the place.

bad viewport performance,bad edit mode performance, bad instancing performance, catastrophic array modifier performance… catastrophic ctrl z performance ??? thoses are much much more important than a new modifier. It’s like we all assume it’s normal at this point… seriously ?

[LudvikKoutny]

The reason is that all the problems you are mentioning are not being done because they are all very complex problems that require lots of development time and complex planning to solve, where as weld/merge modifier is literally taking a few already written lines of code and turning them into modifier code. An afternoon for an experienced programmer at most. The ratio of development resources needed to benefit to the end user is extremely favorable here, yet the feature has been repeatedly rejected due to some of the Blender developers not being familiar with even basic non destructive modeling workflows.