[survey results] Features people want the most + MTG thoughts

[Zughy]

Hey there, this is a summary of this other topic => viewtopic.php?f=3&t=24631
I created it to put people thoughts in an orderly fashion and I think it'd be a nice idea to do similar surveys once in a while - like 6 months.
The survey contains the opinions of 15 people: I really wanted to have a larger poll but, well, I spammed the link three times (in three different occasions) in the IRC; I didn't want to be loud.

Every asterisk represents a vote. Every question assigned one vote but the first, which assigned two votes.


Things people want to see implemented the most (sorted in descending order)

New main menu ******
Better mobs API ******
Dynamically modify nodes def while in game *****
Dynamic assets loading ***
Better water ***
Better MTG ***
Ambient sounds **
SSCSM **
Better lights **
Better entity spawning **
Per-player creative mode ** (coming in the next release, 5.3)
MTG removed as default **
Nodes management **
Input management **
Better GUI networking **
Mouse binds **
Bind one action to multiple keys *
Entities' inventories *
More generic type for texture packs *
Built-in text on node faces *
Better playeranim mod *
In-game mod management *
Better inventory *
Bare OpenGL, without Irrlicht *
Separate Irrlicht from the code that doesn't render anything *
Bugfix entities glitching in solid nodes *
No latency in single player *
Better sound design in MTG *
Better general graphical representation *

Should be more obvious that MT is an engine?
Yay: *********
Nay: ***
???: ***

What do you think of MTG? (1 = very bad, 4 = very good. BEWARE: I tried to convert in-depth comments in simple numbers, as I didn't ask for a vote. If you don't trust this oversimplification, go check the original post!)
1: *****
2: ******
3: ***
4: *



Hope this can help shed some light a bit :)
P.S.: it'd be great if some admin could close the survey topic

[Wuzzy]

New main menu ******
Better mobs API ******

I agree. I'm glad to see that those have risen to the top.

Main menu is being worked on, but there's no real result yet. Minetest is still sorta kinda in the planning phase. Rubenwardy has written some code to make the main menu easier to code (I believe), but it's far from complete. It seems like the progress on this has been stalled for a while now. :(

Sorcerykid has posted a PR on adding more functionality for entity movement and other entity functions, making stuff like “smooth turning around” possible, lag-free. This is very cool and exciting stuff and I think it goes into the right direction.

Better entity spawning **

That one is not really about “better” entity spawning, but more about actually making it functional in the first place, because currently, it's very buggy and a game of chance (entity may or may not spawn, and you can't predict or detect success). A very very serious bug with entity handling IMO, that, for some reason, has not a high priority.

Better SSCSM **

This is misleading, too. As far I know, we don't have SSCSM at all.

Dynamically modify nodes def while in game *****

No, please don't. This sounds like a potential bug factory to me. If nodes modify themselves at runtime, this might cause all sorts of glitches and overall confusion, and it might invite very messy coding. I'm not saying it will be impossible to deal with it somehow, but I predict that things will go South fast, if we allow this to happen. It might become even worse when you're a mod that depends on a node that another mod modifies at runtime. This idea just sounds wrong on many levels. Maybe it would be better to ask the proponents of this feature what they actually want to achieve (concrete gameplay features), and then look how to implement them differently.

Playeranim implemented by default *

What you mean? Player models are animatable. There's just no default player model. But this makes sense, given the nature of MT. We can't expect all games to use the same blocky 3D player model. I think this suggestion has zero chance.

Bare OpenGL, without Irrlicht *
Separate Irrlicht from the code that doesn't render anything *

I think those two are pretty controversial. However, one coredev proposed something similar, to actually include Irrlicht in Minetest itself, and use it to fix bugs that desperately need fixing. I don't know if this is a good or a bad idea. Irrlicht is not dead by the way, there's still active development. But only the stars know when the next release will be. :-(

Better MTG ***

This is an useless suggestion because it doesn't say what should be done with MTG / which features need to be added.

I agree with most of the other suggestions.

What do you think of MTG? (1 = very bad, 4 = very good)

Where do the star ratings come from? I didn't give MTG a star rating. :P

[Zughy]

What do you think of MTG? (1 = very bad, 4 = very good)

Where do the star ratings come from? I didn't give MTG a star rating. :P

I extrapolated it from what people wrote. It anyone thinks my summary is misleading, they're free to check the in-depth answers :D

I fixed the SSCSM part, thanks :)

[aristotle]

...

Playeranim implemented by default *

What you mean? Player models are animatable. There's just no default player model. But this makes sense, given the nature of MT. We can't expect all games to use the same blocky 3D player model. ...

Hmm? Are you saying that we might create a model in a program such as Blender, import it and use that instead of the current one? Players and humanoid NPCs both using the same different model?
If yes, is it something for which lua APis already exist??? :O

[Hume2]

Dynamically modify nodes def while in game *****

No, please don't. This sounds like a potential bug factory to me. If nodes modify themselves at runtime, this might cause all sorts of glitches and overall confusion, and it might invite very messy coding. I'm not saying it will be impossible to deal with it somehow, but I predict that things will go South fast, if we allow this to happen. It might become even worse when you're a mod that depends on a node that another mod modifies at runtime. This idea just sounds wrong on many levels. Maybe it would be better to ask the proponents of this feature what they actually want to achieve (concrete gameplay features), and then look how to implement them differently.

I would do enough if the node appearance didn't depend only on the node ID but it could depend on the metadata also. So I could have two nodes with the same ID and both could look differently. It could be used for example for stairs, there will be no longer an extra node for each type of stair, there will be only one universal stair node which could appear in many different ways. If this was done only for nodes which appear rarely, it wouldn't lag much.

[paramat]

What do you think of MTG? (1 = very bad, 4 = very good)
1: *****
2: ******
3: ***
4: *

I think it is misleading to present these results in this way when people did not actually vote for these categories, faced with these categories they may have voted differently to what is presented.
Please at least add text to the first post that explains that these categories and stars are your interpretation of what people wrote. Or better, remove these and write your interpretation of the responses.

MTG certainly deserves criticism. However, please consider that if you had worked extremely hard on something for 5 years, you would not want any highly critical feedback of it to be misleadingly presented according to someone's interpretation. However, i understand you had no bad intentions.

[Andrey01]

New main menu ******

+100. I think it`s extremely needed due to two reasons:
1. The menu interface of MT engine and its games (including MTG itself) should be clearly separated from each other. Under MT engine menu I consider starting independent one when a player just booted up it. Via such menu already happens a switching between different games.

2. The menu design is too technical and pretty primitive in respect of the appearance and the GUI elements functionality. I`d suggest to add more visual feedbacks (highlighing table cells, tab headers in a bright color, accordingly selection in other).

Better mobs API ******

Also +100. The current AI of all mobs mods is pretty low-level. They can`t normally bypass obstacles towards to their way, go upstairs if it`s required, jump over blocks if there is an abyss between them.

Dynamically modify nodes def while in game

I agree at 50/50. On one hand, making nodes dynamic enables to change any properties right at runtime without a need to register additional nodes. On other hand, if certain node was changed, then ALSO remain ones that are put in the world should be changed. I think this feature is not realizable.

Better water ***

I agree for improvement and complication of the water physics, increasing the transparency and accelerating the water movement.

[Zughy]

What do you think of MTG? (1 = very bad, 4 = very good)
1: *****
2: ******
3: ***
4: *

I think it is misleading to present these results in this way when people did not actually vote for these categories, faced with these categories they may have voted differently to what is presented.
Please at least add text to the first post that explains that these categories and stars are your interpretation of what people wrote. Or better, remove these and write your interpretation of the responses.

Absolutely, my bad; done

[eyekay]

Playeranim implemented by default *

Uhh no I said that there should be a less hacky way of doing it, not that it should be there by default..
And I meant the playeranim mod (for moving the players head on z axis)

[Zughy]

Playeranim implemented by default *

Uhh no I said that there should be a less hacky way of doing it, not that it should be there by default..
And I meant the playeranim mod (for moving the players head on z axis)

Fixed