While this isn't so much a feature request as much as a discussion as to what we want out of the building editor.I think it's definitely vitally important that it gets done right and that it benefits both beginners and advance users, and store artists.

So here are my thoughts on what a perfect building editor could possibly look like and work.
With the new year fast approaching before getting stuck in new
content, I thought I would share some ideas and experiences with
regards to the building editor.

While the idea is certainly to keep it easy to use, which is fine for
the general users, if the intention is to build a 4 x 4 room, much
like classic did.Fair warning is that TGC will likely pick up a lot
of flag for 4 x 4 room assembly as it wouldn't allow for any thing
special or unique without having to do massive amounts of tweaking and
or adding own models.

While keeping that in mind, I have been thinking and pondering about
how a system can be implemented that both assists new users and offer
more advance editing without it being overly complicated.

What I have seen for axis game factory for example is that allow
placing a pre uv mapped cube as you scale the cube, the uvmap
subdivide the seamless map without scaling the map, meaning you can
have a simple 128 x 128 seamless texture map repeated infinitely
without have to place an additional cube.While it would certainly
increase the polygon count on each subdivision, it should hardly pose
and problem with regards to the number of draw calls as it would only
contain 1 cube.

While this works fine for a cube it becomes slightly more involved
when you have to create a room, easiest method I can think of to make
use of the uvmap subdivision, is to have a room blob with 4 walls
ceiling and floor and corners, should you wish to increase or decrease
the room size select a highlighted wall and pull it into the direction
you want to scale the room, select an edge of a wall with the use of
simple convex and concave features allows to only move that edge of
the wall to create an angled wall.

To keep it simple, allowing it to move in increments of 5 or 10 units,
means that any thing placed next to it or on top of it can easily be
snapped in place.

For example you can scale the room wall 450 units, it should create an
edge between unit 400 and 450, which will allow you to easily edit
edges and only that edge, adding the feature to edit and add more edges to
you can definitely create a more unique creation then the
standard 4 x 4 room.


There should definitely be the ability to highlight a specific wall or
roof, to change the texture of that specific face scale it, delete it
or CSG without affecting any thing else in the room.

There should also be the ability to add an entity, like windows, SCG
cutters, crates, light, pillars ect... that is duplicated as the room
is scaled, with an added widget menu that allows deleting CSG
functions, or which ever entity has been placed, which should then
revert that specific room section to default again leaving the rest of
the room blob untouched.

While scaling edges in increments will be much easier to deal with,
there should definitely be a freehand option as well, which would
allow users to create even more unique options in general for the
advance users, while it is definitely harder to implement round room
objects, there would be no need with freehand scaling as well as
increments with the added ability to add and create edges, and the
ability to push and pull edges.

Some thing that could make it easy is to have floor plan guide per
wall face with snapping ability, right click the mouse select guide,
angular, rounded, square ect select the size of the guide 100 units
200 units ect or any user selected number increments of 5 or 10
....and scale the face or edge exactly on the selected guide.

This would definitely mean some unique buildings.

Another useful feature and some thing that will be worthwhile in the
store is "floor plans" after a artist has created a building for
example with all the entities and windows ect, he can save the various
floors plans that may have numerous levels and variations as a simple
outline guide, upload it to the store with various room blobs and
accompanying entities. User can then place a floor plan choose any of
the different textured floor plans, place any room blob on the floor
guide, and gameguru automatically builds the floor plan from the
single room blob and entities included with the store upload.

After which the user can then edit the building as they see fit,
remove the ability to save any store bought "floor plans" and give the
users the ability to save only the building, should resolve lots of
issues with regards to users using some one's "floor plans" and
uploading it to the store.

Floor plans have positive features as you can literally create a whole
city for example or desert town, or space station ect from a single
floor plan, and simply up to the user to edit as they see fit, even
use their own room blobs and or entities that follow the naming scheme
of the default ones.

This is simply my general idea on the building editor, a block by
block method is really time consuming, and to me holds very little
rewards in the, a method similar or better even expanded on like above
would definitely put gameguru back on the map.

I should note, that the building editor also has to take into account
isometric, 3rd person, platform games and not solely on 1st
person.

Seamless texture repeat shouldn't increase poly count, so that isn't a problem. Ability to scale/repeat +or- a texture on surface by percentage in editor would work pretty well for that. I would add that it would be nice also to add an ambient colour palette for individual entity surface colour which makes it simple to alter colour and brightness of a single entity's texure, this makes it easy to match default and store bought items made by different artists. This ability will help users to create scenes with mismatched objects where even a greyscale palette will make the surface lighter or darker to match with the dominant objects and lighting in your design.
If you have a colour palette it gets even better.

Sounds like a good idea. I can see this being handy.
I have no idea what is now planned for the construction kit but I reckon they should ask for some user ideas on how it might work, so pretty appropriate thread really, we users need to take the initiative and speak up before things get started. We all have a good idea of what is planned going by the voting system so if we keep up with development we know what to ask for.

@rolfy.

I think also useful would be normals especially if you made a curve or pillar or round objects.I definitely want the ability of adding edges, and manipulating those edges or a group of edges, convex and concave editing should be a big considerations

I wouldn't reckon TGC would not want it to become full blown 'ease of use' modeling kit as they would concentrate on the game creation side of things , a lot of users may expect no less but I think they might at best put a basic system together. I seem to remember Lee saying something about blender support which would be ok for blender users. All the same if I want to model I have no issues with using a modeling program, I do get it that some users dont model but they would have as much difficulty learning to use anything that's complicated in GG. Hence my point that I doubt they would go too deep with editing functions in the construction kit. Booleans would be nice.... right ?

Boolean operations would be cool, but they can be limiting as well, my 3d editor is prone to crashes as a direct result of certain incompatible operations, the can of worms it opens, might not be worth it, if the editor is prone to crash.

The beauty of the system I am suggesting is, it allows easy editing, for normal users, as you simply pull a wall face to the desired size and all walls, floor and ceilings is duplicated along the axis you have pulled it, simple enough a monkey can do it.

More advance features like adding edges and having the ability is needed sorely, it is one of the reasons why gameguru is receiving so much bad press, as it doesn't have any ability to create primitives.

Having the ability to push, pull, angle, to you hearts desire, makes primitive creation a moot point, and I can promise you the gameguru uptake will be 10 fold, and much better reviews in general.

The second reason why a more involved system is needed, is polygon limitations, 30 000 polygon model is out of the question, splitting polygon in to sections isn't user friendly to say the least, for your own use it is great but not every one would be able to deal with 100's of split sections.


It is good and well they are touting it as a easy game creator, but they have time and time again burned them self to a crispy mess on steam, if you go by the reviews, advance or serious developers, need and want more advance things, simple fact.They have been shutting out those developers and they are paying for it dearly.

So definitely worth it to shout and scream for some attention before they get stuck into adding easy building editor.

Some thing that you may only be vaguely aware of is "entity composition" mode which as far as I am aware of will form part of the building editor, will allow you to create a 60 000 odd polygon model split it up in sections. Import the models section, adding windows, doors, entities ect, and create a stand alone model for the engine, and the engine will be able to deal with it by rendering it in sections while still only having one draw call.

it would mean having a desk draw for example with individual draws for one, and many other possible animation options.

So at least if we don't get building creator we hoping for "entity composition" would be the next best thing albeit it would be twice the amount of work, but regardless I rather have both easy to use tools, that still cater for advance users, when one isn't up to the task and calls for a more hands on approach.

Either way TGC needs to really sit down and carefully plan the building creator, so it doesn't end up as bad as the unfortunate construction kit which was just terrible....

I developed a Boolean method for DBPro that is fool proof. It is very fast and will not crash.
It handles the UV and normals automatically. I think it would work great in GG.

That would be awesome in GG, especially if they could get it to interact with terrain, i'd love to be able to take a bite out of a hill and make a cave or something like that.

@Stab in the Dark software

CSG cutting with boolean opperations causes the most problems and often adds additional polygons, intersect and joining meshes will be a moot point considering that the engine will in any case "build" the entity after editing it.

We are just left with CSG cutting, I think for all intends and purposes lee is going to keep it simple, in the shapes you will be able to CSG, to cut down on polygon costs and possible crashes as a result.

It should be noted CSG in general like classic didn't deal with the uvmaps after cutting, so that is also some thing to considering, simple cutters should be able to deal with the uvmaps, more complex cutters and applying a uv map, is some thing that needs to be manually done.

@Belidos
CSG isn't suited to terrain editing, for that you either need to create a terrain with the cave sections, or add a voxel terrain editing features, voxel editing of the terrain can become costly on polygons with a engine under a lot of strain to preform you definitely want some thing like voxel editing.

That said if you have a look at the death valley canyon level I created for gameguru it is possible to create cave sections, by using entities and the scale function of those entities just requires some imaginative designs and you can create any thing you want or need.That said voxel editing is definitely some thing the terrain editing needs, and will become standard in most commercial engines in the future.

The method I developed is optimized and reduces the polygon count. Did you watch the video?


This is not the CSG that was in FPSC Classic.
The method automatically handles the UV maps and normals no need to do it manually.

@Stab in the Dark software : Awesome video, thanks for sharing! Your CSG solution looks about a light year away from the last time I tinkered with the process, and really does the business when it comes to repairing the final mesh. I especially like the retained texture of the cutting mesh to form the 'inside' texture of the target mesh, a very neat trick when you consider the original artist would not have needed to define what the inside of their model is made out of I am currently swirling some ideas about for the new 'Easy Building Editor' system voted for by the community and the final plan may require some CSG magic (it worked well enough for FPSC to punch doors and windows into any mesh but I think the EBE (easy building editor) might require better CSG).

I like the video. Would be great if GG could do this sort of thing generally. I often place objects through others when making a busy looking scene. Would be good to have the option to reduce poly count this way, not to mention the benefits of being able to cut holes and such into walls or rocks in editor, rather than make new entities separately.

Also overlapping items in game engines is very shoddy way of working (it's not your fault though you're forced to do it here), overlapped objects cause all sorts of issues with rendering the scene, the least of which is flickering graphics and increased load on the engine.

I've seem Magicavoxel in action, and it is great.

3com