Make all the other monsters attack you from upside-down and the Vermis battle will be easy to implement. 

Did you put the UE3 camera in ToeTag? If so, WINDOWS CONVERSION NOW!

Oh, rewrite Quake's map format and compilers so we can have subtractive geometry too. I can barely stand working on my Quake map now... "what do you mean I have to create 4+ brushes just to make an indent in a wall!?" 

Scampie

No, sorry. Just saying I like that camera system. ToeTag is using the standard Mac 3D camera set up.

Doesn't Quark do subtractive brushes? Or maybe it's BSP. One of them does... 

but it almost always turns the brushes to shit - like most boolean operations. 

He's talking about a system like Unreal has. You place a brush that acts as a negative space in the map. When the BSP compiles, it cuts that brush into the world rather than adding it to the world. It's hella awesome and makes BSP work much easier/quicker than in Quake.

Moving a window, for example, means moving your subtractive brush over a little. You don't have to resize all of the surrounding brushes to surround the new hole you want instead. 

Rather than making it out of shapes which have sides that the engine doesnt even need, and we waste loads of time with compilers trying to remove them! And I suppose it eliminates leaks too? 

That does sound pretty good.

Straight away I'm thinking of making a map only out of subtractive brushes, just to be awkward. Could be useful with a subtractive mask system, but probably too much for poor ol bsp.

Wasn't there some game that did work like this? 

how hard would it be to make an editor like that? Ie. leafs and nodes.

Then you could possibly even directly edit BSP:s.

Or you could couple it to the engine.

Then you end up with cube style stuff.. 

In my experience, you don't want that. Treating a level editing tool like a modeling tool (which is what this amounts to) has never been shown to be a productivity boost.

It's better for level designers to work with some sort of abstraction (i.e. brushes) and have a compiler that does the right thing later. 

That's what I was thinking of. 

thinking of making a map only out of subtractive brushes

One of my first (unreleased) maps in 1997 was done like this. I thought i was a genius (no leaks!) until, around adding room 5, the qbsp times were already climbing past 5 minutes. 

heh - a P133 w. 16MbRAM and Windows 95? 

pentium 166, 32mb ram, win95. 

yes, quark has those. Called "diggers" or something. I'd use them carefully though. 

... or hole creation is explained here:

http://quark.planetquake.gamespy.com/infobase/maped.tutorial.advanced.html#creatingholes 

quark has some cool features! Not sure I want to switch now though... not after 10 years of Worldcraft.

Holy shit. 10 years. Argh. 

 

I've tried QuArK, but I could never get into learning how to use it and always went back to wc. The massive amount of features makes it a shameful retreat.

Maybe its the interface - I don't like most modern interfaces with their helpful "features" and round icons for square buttons (fuck XP start menu) but the way QuArK is laid out doesn't gel with how I think about mapping in Q1. Odd. Or lazy. 

No, that's fair. I always felt that I should be using Quark as well but I could never get into it. It never clicked for me. 

Diggers are very good when you're unsure how to build an area.
Tag and glue features are also very nice.
For those who think quark is complex download an older version. And you don't need the interface. Quark has a lot of shortcut keys too.
But it it doesn't suite for quake too good. And is a bit buggy.

Bsp is about the oposite of quark: it's cleaner,(look at the map source,less compiling errors, Hom effects)
but has less features.

Ogier is the most 3D editor I have seen so far.
And radiant seems to be similar to Ogier but not as efficient.

haven't tied Worldcraft yet. Doesn't run on my sys. 

Oh god, that digger is ugly. In their example, 12 tris are created more than a properly built and mitered wall of that example. And since it's done post compile, you can't even fix it.

MITER 4-LYFE 

The point of them is the trade off though. Easier level building/tweaking in exchange for messier brushes. And, really, if the level runs correctly does the brush configuration really matter? 

When you do coding, do you favour methods that work, or methods that work quickly? 

diggers=Compile it and see how it looks ingame.

When you want to have a door somewhere else: just move the negative, otherwise you have to modify more than one brush and doing this severeal times would be a lot of work.
When you are sure build everything around the negative brush. 

"When you want to have a door somewhere else: just move the negative, otherwise you have to modify more than one brush and doing this severeal times would be a lot of work.
When you are sure build everything around the negative brush."

That sounds sensible. If you're concerned about how the brushes are laid out, go with this strategy. Fast iteration times = win.