I'm considering remaking the classic Slipgate Complex. Right now I only have very minor brush work done for the area where the player would start in the single player game. It's still just in my basic 'thinking about it' phase.. but if I do go through with it, it's a map I plan on taking just a bit more seriously than my Cistern remake.

I doubt it would be anything spectacular, mostly because I lack patience with finding textures and shaders, and lack any real talent when it comes to architecture design and such, but it's a thought that's been growing in my mind for the last couple days. I went through a bunch of the Quake maps, and came to E1M1.. only other one I'd remotely trump over it is another stab at redoing E3M7, which is a map I tried making before. But I'm not sure I want to do that.

Also know it's another Quake map, which would be my 3rd if it was finished.. but the only Doom map I'd remake has been done somewhat decently (Gantlet) and Q2 never interested me much. I'd go with a wolf-themed map, but I'd save that for if I ever got more creative and original with design and layout.. which I know I'm not.

Just been motivated with taking more stabs at map making recently for some odd reason.. so we'll see how it goes, I guess.

The more remakes the better.  Gen has always been a little lacking in the maps department owing to the lack of existing remakes.  Between you, Tabby, and Reboot, we've been getting some good stuff lately so I'm all for it. 

So, how`s the map going?.

If you`re stuck for textures take  look here  http://facelift.quakedev.com/retexture/ .

The trick, according to the regular mappers I've talked to, is to build the whole layout of the map in simple brushes. Then playtest it a bit, check the size and so on (so that you don't have to re-do actual detailwork later, if something needs to be changed in the layout or scale). And then you start working on the map in detail - preferrably in incremental steps of detail, so that you don't get stuck with one hyperdetailed section (and the rest of the map left to do).

Obviously, worrying about layout is not as big a deal for remakes (as I've expierenced first-hand, too). But starting basic and going into real detail later does seem to help. I also force myself to work towards finishing something (at least every other serious attempt), so that you mentally connect lame starts to succesful finishes. In the end it's just a matter of will(power): if you really want to make a map, just go for it. :]

What I meant is this: the difference between a remake and a new map, is that you don't have to worry about how well a layout is going to play when doing a remake -- you don't need to playtest paths, connection of rooms, shortcuts and relative scale issues (unless you're going for a remix). For a new map, you should (according to the regulars at Q3W) 'worry about the layout': make several early, ugly and basic alpha versions and playtest those.

Obviously the layout itself is important, and so is getting possible vis-issues out of the way early. :}

Woah. I don't know about that quirk. I've only used GTKRadiant for Q3, and every newer version was better, so far. Dunno what might be causing that, but it can't hurt to give a different editor a try.. (is qradiant still even in development?)

By detail I mean everything from lighting to precise placement of textures and brushes. Basically all you do to get from a blocky, fullbright hallway to a well lit hallway with convincing geometry (like brick- and woodwork + damaging/wear) and well placed textures &cetera. Just anything that goes beyond the basic layout, really.

Lighting doesn't take me as long as getting details in the brushwork right (like bricks falling out of a wall, or the woodwork in a ceiling), but tweaking it for the final version does -- because rendering takes ages.. :]

I don't really know a whole lot about the mapping process, but speaking from the perspective of developing models and code, I find a little planning can help save a lot of work.  For modeling, I do a lot of "scratch" work before I start really building a mesh.  I get some ideas on what I think something should look like, then I think about if it's feasible to mesh it or not.  Breaking a model down into sections helps as well.  This bit should be part of the main mesh, this should be a detail bit, this I can do with a shader overlay, etc.  If I were to apply this to mapping, I'd think you have three major types of areas - rooms, terrain, and corridors.  A room would be any enclosed space, and I know some of those can get complex, terrain would be any amorphous outdoor area, and corridors would be short connecting halls, stairs, or other spots that link one room or outdoor area to another.  Each of these would then be broken down further into sections.  You'd have major structural geometry like walls, floors, ledges, etc; supporting structural geometry like pillars, angled or recessed sections, stairs, ladders, etc; and detail geometry, like lamps or torches, bits of broken stuff, decorative trim, etc.  That's for the geometry at least.  For texturing, I'd probably get the major geometry done first, texture that, add the support geometry, texture that, then start looking at detail.  Lighting would probably be last because it's a lot easier to decide on texturing before  you start darkening areas.  For lighting I'd follow the same process - major lighting like ambience, followed by room-specific lighting, then extra detail stuff.  That's my thinking process anyway.

An example of some recent model work I've been doing and how this greatly helps.  I don't know how to rig models or use bones, so I do everything by vertex manipulation still, so keep that in mind in this description.  Say I want to make a motion on a torso model on a player.md3, let's say an axe swing.  First, I'll look at the main mesh parts.  I have a torso, two arms, a weapon tag, and a head tag.  The weapon will follow the weapon tag.  Since most of the severe motion is going to be in the arms, I'll keep the torso straight and start with the arm holding the weapon.  I'll start positioning that arm, and the weapon tag, so I can see where the axe will be.  Once I'm satisfied with the overall motion of that arm, I'll work on the other arm, and see how the two work together.  Once the arm motion is satisfactory, I'll work on the torso tag.  All the mesh components will turn as I rotate that tag since the tag is parented to the arms and weapon tag.  I'll turn the torso and lean it as you'd expect for the given motion.  Finally, I'll look at the head tag.  Where do I want the model to be looking when it's swinging the axe?  How would the head follow the torso's inertia?  By keeping the model aligned to X,Y,Z initially and working on the major actions first the detail stuff, like the head motion, I only have to do once.  If I worked it the other way around I'd never get finished.  The same goes for coding.  If I'm working on a major function group I try to get the logic framework functional first, then get into the details. 

For anything it's tempting to get caught up in the detail parts too early.  That's a trap that leads to frustration, because you get one area looking or working really well, then realize you have to do it all over again for something else, and then try to get it all working together and hope the new part works as good as the old.  I tend to work on either modeling or coding, but not both at the same time unless I have to write a support function.  Right now I'm working on only the thirdperson model stuff for Gen.  When I was working on particle code, that's all I was working on.  The same went for muzzle flashes, though I'm not done there by any means, I tended to work on one class at a time.  By layering your work it makes things more consistent.  I think for making maps the same would probably apply, regardless of how you prefer to work and what order you prefer working on things.

After all that, if I could give you only one bit of advice, it's to not get discouraged and give up.  The reason all the stuff in Gen works, which over half of it probably shouldn't, is not because I'm some super-talented coder or designer.  I'm not.  I'd consider my programming skills mediocre at best.  What I am is relentless and tenacious, and just too stubborn to let the seemingly impossible get in my way.  I don't look at a problem and say, "Oh this will never work," or I'd have given up a long time ago.  I look at it and ask myself, "how do I get this to work," or "how do I get around this," and then proceed from there.  If something doesn't work I back up and change my approach, and keep trying until I find something that does work and get the program to do what I want.  I can assure you that if I can program Generations to work as it does, then you can certainly make good maps.  Just... don't be too hard on yourself if things don't turn out right on the first attempt.  Your Bad Place remake is an excellent map and a blast to play.  I know you're capable of great stuff if you just keep at it.   

Ah yes, Zeroradiant.. that might be good -- I haven't tested that. If you want some answers as to how you should deal with editor quirks, it may be a good idea to ask over at http://www.quake3world.com/forum -- there are helpful editing fanatics there. I myself might help you out with GTKradiant, but not much else.. :]

Mink: I agree with Pho: not giving up is the most important step. There's always a 'pain phase' with everything that you need to learn, and the only way to make the most out of something (and actually start really enjoying it) is to grit your teeth and see that phase through.

If something feels tedious, either keep at it until it becomes second nature and you'll do it so quickly that it doesn't bother you anymore -- or, do some research and come up with ways to do the same thing more efficiently (like using clipping to make shapes, instead of moving vertices). Get to know the tools better, and you'll be able to focus on the fun and creative stuff.

A good way to prevent needless and frustrating fiddling with details, make sketches of how what the end result should more or less look like (or, if you don't feel like drawing, find pictures of hallways, architecture or geometry that match what you'd like to do). Analyze the shapes and make a simple plan of translating the sketch/picture to brushwork (worry about textures later). That way, you'll: a) only build details once, b) have a clear way of matching up your work (at every step) to the desired outcome, and c) get a handle on deciding whether and how your planning matches up to your ability to get the proper results.

I'm a plodder myself, which means that I'll keep toying with things until I get what I want. That suits me just fine, because I don't mind 'tedious' work, and it provides me with plenty of ways to experiment and improve. But if you're more likely to get annoyed or bored, the answer is probably to systematize: put more effort in creative work and mental exercizes, so you'll do less boring stuff. The trick is to turn a project into the kind of challenge that's enjoyable and rewarding to tackle.