FOr those of you who have experienced cracks in your arches or pipework, here's a look at what causes the problem and how to remedy it.

1. Can see here cracks in the arch on the left and right of it: http://www.phait-accompli.com/q/s4/pre/prob_arch1.jpg

Here is the editor shot, you can see small misalignments: http://www.phait-accompli.com/q/s4/pre/prob_arch1-ed.jpg

2. And here the arch is fixed http://www.phait-accompli.com/q/s4/pre/prob_arch2.jpg
after straightening up in editor: http://www.phait-accompli.com/q/s4/pre/prob_arch2-ed.jpg 

What is the difference (or the different use) between self.think and self.think1?

I can find a definition of 'think' as "function invoked when entity must act
", but nothing for 'think1'. 

.think, as you probably know, is a special field according to the engine. When the local time on the server exceeds an entity's .nextthink, the engine automatically executes the function stored in .think.

.think1 on the other hand, is not a special field, it's an added field. If you were writing a mod from scratch, you'd have to include .think, but not .think1. The only time a function stored in .think1 is called is when the QC calls it.

So, when does the QC call .think1? Well, only in one place to the best of my knowledge, and that's in subs.qc, in the functions SUB_CalcMove and SUB_CalcMoveDone. SUB_CalcMove essentially calculates how long you have to wait while moving at a given speed before you reach a point on the line, and then sets the entity to move at that speed, and think at that time.

When you have a object like a door or a platform that moves to a destination point using SUB_CalcMove, you might want it to have a think function be called when it arrives. However, the .think function is already used up, because when the object reachs its destination you must call SUB_CalcMoveDone. SUB_CalcMoveDone basically sets the origin of the moving part to exactly where it should be, there might be rounding errors otherwise. So you store the other function you'd like to call in .think1. When SUB_CalcMoveDone is called, you'll notice it checks for a .think1, and executes it if it's there. SUB_CalcMove sets this .think1 automatically.

Oh, and SUB_CalcAngleMove does the same thing. So two functions use it. Generally if the object isn't some kind of a door or bsp object, it won't use .think1 at all. Monsters, for instance, use their own navigation code rather than this function, so think1 is just empty. Which is a bit inefficient really, it seems like you could merge one of the th_stand fields with think1 and save an entity field. Oh well, the quake QC code wasn't really written with total efficiency in mind... 

I've been trying to make a spiraled hollow pipe using the 0:2 2:1 1:2 2:0 ratio (the standard czg/spog/fingers curve). Now, CZG's curve tutorial (http://czg.spawnpoint.org/curv_tut/curv_tut.htm) stops one step short of this. The last pic he shows is of a spiral platform, if I understand correctly, but I want a curved pipe. Help please? The idea is that I want to have a solid cylinder core with a ascending curved pipe wrapped around it, that the player can walk through.

While I'm on the topic of that tutorial, many of the steps CZG goes through in that tutorial, based on worldcraft, seem impossible given my current understanding of radiant, which I am now using. You can't stretch or skew multiple brushes such precise amounts... or can you?

So... prefabs, ratios, more tutorials, etc would be much appreciated. I've been banging my head against a wall for too long on this. 

I've been thinking of such things too. Go up a few posts and checkout my bit about cracks appearing in arches/pipes as you're bound to encounter those.

Maybe I'll see if I can do what you're describing, but it'll be tough. As for Radiant - would holding ALT while dragging a vertice do it? (guessing) 

Thanks for your help: little by little I am getting to grips with QC. 

AguirRe - I'm confused as to how to change the width of a spotlight. I did it once a long time ago, but after reading the readme for your lighting tool, and fooling around - all I can seem to do is move the spotlight's angle around - it doesn't get any thinner, which is what I want to do. 

Phait, If you make your curves correctly, you will NEVER have such cracks.

I'm gonna try and get a little more specific than my previous and post some editor screenshots, as I have no webspace to post .map examples of my attempts.

#1 = http://img180.imageshack.us/img180/7114/spiral19yy.png
OK. First we have a 12 point cylinder, with each edge raised 64 units higher than where it began. I used vertex editing with split faces, because while doing it with split faces off or edge dragging would save the the flat face of the platform, the edge points don't meet up anymore and become misaligned by tiny amounts. By reading CZG's tutorial again it appears as though triangulating the platform face one way or another is the only way to make the edge points meet up.

Actually, I found an exception to that while messing with doing the same thing with 24 point curves (0:2 1:4 2:4 3:3 4:2 4:1 2:0). Observe:

#2 = http://img138.imageshack.us/img138/2264/24pointloft1dv.png
As long as the curved brushes to be spiraled remain essentially skewed rectangles, you can drag their edges to loft them without any misalignments. But as soon as they are, umm... parallelograms, (like the middle and edge brushes, with differing angles on the short edges) you can't. Hence why in that pic they are flat, whereas the rest raise higher. I just thought that was interesting.

#3 = http://img260.imageshack.us/img260/4605/fuckedupskew8ik.png
OK, here's where the trouble starts. We have a nice large 12 point curved pipe. Nice, except for the selected section, which i have attempted to skew upwards by ctrl+leftclicking. Really fucking ugly. So... since those brushes are not "skewed rectangles" like I worked with in #2, I don't expect skewing or edge dragging to work at all. Shall I have to manually vertex edit each face upwards, causing every face to be split?

#4 = http://img22.imageshack.us/img22/4233/clampgrid88qk.png
Let's look closer at where the interior points of this pipe fit onto. Yikes, they only clamp on a grid of 8. That's not... that bad, but i'm not entirely confident the ratios I'm working with are right.

#5 = http://img230.imageshack.us/img230/8914/loftgrid88mx.png
Let's try and make that bottom brush loft upwards 128 units. Whoops, radiant won't let me push the last vertex up the last 8 units. I guess that means it would make an invalid brush. Fuck. So I'm nowhere farther than when I started this post, but maybe you all have some clue of what I'm going on about.

Is there some ratio that the whole pipe would have to be lofted by, that relates to the interior ratios of the pipe itself, that are required for this to work??

I hope those imageshack links work, otherwise this post will make very little sense. 

Phait, If you make your curves correctly, you will NEVER have such cracks.

That was my point of posting what I did.

BTW, you got Radiant working on OS X. Does it still require that X11 bullshit? 

with the idea because I have never made a coil before, and thess are the results: yeap, I use Radiant too (though I use 1.2.11 now because it is the latest version that supports my system downgrade).

http://img50.imageshack.us/img50/7129/joequake0146ua.jpg 

to making any prefab is to at least scale the x,y and z coordinates to double what you intend the object to be in game.

I'd be glad to write of a little tutorial but someone would need to point me to a no frills screen capture program, or if there is screen capture built in to Radiant, tell me how to use it.

BTW, once you have the coil built, hallowing it out for player walkthrough is tremendously easy. 

You set the width of the spotlight cone with the angle key (default 40 degrees). In my Light tool you can also specify an inner cone for a more gradient spotlight. Please check the readme for details. 

As you mentioned, the four-sided planes made on these brushes cannot be contorted so their edges meet up. Unfortunately, because of their shape, you can't split these brushes into triangles. Without using lots of brushes for each of the 12 sides on the pipe, and without hollowing, I don't know how to accomplish this.

So basically there wasn't much point to this post.

Are you still in Greensboro from August through May? 

 

Phait: some ingenious bastards created an installer for radiant 1.4 and 1.5 makes it almost idiotproof to install. No fink dependencies or compiling your own source bullshit, it basically does everything for you and makes an application package with all the radiant subfolders in it. take a look if you have a spare mac: http://www.redsaurus.net/gtkradiant/index.php Runs beautifully.

Headthump: If I recall the pipe i was taking screenshots of was 512 units across, which was the size I wanted it. Theoretically those curves can be made at almost any size and scaled almost to any size; it just worried me that such a large pipe had interior points on such a small grid. If I try to make architecture inside the pipe that matches up to the same proportions, it will probably end up aligning to a grid of 4, which is rather smaller than I'd prefer to go. But I suspect it'll be inevitable. As long as all the points match up, QBSP's cool with it, right?

That screenshot you got there is pretty close to what I was envisioning. How well does that snap to the grid? Will it scale well to about 8-16 times that size? and what do you mean by "hollowing it out is easy?" If you mean using the hollow command in the editor, that is a really awful idea, and a surefire way to make qbsp hate you.

RPG: Yeah I know it's hard, I'll keep trying. And actually yes I'll be coming back to gboro sometime this Aug, even though I graduated last may. You still in, umm, chapel hill was it? I'll try and catch you on irc sometime. 

Spare Mac? I have a Powerbook :P

Seems Radiant even with installer, still depends on X11 - which I had installed once - I have Radiant on CD for OS X... but ah, just confusing setup. Anyway - it also requires 10.3 and I'm on 10.2 yet. Oh well. 

correct, the scaling will work fine at larger sizes.

lol, hallowing is a horrible solution to this, unless you are dealing with large areas and a well contained surface (that doesn't split up adjacent brushes in the compilation).

Instead, I would start out with three seperate spiraling brush builds to represent the cieling, wall, roof than the way I started here.

My example above was an octogon, given it is all 45 degrees, the math stays easy. I'm working on a 12 poly sided spiral now, and to be honest I'm not encountering the problems RPG mentioned. It makes me worried I am missing something.

I avoid using the vertice tool, and instead I move the sides into position and cut. It is less messy that way. 

phait, doh you're right, but it's apple's X11. I suspect macradiant will always be X11 only unless somebody makes Aqua-native GTK2-wrappers or something, which will probably never happen. And you should upgrade to Tiger (10.4), stuff like Spotlight makes it really worth it.

Headthump, can you show me a .map file? mail me if you'd like: owen dot greenwood at gmail dot com 

I'll send you the first map. The second isn't ready for puplic consumption. 

Check your Guilford e-mail, and then update your profile if that's obsolete. :p 

but the mapping help thread is surely not the ideal place for this conversation. :P 

that map I sent you, Grahf. I wouldn't be able to sleep at night without straightening that stuff out. 

Thinking about recent questions posted here, I built an acending hollow pipe (that the player can walk in) spiraling around a cylinder. It worked out well, so I cloned it and mirrored it to make a hollow tube figure-eight. Again, no problems.

I am curious how any of that would be possible without using the vertex tool. Can curved tubes be sloped without vertex editing? 

I think vertex editing is the only way. In Radiant, edge dragging or skewing makes really messed up brushes that don't align right.

Do you have a picture or .map file of your creation? 

i use them; they work in some situations; vertex manip works in others; you just have to know when to use each one.