News | Forum | People | FAQ | Links | Search | Register | Log in
Mapping Help
This is the place to ask about mapping problems, techniques, and bug fixing, and pretty much anything else you want to do in the level editor.

For questions about coding, check out the Coding Help thread: https://www.celephais.net/board/view_thread.php?id=60097
First | Previous | Next | Last
Actualy 
your map does not gave to be open anywhere to leak.
An entity by mistake placed into the void can cause a leak too.
At least thats true for q3bsp, so i asume that is a valid statement for q1 maps as well.

What shamblernaut was talking about are different network protocols.
15 is the original, 666 is an enhanced one, by Fitzquake methinks.
It's important because Quake works with a client-server protocol, so even offline in sp the gamecode communicates internally over this. 
 
if you're mapping for Hexen II then the protocol 15/666 question isn't relevant. 
Aside 
Leaks can be very tricky. Following the pts file will lead to the initial brush, but it always starts at one corner fading on a irragular vector into the void.
Very useful to set the sky brush dark and look with a 3d viewer along the pts line. 
 
About the protocols, got it, thank you for the precisions.

The pts line didn't show clearly the leaking point but went from inside to outside in an irregularly shaped part of the map (a cave). Certainly not a coincidence (thank you so much Wakey for having put me on the leak's track!)

I never understood where exactly was the leak but I found a workaround.

I encompassed the whole area into a giant square box. By chance the cave was a dead end and was connected to the rest of the map only through a rectangular tunnel. I just opened a square hole in the giant box wall to let the tunnel go through, et voilĂ !

Hop this trick could help someone in a similar situation in the future.

As a side note, strangely enough brush entities or water brushes partially in the void don't seem to cause a leak. I have the impression that the only "entities in the void" counting for leaks are model/point entities...

Thank you again to all of you for your help. 
 
the "put a big box around a leaky area" is an old trick, and it does work, but you'll have longer compile times and a larger BSP than if you actually plugged the leak. Depending on the cost/benefit analysis, might be worth finding the actual leak, or not. 
Planning Routes 
is something i want to learn more in-depth.
Just can't properly wrap my around yet.

How did you guys learn it and do it now?
Learning by playing doesn't seem to do the trick for me.
My guess is that most of you will already have a rough idea of the routes in a map in your head when starting to block out a rough layout - is that right?

To you take notes and sketches beforehand?
Like "I want this traps in that map, that many arenas..."?

Especialy interleaved routes with some backtracking and opening paths give me a headache, but my guess is that im thinking about them the wrong way. 
Yes. 
Sketch layouts first - including approx player routes, likely combats etc. 
My Nitwitz 
32768 brushes - 620 entities standard quake

I often asked myself what happens on the edge of the 33768 brushes. Having a map that big gets weary adding one brush or deleting one entity.

Still the knowhow behind it isn'nt clear to me.

Let's say I have map with that account. It is widly surrounded in the editor with all kind of edges. I assume the compiler uses all that outer brushedges to come to that 32768.

Then, would it be of any use to pull all outer edges to one line so the outside would be a cube with only 8 sides. It should make a difference, or am I just short sighted as these outer brushes get deleted after compiling? 
Brushes Are Not Faces(sides) 
a brush can have up to 255 faces in q1,
what matters is therefor the facecount.
outside faces seeing the void (if your map is sealed, no leaks...) aren't considered when creating the bsp. 
To Add To This 
brushes are made out of faces of course. 
Host_Error: Mod_LoadModel: 9 Not Found 
Hallo.

Got this error message on my map: Host_Error: Mod_LoadModel: 9 not found.
Seams I've wrote a wronge value somwhere.
Is there a way to find the entity/brush in trenchbroom, via notepad,.. or another way to fix this? 
Entities Come Up With Nr In Map File 
Trenchboom>View>Switch to Entity Inspector
or when it won't come up, I would search with Texpath in the *.map file all entities. A bit meak. 
@madfox 
Thanks! 
Still Cant Think Up 
a map layout.
Does anyone of you have some general tips when it comes to layouting your map? 
Steal It Off Something Completely Different. 
Like, I dunno, a HL Alyx cutscene.... 
Oh Cmon 
That would be way to obvious xD
Then rather some Boom Eternal cutscene.
But i'm not so keen on learning how to steal at all, i just don't want to come up with
some linear hallway-room-hallway-room kinda thing.
You get the idea. 
Sweet Inspiration 
Forget your intention of completing a finished map. Just imagine you are in a corner of a room. Start sculpting a corner without the intention to finish something off. Make a small ledge on the floor. Try to compare on the moment what your intuition is telling you.

I am working on three maps now, and stop immediately when I feel I want to finish them. That's why I am no speed mapper. It can take years and I don't care. 
Was Kind Of My Workflow Too 
But it hasnt yielded any release in years.
The last one was a WoP map in 2015.
Most of the time i experiment around with brushwork, but that doesnt help in upping my skills to create a map interesting to play. 
Methodology 
Make your method more structured.

When you just do random stuff finishing can get near to impossible. You don't know when to stop or where it's going so it just becomes an unending slog.

When you set yourself limitations and challenges, it becomes more satisfying to complete and you have to be creative to make something interesting.

Blockout
Build the entire layout in simple form, without detail or gameplay. The layout itself is up to you - I like to make things loop as much as possible and expand any dead ends into self-contained segments with unique challenges - button hunts, locked in a room with a shambler, whatever. Loops mean the player enters the same area from different directions along the critical path; try and avoid forcing backtracking unless you're going to change the structure of the level on the return or drastically change the gameplay (repopulate). You should be thinking analytically and planning what you want to do with these areas later on. If a particular area gets visited a lot, be sure to make it bigger.

Detail
Make yourself a handful of object to decorate the level with. Ideally these should be from an external bsp to let you make big changes with minimum effort further down the line. Does take a bit more work at the start, but it's worth it. Sprinkle the deco all over the place. Areas that deserve something special, make it. This is the more artistic part, so you'll be making your ugly blockout nice looking, modifying or expanding it where needs be.

Gameplay
Fill it with monsters. You should have planned ambushes and set ups in your head already during the blockout and deco phases - now you get to fill in the blanks. you should be playtesting constantly, on hard skill (nobody is better at your map than you). You shouldn't have to make any structural changes to the level now, but if you do or you have a cool idea, go for it.

Playtesting
Get a few people together and test it - demos are ideal. This is where the assumptions you've made on how the level with function with a rel player are tested. Listen to each complaint and understand what it's about.

Release
Wait for a month or so without working on it. Then replay the level and make any changes that you deem necessary. Once those are done and playtested, upload the zip.

Postpatch
There's probably some bugs. Be prepared to patch the release after it's done. This should be limited to fixes only, no new stuff. 
Kill Your Deadline 
I understand your intention, but from where the urge to deliver an outstanding map?
My last map is from 2014 and was one of my hardest cases. It can bring you on the edge of tranquility, as the map starts leaking or lightning gets abused.

From the moment I stopped wanting to make something good I made all these small maps with no pretention that fitted in together.

Looking at I'd maps it comes to me that they are not that hard to make. The challenge is to give it something unexpecTed. A moebius ring or a part with moving walls that gives several options for leadways.

If I'm on the point to produce something wild I just start playing AD. 
Everyone's Got Their Own Method And Driver 
My driver is the sooner I finish one project, the sooner I can start another.

Each project has new and interesting stuff to learn.

But, to each their own. 
Wakey: 
one easy idea is to start by building a "hub room" which has multiple entrances and exits, so the player has a familiar space they return to multiple times.

The hub room could contain locked doors (so you have to go find a key and come back) or multiple elevations (so the second time you visit, you are on a higher level you couldn't get to before.)

Some good examples of hub rooms are: the central room in e1m5, the starting room in e1m6, the room with 3 wind tunnels in e3m5.

Once you have a hub room, then you are mainly just building out the looping paths that branch off from it and loop back into it from a different angle. 
That's More Practical And Less Esoteric :) 
 
Yes I Just Watched Unabomber On Netflix 
 
Thanks Guys 
Already making progress.
The explanation of loops, how to deal with dead and ends and the explanation of the hub room helped a lot!

@madfox: It doesn't necessarily need to be outstanding map, but at least it should be worth beeing played.
I think i should at least try to make the best out of it, nobody wants his time wasted by a boring map.
Had my fair share of ugly releases already, like that messed up monstrosity Chili Quake XXL.
Gosh, i have to do a remake of that, to save at least some of my dignity xD

@metlslime: yeah, e1m5 and e1m6 are very good examples for learning purposes.
They use both a hub room and loops, but are not overly complicated or long.
So for examining and studying them both are ideal.
Cool about e1m6 is that the goldkey is visible from the beginning.
On the other hand it's one of those maps you can cheat through by just grenadejumping there.

@ijed: Thank you for your very detailed post!
Lot's of useful information in there.
Never thought about letting a map rest a month or so before the final releas, but it seems very wise.
You might start to see stuff that you overlooked while still working on it. 
First | Previous | Next | Last
You must be logged in to post in this thread.
Website copyright © 2002-2024 John Fitzgibbons. All posts are copyright their respective authors.