I'm totally drunk, and i can barely operate the keyboard or focus on the monitor. So it's a wonder i can still spell or type. I guess it's becuase i'm a bad-ass. Anyway. 

If you've been having problems with Q1 light banding effects in e.g. arch faces or infinite (delay 3) lights saturating entire areas, you might want to take a look at my latest Light 1.26 version available at http://user.tninet.se/~xir870k .

There are several new options for more flexible light distribution. Also, there are several new emulation modes for the ArghLite/IKLite/LightDLX tools. 

For the record, necros' map was totally crazy in the brush department, so the solution may be, "don't make totally crazy in the brush department maps" or something.

The map in question, (which, as I said, was a result of a decompile of one of my bsps,) was composed of 5000 1 unit thin brushes (the original map was 1000 brushes) , each representing one face in the map. So yeah, it was crazy in the brush department.
It leaked too... 

it was on a monster_shalrath, so i just moved all the shalraths around. stupid way to fix a problem but it worked.

ps: and it's moaning now, eh? :P 

Hay!

I've seen some maps with armors hanging in the air, how's that done? 

Hay!

I've seen some maps with armors hanging in the air, how's that done? 

1. func_wall is a support for that armor
2. set trigger_once with kill_target key set to that func_wall
3. put info_player_start into that trigger 

just set the 'suspended' option on the item. 

OK, I'm using Hammer and Qonverge... oops I mean WorldCraft 3.3 and Qonverge, and get this error " Error 35: error reading from file. " after it starts to process Hull 3. Both Hulls 1 and 2 apparently have leaks somehow, { "Warning 10: reached occupant at <x y z> no filling performed" ) even though I've made sure the entire map is leak proof... I think anyway. i've never gotten this error before, and haven't found anything on here about this error. Can anyone help me fix this? Would be most appreciated. Happy Thanksgiving! 

Thats a new one for me, and I've been using Qonverge and hammer uhh wc 3.3 for ages.

Just a thought, but isn't hull 3 for use with player crouching? (I might be wrong) If it is then there should be a parameter you can set so that the compiler doesn't compile the map to be "crouch compatible". Which is q1 matters not a lot :) 

with this parameter hull 3 wont try and be created so your error might dissapear. 

is it? Or do you know? I dunno maybe I should try to see if there are some leaks that I didn't get? 

didn't work. I tried it when I compiled, instead I just got the same error after hull 2, so that wasn't it. 

the problem runs a little deeper then :/

I dunno if I can help any more as I've never had this issue before :( I'm sure someone else around here can help, the guy who wrote the compiler has been known to peruse these forums on the odd occassion :) 

of DuBSP (which is a part of the Qonverge utils), but I might be able to help you if you send me the zipped map+wad. 

I was running the forumboards and came to someone, who had construckted with WorldCraft1.6

He got the error at the Quake1-console:
- too many efrags! -

Does anyone know, what this means? 

I believe this is something to do with complex brush models split over too many leafs in the BSP. Each bit is split into a separate "entity fragment" of which there is presumably a maximum number. It can happen if you use that trick to reduce the number of bmodels by combining ones from different parts of the map into a single func_wall or whatever.

On the same principle it can probably also happen if you have a very large bmodel or one in an area of very complex brushwork. 

While I'm here I'll ask a question. It's been suggested to me that I make a version of gmsp3 using hi-res textures and all the latest technology for use in engines such as DarkPlaces.

I'd like to do this (as long as it isn't too muxh work) not because I think it will necessarily be any better than the original but as a learning exercise for myself.

Not having kept up to date with this new fangled stuff (& never having glimpsed the doom3 leak) can someone suggest some beginners tutorials that would get me started.
Thanks 

Copied from Google usenet search :

I posted this question a few weeks ago and nobody even tried to
answer it! However, I did find out what was causing it in my maps. I
don't care for ID's torches, so I created my own using a thirteen sided
spike and the light_flame_small_yellow entity. I had tons of these
custom torches in my map and it ran great up to a point. I use
WorldCraft to create my maps and I found that if I moved the entity to
where it actually touched the brush that I would get efrags.

An efrag is caused when an entity is actually partially inside a
brush. The Quake engine is capable of handling a certain number of
efrags, but beyond that number it creates the "too many efrags!"
message. WorldCraft's entity sizes are not always correct, for example
the health entity is way too small in WC.

The entity for the
light_flame_small_yellow is too small in WC also, so by lining up the
brush and entity so that it looked like they just touch, I was actually
putting the entity partially inside the brush. This is what created the
efrags in my map. The solution is to put the light_flame_small_yellow
entity about 8 units above the brush. This doesn't look as good since
the flame doesn't originate at the botton of the entity, it floats
slightly above the bottom. I have noticed this problem with all of the
flame entities.

As far as not vissing the map, you will get the efrags message
whether or not you run a vis. The error is caused by the entity being
partially inside a brush. Curiously enough, light still creates lighting
effects even though the affected flames disappear from the map.

Apocalypse
Andy Knies
aknies@midohio.net

Maybe this helps? 

Is there a limit to the number of intermission cameras a map can have in a Q1SP map? I put four cams in this map, and those work fine. When I add a fifth, either the first four continue to work but the new one won't, or the new one works but one of the old four won't. (When I say "won't work" I mean it won't use that cam for the intermission.)

I guess I've already proven that more than four cameras won't work and I'm just looking for someone to confirm it. 

I can't believe I am giving a first rate mapper advice, but here it goes -- The information on editing gloss, luma, bump/normal maps here is applicable for any of the accelerated builds.
http://tenebrae.sourceforge.net/index.php?page=technical.txt
I don't think you'll need a tutorial on this aspect because if you understand the directory structure used for the accelerated Quakes (Quake\Tenebrae\Override -- for Tenebrae, Quake\Id\Textures for Darkplaces) you have most of the problem with texture placement solved.

If your concerns are with understanding the underlying technology, this guys page, http://www.paulsprojects.net/ can be of help. 

I played gmsp3 on the latest Telejano 7.2 a few weeks back, and visually it already appeared pretty damn sweet. Of the three Quake acelerated builds, Telejano mangles the original textures the least. 

Thought I might try mapping again. Have been using Fitzquake to play the latest levels.

Copied my old Bsp Editor directory from my old slow computer onto my new fast one and am therefore using all of the old utilities - qbsp, light and vis. All seems OK (the editor works as it used to) until it comes to launching Fitzquake after a compile, and then I just get a crash.

If I try to run the compiled level from the game, it doesn't crash but I get some video problems and the game will not continue.

I suppose the question really is - do I need different compile utilities for Fitzquake. 

Is it actually a crash, or does it quit with an error message?

Is the error message "couldn't find gfx.wad?"

If yes, check out RPG's question and my response starting with this post: http://www.celephais.net/board/view_thread.php?id=4&start=407

However, what about the video problems? I need more information to help with that. 

Looking at FindIntermission() in client.qc, it seems that you're right. That function appears to randomly pick from the first four info_intermission entities it can find.