The scrags have never failed to trigger the door (bridge). Here's a screen shot of the area in question.
rick0001.jpg

So this morning I try like 10 or 15 times always getting the HK to kill at least one of the ogres and the scrags got activated every time, no failures.

I didn't know the HK could also target something and still have a path. That's actually how I wanted it to work in the first place.

Player kills HK, ogres get mad and come to attack. Player kills ogres, then here come the scrags. Player kills all the scrags and then the walkway extends.

I will try it that way first. I don't really know what to think since I can't seem to make it fail now. 

that happens to me sometimes-- i'm trying to track down a bug and i get so into it, i either fix it without noticing or the problem was caused by something else entirely. 

Sorry for posting this out of the blue after not posting anything for ages and expecting help :)

I've been thinking that it's about time I tried to finish up a couple of my old maps since I did a lot of work on them and they were not far from completion. Unfortunately, Worldcraft runs like ass on my Windows laptop for some reason and I'm mostly using Ubuntu now anyway so would prefer to stay in Linux if possible.

Since I converted mostly over to using Ubuntu I can't easily use an editor that I would like, and in fact I don't even really know which editors can be run in Linux. Today I wanted to use Photoshop without having to boot up my Windows PC so installed Wine, which led me to trying to run Worldcraft. So far I haven't been able to get it working properly (1.6a has direct3d issues, and I can't get Quake Adaptor installed because it checks the location of the WC install and won't let me continue) and was wondering if anyone here had any suggestions for Linux mapping. Should I just suck it up and try Radiant (does it work for Q1 in Linux?) or is there an easy way to get WC3.3 or Hammer working?

I always wanted to make my own editor to be honest, but I don't really have the time, skill or patience to do so :/ Would be really awesome to make one for WebGL like Warren attempted some time ago. Unfortunately, the intel graphics card on my Thinkpad isn't recent enough to support WebGL :( 

Quest of course! Seriously though, a linux build of GtkRadiant 1.5 or Netradiant would be your best option. I think Netradiant has pre-compiled builds for download so you should probably try that first. 

You need to set the installation path to:

program files\worldcraft

instead of:

program files (x86)\worldcraft

Then it should work. 

just a shot in the dark, but willem made toe tag for the mac and it's open source. maybe you can compile it for linux?
http://www.wantonhubris.com/toetag/ the ui is very different from both WC and GTKR though, so it can be hard to get used to. 

 

Radiant is the typical choice for mapping under Linux, and Netradiant is the most up-to-date and bugfixed version. Works like a charm. 

http://dev.alientrap.org/wiki/7

http://ingar.satgnu.net/gtkradiant/files/gamepacks/

Use the quake gamepack, not the darkplaces one. 

Until last year I had only used Worldcraft, but then I switched to Netradiant and have never looked back. I remember it took about a week to get used to it.
I doubt I could even remember how Worldcraft works now.

I have a map that's closing in on 6000 brushes and it still runs smooth with XP on an old Athlon 64 3400 with 1 Gb of RAM. 

I've been doing lots of programming lately (javascript+canvas) actually, so ended up not doing any mapping the other day, though I did manage to get Worldcraft 3.3 working properly in Windows XP, so maybe I will just use that.

The problem I was having with setting it up in Linux was the installer for QuakeAdaptor wouldn't allow me to install in any place other than program files/ and even though I was trying to install there through Wine, I think it somehow knew that the actual path was different. Maybe I can copy the files from Windows and get it working properly. 

Not that I have tried it, but I think you can run WC from a USB stick for example. 

I've put up a 1 hour video on my yt where I look at an unreleased hl2 single player map I worked on some time ago and talk about level design shit. Might be interesting to some of you guys perhaps http://www.youtube.com/watch?v=k7FZAuWfNNg 

Issue with Quoth Delayed spawn. With some larger spawns (dozens of mobs) it sometimes doesn't auto wake them. I even added a relay to retrigger them but it still doesn't wake them up... thoughts? 

I'm going to channel Raymond Chan here, are the hordes spawned by the death of other monsters? Like waves once a certain number die? 

Trigger multiple -> Lightning trap -> Trigger once -> relays 

It's a bit odd all the other fights seem to work fine. This particular spawn however is spawning both a zerg in a spawn room that then come through teleporters AND a bunch that spawn right next to the player. If they are not waking you don't notice because they aggro you anyway :p 

So the lightning trap is shooting a hurtable trigger_once during this sequence? Because I could see that setting the wrong activator and so not waking monsters up. One test of this would be to replace the hordes with just a regular monster whose back is turned and see if that trigger will wake them up. If this monster also fails to wake then read on to find out the problem...

The concept of the activator is a fairly opaque topic in quake mapping. When an targetname event gets called, one of the entities in the game is designated as the activator of that event. If the trigger was a touch trigger then the player who touched it is the activator. Similarly relays propagate the activator of the original event.

What about when the trigger is driven by damage? Well, the QC has to work a bit harder to get this right. If the damage was done by a player rocket, naive code might make the rocket the activator. Really what we want is for the owner of the rocket to be the activator, which is why T_Damage takes two entities: the inflictor and the attacker. In this case the rocket is the inflictor and the player is the attacker. The concept of the attacker is also needed for correct ai and kill message attribution.

So I'd guess that the problem is that the lightning trap is being identified as the attacker and therefore becomes the activator. Lightning traps don't use their activator as the attacker, self is the attacker (mostly because the lightning uses the thunderbolt attack function which doesn't distinguish between inflictor and attacker). This is typical for trap shooters though, spike shooters would behave in the same way. 

Forenote: Bear in mind that if any of the monsters from a horde spot the player, they will wake all the nearby monsters which will overwhelmingly reach the whole horde. Be careful of false positives in your testing.

There's actually a much more common source of activator bugs in maps, which I think have been discussed recently: monster kills (I assumed this was the issue at first). When a monster fires a trigger on death, their killer becomes the activator of the event. Usually the player does the killing but infighting can ruin it! So it's doubly worth looking at how to fix activator issues.

One suggestion would be just change this crazy QC stuff so that monsters who are activated always attack the player regardless of who caused the activation. I don't like this fix because it only works in single player. In co-op you might have multiple players, so you have to decide how to sensibly pick one for monsters to attack. Even if you think you've fixed this, what about if there are no players!

A simple fix on the mapping side of things is to make sure that events to wake up monsters are only bound to things that players can do, like touch triggers, buttons and item pick-ups. A lot of the time this is reasonable and worth bearing in mind. I don't think it's ideal everywhere though; if you imagine a climactic arena fight with multiple waves of spawning monsters, it might destroy the intensity if the player gets to press a button at their leisure to restart the combat.

So in these circumstances I would recommend a new pattern: Use the trigger event which may have the wrong activator to spawn a trigger_once over the combat area. Make this trigger_once target the monsters on touch. Since touch triggers only fire on players you can be sure the monsters will have a valid target. There is no chance in co-op that the monsters will go after the player who is AFK over the one in the arena. If no players are in the server, we just wait for a player to enter the area. 

i ran into this problem back in ne_tower.

you could try making a giant info_notnull trigger_once in the area the player will be in that is only turned on when the lightning trap hits the trigger once, and have THAT trigger target the relays and monsters. this way the player will be the activator. 

i should refresh pages before posting. -_- 

There is no chance in co-op that the monsters will go after the player who is AFK over the one in the arena.

Reads better as:

There is no chance in co-op that the monsters will go after the player who is AFK in spawn ahead of the one in the arena. 

For once my verbose answer is in before Necros' concise one! 

I love you guys!

(did I already post this?)