Dungeon Master Forum

Since Money has successfully conquered dsb-DM (despite finding a few bugs), a fully fixed DM is up, and can be considered FINAL.

I mirrored the fixes in CSB, which however has a lot of tricky mechanics that I couldn't test. So dsb-CSB is still a "beta" version.

Also, comes with this release:
- Chaos Hack (a random dungeon generator for DSB) - still no winning condition
- A small demo puzzle dungeon, called "Devices" (for which the working name was first... "Cube" and then... "Square" ...)
(the end is considered reached when you reach a staircase. Big up for the winners... trying to map it by hand will be a lot of fun also ! )

Download link: http://www.ojnk.net/dm/dsb/jorfiles/

Joramun wrote:a fully fixed DM is up, and can be considered FINAL.

Now you're asking for trouble. Look what happened to me when I called DSB 0.45 "stable."

I've edited your post to include the download link, though.

CSB. Error:

Location: 8, 14, 27
LUA ERROR: base/triggers.lua:894: Invalid instance 2999

EDIT: Fixed.

Is CSB up to date? I downloaded the CSB file from the link above, but the dungeon.lua file is dated 13th May 2010. I found a big gameplay bug that makes Neta unsolvable in the CSB+Magic Map posted elsewhere, and I assume that's the same version of CSB as you posted above... but want to make sure before posting a full report.

EDIT: Here's the bug anyway: The Vexing Vexirk can be chased away using fear attacks. If you chase him to the VI alcove at the junction of the ways, he's supposed to teleport away and leave behind some items, including the square key needed to progress. He teleports, but nothing is left behind, so you can't complete the game.

What was the reason for flipping the prison upside down and putting the entrance at the other end?

This CSB has been adapted from the RTC version. The prison is therefore in the EXACT SAME position as in RTC.

Sophia has nicely update the file everytime I send her a new version, so this one should be rather recent.
The wrong date may comes from the fact I re-zip the CSB files in the same csb.zip every time, so it probably doesn't updates the files dates. Otherwise, it's a big mistake on my part.

Gambit, I don't know any of the CSB mechanics you're talking about, so please tell me how the dungeon is expected to behave.

The CSB+magic map is not supported yet because It has been made by someone I don't know, from a CSB version I don't know. I'll look into it.

I'll keep everything up to date once all these mechanics are more clear to me. It's specially difficult because RTC mechanics are different from DSB ones.

Joramun wrote:The CSB+magic map is not supported yet because It has been made by someone I don't know, from a CSB version I don't know. I'll look into it.

Don't worry too much about this one. We're working on making the magic map into an importable archetype. Once this is done, you can just drop it into any dungeon, including your current CSB.

Sorry, forget that I mentioned the magic map, that's not the issue here.
I also tested your CSB map supplied above and the problem is the same.

Joramun wrote:Gambit, I don't know any of the CSB mechanics you're talking about, so please tell me how the dungeon is expected to behave.

I looked in DMute, RTCEd and ESB at the mechanism I described above. They seem to be correct, but I don't know ESB well enough yet to spot any mistakes. Basically, when the Vexirk walks on the tile by the vi alcove (#1 in the map below), he's teleported downstairs to a single hidden cell where a damager kills him, then teleports what he was carrying back to the original trigger. One of the ibjects is a key that the party need to progress.

In your CSB conversion, these objects are not teleported back, so the player can't progress.

I think the problem might be that trigger 2986 should activate teleporter 3261 (on the floor below)?

That seems to be correctly handled by the trigger 3262 on the same tile as the teleporter, which is monster activated. Is it something to do with the damager not working correctly?

I looked into this. The problem is in the timing and a slight difference between the way RTC and DSB handle events when there are multiple things on the floor.

The trigger is activated by a monster, but the damager fires first, so the monster is already dead... so there is nothing to trigger the trigger.

Yes, I noticed this in DM also. In DSB, the order of event is very important since multiple triggers on the same tile are handled one at a time.
There are easy ways to fix this though, the first one being, merging all triggers in one

A better way would be to copy the two items on that tile into the Vexirk. Then simply have an object teleporter permanently switched on with the damager. I've no idea why it's currently built the way it is, seems unnecessary to even have a monster activated trigger, since nothing else can get in that cell in the first place. (Unless it's a target for one of the random gigglers?)

I think it's designed that way so you get a greater reward for solving the puzzle correctly rather than just killing the Vexirk.

Anyway, I investigated more, and found that what I thought to be the problem wasn't really the problem, because DSB's trigger queuing mechanics took care of that. So, while triggering order does tend to matter more in DSB than in RTC, it's not quite as picky as I was thinking it was. (Yes, after all this time, I don't totally remember how my own engine works! )

The problem was simply that Vexirks float. It's an often-forgotten fact, but they do. So I had to set the trigger to "Ground & Air" and everything worked fine.

Yeah, surely the idea is to peacefully get some items in return for not killing it, instead of killing it for its items if you corner it?

And I remember wondering why a wall puzzle wasn't working when I had a vexirk in an illusionary wall. Took a while to try a different monster type, and realise the vexirk floated!

My recommendation doesn't prevent you from solving it peacefully. What I meant was: there doesn't appear to be any good reason why that single cell already has some items in it, and an inactive teleporter. You could just give all the items in the cell to the Vexirk in the first place and leave the teleporter switched on. It seems an odd design choice that's all.

Gambit37 wrote:My recommendation doesn't prevent you from solving it peacefully. What I meant was: there doesn't appear to be any good reason why that single cell already has some items in it, and an inactive teleporter. You could just give all the items in the cell to the Vexirk in the first place and leave the teleporter switched on.

That doesn't give you an added reward for solving it peacefully, though, because then you'd get all the items for killing the Vexirk as well. I think it's made the way it is so that you get an additional reward for doing it peacefully.

Ah right, yes, I missed that, cool. Those clever FTL guys

It's peaceful only if you don't know the vexirk actually gets killed...
What is the Vexirk International Rights United Society doing ?

They're cool wit it, it's very Obi-wan like. The game strikes him down to become more powerful than you can possibly imagine.

@Sophia: if I edit the CSB dungeon with that fix, will my save games still work? I'm rather getting into CSB again!

The dsb savegames are exact copies of the dungeon state + scripts at the time of the save.
Reloading a savegame will simply reload the old (thus wrong) scripts and dungeon layout. I'm sorry...

OK, no problem, I didn't get far yet anyway Those DSB worms are a lot meaner, too! I'm not looking forward to a room full of Munchers or Demons!

Joramun wrote:The dsb savegames are exact copies of the dungeon state

This part is correct. The scripts, on the other hand, are not saved in a savefile, so errors in the game script can often be fixed.
(This is also why you have to distribute your custom .lua files, or compiled versions thereof, even if you are distributing a compiled dungeon.dsb)