Baal

@nbohr1more @Dragofer @fllood Who can provide me with the files for "The White Rose Hotel"?

Thanks for the info on that. The way I have set it up now is: LMB -> Cycle Select (the previous selection is deselected) Shift-LMB -> Select (select the target and keep the previous selection, drag selection is also done this way) Alt-LMB -> Manipulate Ctrl-LMB -> Surface Selection This works pretty well (after you have unlearned the old bindings).

I would like to take a look at "The White Rose Hotel". Nobody is working on that at the moment, right?

I don't find drag selection to be very useful in practice. I never use it. So that would be a welcome change.

Selection is left clicking while manipulation is left dragging. It works perfectly this way in other editors (Trenchbroom for example). And Cycle selection is the mode I use almost all of the time, so it would make sense (for me at least) to use simple left click instead of having to use a modifier. If it's to much work to implement that, I understand; but it's a huge useability improvement in my opinion. Edit: Rebinding Manipulate works ok. But it's not ideal.

I just changed the mouse bindings to cycle select with left mouse button in the camera view. Much more comfortable than Shift-Alt-LMB, but now I can't use the rotate and move handles anymore. The 2d view has the same problem.

Is the 'userinterface' branch on github stable enough to work with it?

Well, not just because, but because it's fun. Are you telling me power outlets are fun? I have to try one then. I do think that in-game shops are a great idea. But you don't need them to justify nicking everything; you are a thief after all.

Looting, for me, is part of the core gameplay of Thief. It's very basic, but simply fun and rewarding. Find something shiny, stick it in your pockets with a satisfying bling. It doesn't really matter if you can't use it for anything and I would not complicate that simple principle. You wouldn't spare a monster in Quake just because you don't have to. You blast it away with your shotgun because it's fun.

On the subject of missing loot alerting ai: I think that could be interessting and fun, only if the player can actively prevent the guards from noticing it. You could, for example, turn off the lights, knock out guards that could see it or close and lock doors on their patrol routes etc. Otherwise you might have to just leave it alone, and I think that would be unfun and frustrating.

Thanks, I didn't know that. It does not work in the dockable layout, though. Nothing happens.

The possibilty of maximizing the 3d or 2d view temporarily on a keypress would be nice to have.

This is exactly how I did it, too. The whole thing is a func_static so I don't think small grid sizes are a problem.

Does dmap generate a better triangulation with this approach? I need to try that.

Ok, that's better. But I think you can't have the curves be parallel that way.

But that will change the curve. You don't want that.

It's not as complicated as it sounds. I find brushes are much easier to work with than patches.

The way I did that, was to simply build brushes to fill the gaps.

Well, it's obvious that nobody ever had the slightest clue why guards suddenly were alert (me included). I propably thought it would mix things up a little, but it just causes confusion. It has to be fixed, and the easiest way is to just remove the spawnarg. Otherwise I have to make it clear what exactly will be missed; and I don't know how to do that.

Hm, I don't know. Normally you can just nick everything. That would change the expected, established rules. Does anyone have a strong opinion about that?

Oops. I completely forgot about that. I thought the cup was the only one. That feature seems to cause only confusion and the spawnarg should propably be removed on all entities.

Yes, but that is so inconvenient that I would never put something in a parent. So, while the restriction doesn't have to be there, in practice all parents would be empty. The workflow would then always be to create an empty layer, and then put the other layers beneath it. That should be automated (with layer grouping feature).

After working with the layer hierarchy for a bit, I've come to the conclusion that only the leaf nodes should contain geometry. You can't easily select a parent node alone; you have to move its contents to a child if you want to do that. So instead of making one node the child of another node, they should be selected, then grouped. That would create a new, empty parent with the grouped layers as children. Unless there is some error in my thinking, this makes more sense to me.

No, it's the sequel to Sir Talbot's Collateral.

This is very much work in progress but it's starting to take shape.