Temple of the Strange
- Map name: strange
- Editor used: J.A.C.K.
- Build time: August 15-November 18, 2017
- Monster count (easy/medium/hard): 37/47/55
- Secret count: 5
- Quaddicted download
- Direct download
Around early August 2017, shortly before the initial incarnation of the Valve Developer Union, I had been turned onto the existence of a modern-day level editor for Quake known as J.A.C.K. Being a remake of the old Worldcraft editor, I was right at home with it. I set it up with my copy of GLQuake and started dragging out brushes with impunity. It was rough, I had no idea what I was doing, my tools were held together with twine and positive thinking, but it was good. Here's some shots of my little test map.
Of course, this went nowhere. I remember setting up a few spikeshooters and an ogre before getting bored and scrapping it. I wanted to create something of a ruins map, with muted, tanned bricks, open sky, pillars, decay, all of it. If that effort didn't pan out either, then I was gonna make another. I figured my recent upgrade to the Quakespasm source port was a good time to start over. The day after, I began work on Temple of the Strange, and opened up my former Tumblr to document it.
This choice just isn't mine
I knew from the start that I wanted to give the player a choice of which way they could go. One way would be faster, but the other way had more goodies. The former became the stairwell with the spikeshooters, and the latter became the watery side-cavern area where the player has the opportunity to kill some Rotfish and gain a Thunderbolt while they were at it. The construction gave me all the shit in the world, but eventually, I got it built and got it looping back to the bottom of the staircase.
One primary goal for this level was to build something that wouldn't tax the original game. From my understanding of ancient documentation, the old software renderers would begin to blank out areas if more than 900 world polygons were rendered in one area. Long before that limit, the software renderer would begin to lag. Knowing what I know now, this isn't really an issue; patched software engines are common now (Mark V and qbism Super8 are both very very good), and most maps nowadays far exceed this limit anyway. I didn't think to look for modern software engines, though. The large rune room ended up giving me a lot of shit as a result.
My plan was to have the player approach the staircase, come to a locked door, duck out into a side hallway and go looking for the key. Originally, there were to be four different hallways, but that got scaled down to one in both directions in the final map. Down one way is a
func_door cage with a key under it, and down the other sits the button that lifts the cage. With the lighting hiding the part where the chain goes right through the fucking ceiling (golly gee, nothing but the finest), it didn't look half bad when I got it all working.
Expanding on that discussion on
r_speeds earlier, however, that large room paired with the detailed (by the map's standards) key cage ended up pushing that area's world polygons into the 700s or higher. I tried simplifying the detail down in both, but it didn't do a whole lot. I really wasn't sure what exactly the issue is. This is about the time I learned of the visblocker, which is a bit of world geometry used to block high-detail areas from seeing into one another. This often takes the form of a wall blocking the mouth of a hallway.
My first idea was to rotate the staircase to block the view between the hallways, but that wasn't where the problem lied. Instead, it was through the door where the cage and rune room have a clear line of sight (
func_door doesn't block visibility). I ended up refactoring the cage room to block visibility by way of a small tunnel, and that worked wonderfully. (As an aside, the rotated staircase remained.) By this point, the "ruins" vibe had given way to runic textures mixed with medieval textures, because literally what are texture themes.
The ending to the map was simply a mad dash up a level, over a gap overlooking the earlier part of the level, and into a teleporter. Using the same style of construction as the flooded cavern near the start (jagged brushes intersecting), I built up a small outside area over lava that the player would battle through. It really didn't affect gameplay much, but I think it made for some pretty cool screenshots when I was finishing it up.
One cool discovery I made as I was finishing up Temple of the Strange was the existence of ericw-tools, a heavily modified set of compilers for Quake and Hexen II. Up to this point, I had been using the compilers that came with J.A.C.K, which had no special features to speak of at all. I set aside a week to totally redo the lighting and play around with the features of these new compilers, and that became the "proper" version of Temple of the Strange.
(I packed in the original source and build without any of the fun stuff for completeness sake. In some ways, I actually think the original looks better, with its more dramatic shadows and lack of oversaturated "look ma I can do .lit files like them there AD mappers" lighting, but either one's fine, really.)
I'm not exactly satisfied with the fork at spawn. Of the two branching paths, only one is really worth taking, the water cavern. There's absolutely nothing down the staircase route, and even if there was, the player would have to awkwardly backtrack to get to it. I'm still not quite sure what the solution to this would be. Likely a better non-linear layout altogether.
Balance has always been an issue for the level. I have an earlier build dating back to September 2017, and even I have trouble with it on the higher difficulties. I made a pass at rebalancing the level, but I overcompensated and made it too easy instead.
One thing I'm not great with is secret placement. There's no real reason for the player to shoot into the jail cell just before the runic room, but that's the way you get inside. I figure most people just skip the fucking thing and keep going; not like I put anything good in there after I rebalanced the map.
I polished it up with co-op spawns and some deathmatch-only weapons and called it done; my first proper Quake map, in the can. I sent it to Quaddicted in November and it finally went up in March of the following year. It's super inconsistent and long-winded, but for a beginner map, it could've been a lot worse.