Escher mode 

Challenge accepted. 

I'd like to see the tris and how well it lights! 

I didn't try compiling it into a map. It's just a bunch of cubes with offsetting, rotation, shearing, vertex editing and a bit of triangulation.

Anyway, it was a bit disappointing to realize that the alt+rightclick tool for texture alignment is unable to make the texture alignment actually follow that path. Due to this, I couldn't use a scrolling texture to get the same effect of the animation linked by OTP.

The Quake BSP format uses a 2D projection on a 3D plane for texturing. This 3D plane uses a 3D projection to be rendered to the screen. I suspect that a 3D projection on a 3D plane would be required to properly align textures on structures like this. 

With valve220 and by triangulating all your faces it is doable, I did something similar with obj2map, works fine. 

Killpixel: Here's the mesh.

Bal: JACK's generated .map file has a "mapversion" "220" field in its worldspawn.
But what Jack is actually missing is a Contour Stretch UVs Projection tool. This could help a ton of maps.

And I guess that for such distorted shapes, this would require 3D texture mapping, where the texture can have a depth projected onto the surface plane itself rather than only on the screen. I'm talking about a surface with a texture tiled in a non-linear way like this, rather than like this. Hmm, I guess that "non-linear texture mapping" is a better way to describe what I'm thinking about.
Linearly mapped scaling, rotation and shearing can at best produce results like this. It can't make, for example, a rectangular tile fit a trapezoid surface - and many curved shapes are basically a set of trapezoid segments. 

No 3D modeling app does that though, you're only solution is to have more polies basically to hide the how the triangulation breaking up your UVs.

Still with valve220 + triangulation, you have the equivalent of UVs and can do in quake anything you could do in a modeling app with same ammount of tris. 

Absolute madman. 

Lunaran, Johnny Law mentioned that you wanted to make the Steam group public. I can do that, if nobody has any objections.

Any of you assholes live in SoCal? 

I live in Los Angeles dude. And luckily I am an asshole. 

Haha, perfect!

I'm moving to Santa Monica soon, is why I ask. I'll send a shout out after I'm settled to see if anyone wants to meet up for some kind of asshole convention. 

Come and hang out at nu-tf: https://discord.gg/9YkdqCu 

Santa Monica? Nice. I'm in Burbank north of the Hollywood sign.

I hope you like traffic. It's getting bad in SoCal. But you'll have a straight shot to the beach on the latest metro line.

Yeah let's meet when you get settled. 

What takes you to the US? 

If you would be so kind:

Level playtime
Level Structure
Protagonist 

You've picked some issues there that are 99% irrelevant to FPS game quality (I voted as such). 

Level playtime depends not only on the gameplay mechanics, but also on the savegame style and on the talent of the people making the levels.

For reference, Leptis Magna is the largest AD map which I've enjoyed enough to feel like giving it another full run someday. On some of the other maps, I feel a little too overwhelmed and it gets a little hard to mentally track down all the tasks (books, keys, secrets, etc, and their respective locations and puzzles). Without real-time saving and detailed status info (let's say, maps like Foggy Bogbottom would really benefit from a Metroid Prime style automap and a Doom 3 PDA style notes system), the more complex maps become more difficult.

On the other hand, some games out there feature big levels whose encounters/puzzles/etc are not complex enough to take advantage of their available space, and becomes boring. In this case, the way to improve them would be to make them shorter.

TL;DR: Great level design can benefit from expansions to the game's scope, while not-so- good level design can benefit from a reduction in their scope. 

The decision between episodes and a hub system should depend on the length of the levels.

Turok 2 uses a hub system where you can select each map individually, but each map takes a long time to be finished. Quake features an episodic hub system with multiple levels packed on each episode and where you can't select the maps individually, but the playtime of each map is way quicker. 

Whenever these sorts of surveys come up, small maps arranged in episodes are always overwhelmingly preferred compared to a single megamap. For the time and effort it takes to produce a modern style megamap, a mapper could produce a really decent episode of smaller maps (something like the scope of Terra, or DOPA), so why don't we see more of them? It is just that mappers find it more enjoyable making big maps rather than episodes? 

Modern mapping is about dick-waving. You want to make big, convoluted map with super detailed brushing. Gameplay must be at least 8h long. 1000 monsters is mandatory. If player's demo file is less than 20Gb, you're doing it wrong. 

@shambler - Is that so? I appreciate your vote ;)

@mankrip - I definitely agree.

@kinn - The current poll results corroborate the data I already have and it makes sense to me. Intense gameplay can become fatiguing rather quickly. It's nice to be able to get quickie rather committing a chunk of time to something you'll grow tired of half way through.

@khrathor - Hyper-detailed megamaps are awe-inspiring, incredible achievements IMO. The rub is that I have much more fun noclipping around slack-jawed than actually playing them. 

It's nice to be able to get quickie rather committing a chunk of time to something you'll grow tired of half way through.

That depends on the age of the audience. When I was a teenager with no job it was certainly easier to play long sessions, while in the last years most of my gaming has been on the bus. The megamaps in AD would certainly make it one of the top FPSs in the 90's, maybe the top one.

What's weird is that people who have lots of free time nowadays seems to prefer less intensive entertainment such as MMORPGs or binge watching Netflix.

I'd say to just define very well the kind of audience you're targeting, and go for it. There are all kinds of niches out there, you don't necessarily need to target everyone. 

Yeah man. I rate protagonist (as long as he's not as dull as fucking Jensen in DX), map linkage (as long as there is some and it's not unconnected gibberish like the latest Wofl), and map length (as long as it's not crazy short or days long) as pretty low factors compared to a game's, say, theme, flow, pacing, physics, movement, setting, aesthetics, exploration, gameplay-focus, architecture, bestiary, etc etc. 

Map length is a very important property. It hugely affects the gameplay experience IMO, and is also a big factor in deciding whether or not I can be bothered to replay. I agree the others aren't terribly important though. 

@mankrip - That and there are so many games now vying for people's time like never before. As for the target demographic, seems to be 30 year old males (20ish on the younger end, 40-45 on the older).

@shamber - Yeah, I get what you're saying and where you're coming from. When it comes time to make fairly concrete decisions (which it has) in regards to scheduling, staffing, allocating funds, etc, these are absolutely pertinent questions that have 100% relevance to the final quality of the game. These being some of the questions currently at hand does not mean that I haven't, and won't continue to, consider deeply the other aspects of a quality FPS that you mentioned.

So yeah, I'm coming at this from a logistical angle as opposed to a general theory angle.