Very interesting research into the BSP process, Preach.
It's weird how the results we so different from a box map and a full size one.

Have you tried this test with maybe some small maps like DM maps? 

Afraid I didn't try any map in-between. I'd go and give it a shot now, although I expect that the result would be more like the latter than the former. Unfortunately I can't do that easily right now, because I've been revising the code for a future post, and don't have the original compiled at the moment... 

I was mainly interested in pinpointing where it started to become 'bad' and to try to pinpoint what it is about full maps that don't show up in box maps. 

I may have some insight into that from the new tests. It seems like there's a very fine balance to strike with random input - it seems to work best for trying out different ways of resolving "ties". I guess that in the simplest of test maps there really isn't any difference between just tie-breaking and randomising everything - but I think as soon as you add any real detail then the difference matters. 

I'm following this as well, thanks for the work.

Nothing solid to add though. 

is it possible to make rmq using some modern free engine?

i remember i've seen e1m1 on crytek eng 

 

http://tomeofpreach.wordpress.com/2013/06/09/quasi-randomised-algorithms-and-bsp/

So here's the way I was talking about, it gets you some improvement, even in the average case, over the deterministic compile run. I'm pretty sure that because the randomness is so low level, it's almost always a case of multiple tying brushes, and the randomisation changing how the tie is broken.

It's not mentioned in the post, but I did check and this version also performs well on the simple test map file. We achieve the same minimum of 160 (over 100 tests), and I'm beginning to suspect that can't be bettered. The graph shows an improvement in the average case, similar to the one that's in the article. So that's reassuring. 

is it possible to make rmq using some modern free engine?

that would end up being a free quake replacement, and zenimax would probably sue you for theft of IP. 

It's on a free engine... if you mean porting it to something like Unity or sauerbrauten then yes it'd be possible, but in that case it'd be better to make a standalone brand new FPS game using the same ideals as Quake.

Making a classic FPS game isn't too taxing in terms of asset and feature creation, as long as you make a feasible plan and still to it.

An idealistic FPS is almost impossible to make :)

And a AAA FPS would be a great big waste of time.

It also depends how much of a hobby or job the project is in terms of both your own time and if you can pay others for theirs. 

I have a thin and narrow platform (shaped as a square) that monster_ogres run around. It's over 256 units high from the ground, supported by 4 pillar/beams.

When monster_ogre sees the player, they run off the edge, dropping down to the floor below.

I was surprised to see this. I thought monsters don't do this.

I can't seem to find any information on this in google or quakewiki. 

The problem is the UI bounding box for the entity (in TrenchBroom) is misleading/inaccurate. I made my platform as wide(or narrow) as the entity's box, so the entity falls off the edge.

To prove this, I surrounded my platform with walls, and the entity was stuck.

Same issue with the monster_dog box. I put a monster_dog in the corner, and it's stuck. I had to move it 8units away from both walls (of the corner) for it to work.

Pretty annoying. 

the engine checks for collision in a >= way (as opposed to > ) so monsters have to be at least 1 unit away from walls to avoid getting stuck and have to be on platforms at least 2 units wider to avoid falling off.

note also that narrow pathways will be very difficult for monsters to move into because the checks for movement are done in discreet steps. 

I added your information to http://quakewiki.org/wiki/Entity_guide

Feel free to adjust, or find better place for it in quakewiki. 

"max leafs visible: 416 near (-832 -2824 576)"

Although this doesn't appear to be a warning or error message, I still see it as one. I wasn't getting it before, and now I am; and I have no idea what it means.

I've googled it a number of ways, and getting no good results.

What does it mean, and how do I troubleshoot it?

Does it mean too many polys for a certain area of the map from point to point? 

It's just a report

It tells you that of all of the places in the map, the one with the most leaves visible to it (as in when the player is stood there) has 416 visible to it. You can go WAAAYYY higher than 416. The other three numbers are the coordinates of this position. 

So I heard about this awesome content pack called Quoth that adds a lot of stuff (monsters, weapons etc) into the game for modders to use in their levels. However, I have no idea how to make any of this stuff show up in NetRadiant (the level editor I use) or appear in-game. How do I load these custom resources into the editor and use them? 

Alright, Ricky. Added to quakewiki.

hakkarin,
This page: http://spawnhost.wordpress.com/tutorials/quake-1-radiant-tutorial/ talks about Radiant, and mentions Quoth 

What is the maximum distance that can be jumped by strafe jumping and bunny-hopping? I need to know that distance for a map and i don't want to use fences or clip brushes and i want to make the chasm as smaller as possible.

Trying it by myself is no valid option as i can't get the hang of it.

As far as i know, with normal jumping the player can get across a chasm of about 180-200 units. 

I don't know that there is a limit, other than a product of the maximal speed (2000?) and the air time of a regular jump, really.

if there are no slopes, grenades, rockets, or monsters, then you can protect against shortcuts by making the opposing edge slightly higher than standard jump height, so that it can't be jumped regardless of speed. 

Not sure how strafe jumping and bunnyhopping fits in,
but regular mapping subscribes:

Max distance player can jump foreward => 225
Max distance player can jump upwards => 42.5
Max distance player can fall not injured=> 275
Min gap in floor before fall through => 35
Max step player walks before jumping => 17
 

With a rocket-boosted bunny hop sequence or something like that, a player can reach *insane* speeds which translates to insane jumps. I don't think there is any protection against that sort of thing really, apart from something like a midair teleporter or trigger_push, which is a bit cheesy.

(With the grappling hook in RMQ, just as anecdotal evidence, it was possible to reach horizontal speeds upwards of 1000 units per second - I've seen speedrunners go above 600 by other methods.)

Skill should be rewarded. But for 99% of players, 300 units or so should be uncrossable if there is no boost possibility nearby (including slopes and any source of damage) and they have no explosives.

I think. ^^

The slime pool in e1m1 is rather uncrossable (for mere mortals that is) without extending the bridge, so take that as a rule. 

The above goes for *Netquake* servers. Some popular engines are not quite Netquake. 

Those values are all wrong. 

what are actually the right values?