Taken from The QuakeLab: Editing website http://www.planetquake.com/quakelab/ Used by kind permission of Steve Fukuda [The QuakeLab] [Compiling and Editing Problems and Solutions] [Editor] [BSP] [Light] [VIS] [Startup] [Gameplay] Modified August 4, 1997 |-----------------------------------------|-----------------------------------| |Problem |Solution | |-----------------------------------------|-----------------------------------| | |The final path_corner needs a | |func_train crashes at last stop on its |valid target path_corner. If you | |route |want the train to stop at the | | |last path_corner, give the last | | |one a wait of -1. | |-----------------------------------------|-----------------------------------| | |For networking purposes, Quake | | |builds a "precache" listing for | | |each map. It contains a few | | |vital things which are common to | | |all maps (these precaches can be | | |found in the quakec file | | |world.qc), and specific | | |precaches for objects found in | | |the quakemap (such as | | |models/textures/sounds for each | | |of the weapons and ammos, | | |healths, armors). This tells the | | |client to load these objects | | |into memory (if there's room), | | |and to assign these objects a | | |number, which the server will | | |refer to them by from now on. | |SV_StartSound: not precacheed | | | |I can think of two possible | |- warning message that shows up in the |solutions to the specific | |top left hand corner of the screen after |problem. You could add the sound | |using a trigger to fire a burst of |file to the listing of precached | |lightning in a third-party level, |files in the world.qc file, or | |message is displayed four times in a |you could add the respective | |row. |object which that sound would | | |normally be associated with. If | | |the custom map uses | | |lightning-gun sound for its | | |burst of lightning, you could | | |add the lightning gun to the map | | |which would then cause the sound | | |to be precached when the map is | | |loaded. If it uses that sound | | |effect from the lightning tower | | |in E1M7, I'm not sure how you | | |would go about forcing it to be | | |precached. Possibly copying | | |those two electric probes and | | |the switch that powers them to a | | |secluded area of the custom map | | |(somewhere inaccessible) would | | |solve that problem. - Nelson | | |"The Griffin" Hicks | |-----------------------------------------|-----------------------------------| | |The number of polymodels must be | | |kept to a minimum. I stumbled | | |across the console command | | |r_polymodelstats, which shows | | |the number of "polymodels" | | |visible at any given time. this | | |includes items such as weapons, | | |ammo, and health boxes, as well | | |as the weapon model that the | | |player carries in from of him. | | |The instant this count went over | | |20, the effect appeared. the | | |solution: reduce the number of | | |"polymodels" by removing extra | | |scenery. Remove some torches and | | |make the ones you leave in | | |brighter, for example. This | | |number was consistent for me in | |When viewing large areas and/or lots of |my map, but it may vary | |entities, the bottom portion of the |depending on how much or how | |screen becomes garbled but there are not |little architectural detail is | |grey areas. On the screenshots below, |in your map. I wouldn't be | |the average visible planes = 65. |surprised if it's always 20, | | |though. - Ernie Miller | | |-----------------------------------| | |I've been having lots of | | |experience with the problem | | |where the texturing in the | | |bottom portion of the screen | | |gets screwed up, and discovered | | |a new solution: It isn't always | | |related to R_POLYMODELS. At | | |higher screen resolutions it can | | |happen in quite small areas with | | |few surfaces visble, usually | | |when a health box is somewhere | | |in view. Solution: don't run at | | |higher resolutions. At the | | |default 320x240, the texturing | | |problems just don't occur. At | | |640x480, they crop up | | |constantly. - Sonya Roberts | |-----------------------------------------|-----------------------------------| | |The level may have a leak -- In | | |that case, you'll want to find | | |it and fix it. Another problem | | |might be the number of entities | | |you have in this "simple" level. | | |I'm not entirely sure, but the | |MAP layout: Two simple rooms connected |Quake engine probably has some | |by a short hallway. Grey areas replacing |trouble drawing a large amount | |textures in places. |of models in a closed (or wide | | |open) space. Check that you | |- Ran r_maxsurfs and r_maxedges and both |don't have a lot of | |are well below 1000 and 2000 (about |flashing/flickering lights, too. | |half) and the figures are the same for |As for the changing of screen | |the opposing room. Ran r_polymodelstats |size, this might make the grey | |and got figures around 30 to 50... But |areas go away becuase you're not | |after turning off r_drawentities (0), |forcing as big of an area to be | |(ie: turning the entities off,) the grey |drawn and redrawn each movement. | |areas disappear completely and the |The reason people decrease | |normal appearance is restored. (apart |screen size is for added | |from the consequent absence of any |performance anyway, right? Last | |monsters or torches). Also, if the |but not least, what are you | |screen size is decreased (not |running? A 486 or equally slow | |resolution---then the problem is still |processor tends to have harder | |apparent), the grey areas disappear. |time with all aspects of Quake | | |than, say, a P166. And if this | | |is your problem, then I might | | |recommend getting a whole new | | |system -- chances are if you | | |still have a 486, your entire | | |computer's a little out-dated... | | |- Noid=wm | |-----------------------------------------|-----------------------------------| | |This happens when there are too | | |many explosions, monsters, or. | |packet overflow |visible entities in a given area | | |within the player's field of | |- warning message while playing game |view. Reduce the number of | | |entities (usually monsters or | | |explosions) to solve this | | |problem. | |-----------------------------------------|-----------------------------------| | |Well, that happens when you | | |assign an entity's "movetype" | | |field to MOVETYPE_PUSH, but the | | |model that represents that | | |entity is not a brush model (bsp | | |model). MOVETYPE_PUSH entities | | |must be brush models... Sounds | | |like you have an alias model | | |with the wrong movetype on it. - | | |American McGee | |-----------------------------------------|-----------------------------------| | |Shooting a certain area of my | | |own map caused quake to crash. I | | |checked my map file again and | | |found an unbinded func_door | |MOVETYPE_PUSH with a non bsp model |entity right at the end of that | | |file. After removing this entity | |- sudden error while playtesting level |I played my level without any | | |errors. - Philip Poetter | | |-----------------------------------| | |The editor I used (Quest) claims | | |to have removed the scripts that | | |are attached to brushes when the | | |brush is removed. When I got | | |this error I looked at the .map | | |file and found several instances | | |of scripts with no brushes | | |attached to them at the entity | | |portion of the file. Removing | | |these unbound functions cleared | | |up the problem. Be sure you | | |delete what you think you're | | |deleting! - Scott Benes | |-----------------------------------------|-----------------------------------| | |This happens when you make a | | |connection to the outside world | | |by carving and then fill the | | |hole again with an entity-bound | | |brush; for instance, you cut a | | |hole in the ceiling and refill | |When running a made map in Quake, under |it with a button; you can avoid | |certain angles brushes (especially those |this problem by carving not so | |connected to an entity, for instance |deeply, so you have no | |doors or buttons) disappear. |connection to the outside world, | | |or you can put a solid brush | | |(not bound to an entity) over | | |the hole. - Matthias Hoys | | |-----------------------------------| | |Vis the level as if it were a | | |final Vis. Do not use -fast. - | | |Nicholas Lawson | |-----------------------------------------|-----------------------------------| |Making Deathmatch level. It passed QBSP, | | |LIGHT and VIS. Editor used: THRED. | | | | | |Created info_player_deathmatch entities | | |in several places, VIS'd the level and | | |copied the *.bsp to the server and all | | |clients. |Make sure your entities for | | |spawn points are placed a | |*The Problem:* Deathmatch players |sufficient distance away from | |sometimes re-spawn from their dead body |all objects IN ALL DIRECTIONS. | |instead of a random deathmatch spawn |When placed 40 units above the | |point. Sometimes not. |floor, they work perfectly. - | | |Andy Makely | |This gets hazardous to your health, when | | |you re-spawn right in front of the guy | | |who just blew your head off with a | | |rocket launcher. | | | | | |Any ideas? | | |-----------------------------------------|-----------------------------------| |The level in which this error appears |My guess and this is only a | |only appears in that one place, and it |guess is that an improper set of | |is a "flashing error"--i.e: |clipping hulls was generated. | |intermittant. It has a "peeling" effect: |(you are able to walk through an | |like the textures are peeling away from |area you are not supposed to, | |their positions in binary space. This is |thus the hom type effect). If | |similar to the "Hall of Mirrors" effect |for some reason the player is | |from DOOM, but not identical. (see |larger than a shambler it could | |screenshot.) |happen. What you might do is | | |insert a CLIP textured brush at | |Level in screenshot has been VISed at |the problem area, run qbsp | |-level 4, so it's not a leak or an |again, and see if that fixes it. | |improper VISing. Fullbright. - Phil Hall |- Troy Mann | |-----------------------------------------|-----------------------------------| |Compiles just fine, but gives this | | |message in Quake: | | | | | |SV_STARTSOUND: NOT PRECACHEED | | |SV_STARTSOUND: NOT PRECACHEED | | |SV_STARTSOUND: NOT PRECACHEED |I've had this problem with | |(repeats indefinitely) |func_trains before and the | | |problem is that you have the | |The map runs, but the func_train is |sound attribute set to something | |stuck in one position. The map is a |other than 0 or 1, 0 being no | |Tricks & Traps candidate called Mother |sound at all, and 1 being the | |of All Spike Shooters. It is supposed to |ratchet train sound (blech!). I | |simulate a huge spike shooter that fires |recommend 0. - Shawn Allen | |giant instant-gibbing spikes by | | |"recycling" func_trains. There are 4 | | |path_corners. The map was made in Thred. | | |- Evan Thompson | | |-----------------------------------------|-----------------------------------| | |When I shortened the pipe, my | | |VIS time (full) dropped by 15 | | |minutes! - Neal White III | | | | | |(ed. And in more detail...) | | | | | |You have a very narrow, very | |At one particular spot on my level, I |long tunnel to look down. There | |can look down a long tunnel. Quake |are a couple problems with this: | |displays a icon warning with the word | | |"RAM". The tunnel is just large enough | | |for a player to walk through and about |1. The tunnel is presumably | |1000 units long. There are perpendicular |textured, and Quake has a | |walls near the ends of the tunnel so the |significantly tough time drawing | |number of non-tunnel polygons visible is |distant, textured faces. I would | |very small. The tunnel itself was formed |definitely recommend shortening | |by subtracting one 12-sided cylinder |the tunnel, as the amount of | |from another. Hence, its cross-section |memory used to texture large, | |is a 12-sided, regular polygon and it |continual surfaces (with | |takes twelve polyhedrons to make the |increasingly reduced texture | |whole tunnel. |size as distance increases) is | | |rather horrendous. | |As a last resort, I reduced the number | | |of sides to six and the problem went |2. The portals generated by this | |away. All the usual metrics, however, |tunnel are very akward in that | |like looking at the tube with r_drawflat |they are of many differing | |didn't reveal why this particular |widths (in order to accomodate | |situation gives Quake such a major |the tunnel's sides), and they | |intenstinal blockage. Any clues? - Dave |are, where the tunnel meets the | |Wiley |rooms at either end, in excess | | |of 1000 units long. Try a full | | |VIS, and I guarantee that you | | |will still get sluggish | | |performance. SHORTEN THE | | |TUNNEL!!! IT IS THE ONLY WAY TO | | |SAVE YOUR LEVEL!!!! | | | | | |- Shawn Allen | |-----------------------------------------|-----------------------------------| |What in the world can be causing the ram | | |icon to appear and the R_SPEEDS xxsurfs | | |number to be scrolling at a high rate | | |when 'standing still' viewing a very | | |benign section(s) of a map. There is no | | |dynamic lighting (I removed all the | | |lights just to make sure it wasn't some | | |sort of buggy light entity, but that | | |made no difference) or long tunnels or | | |anything else even slightly weird in | | |those areas. - Neil Manke | | | | | |Here's the image. The section in the | | |picture only has 1 plain wall behind it |And here is the corresponding | |so R_DRAWORDER 1 reveals nothing but |solution... | |that. So Quake is not rendering a bunch | | |of hidden surfaces. One thing that | | |should be mentioned is those R_SPEEDS | | |numbers, in the image the 4th number | | |(172) "actual visible polys being drawn" | | |is higher than the 3rd number (131) | | |"total polys being drawn", are not | | |logical and must hint at the prob. How | | |can quake say it is drawing more visible | | |polys than the total polys it is saying | | |it is drawing? Also, loading the map | | |files into another editor makes no | | |difference. | | |-----------------------------------------|-----------------------------------| |I have a problem that I can't seem to | | |find a straight answer for. I created a | | |new DM map that includes 4 DM | | |(info_player_deathmatch) starts and 1 | | |single player start (info_player_start). | | |When I start up a server and load my | | |map, I appear at one of the deathmatch | | |start points. I have some lava and if I | | |take the lava plunge, I end up | | |respawning in the lava. This will happen |Apparently, if you have any | |3-5 times before respawning at another |(i.e. even _one_) | |DM point. If I continually repeat this, |info_player_deathmatch entities | |when I finally respawn at a DM point it |too close to a wall, it throws | |is always the same one (The 4 DM starts |them all off. I tested this and | |are in 4 different corners of this large |indeed it fixed the problem. | |room -- the first time I start I'm in | | |the NE corner, but the remaining starts |I haven't tested how far away | |are in the SW corner). |they have to be, but with my | | |previous map I had the dm points | |I ended up having other problems with |in a corner, right against two | |this map so I shelved it for a while. |walls. - from Al Harrington via | |The problem suddenly resurfaced with in |HAL9000 | |a CTF game we were playing. We kept | | |respawning where we died, over and over | | |again. The other team thought I modified | | |the server to make this happen. :-) | | | | | |Do you know of anything that can affect | | |this? I assume it's map-specific as we | | |didn't see this problem with other maps | | |(same server). - Al Harrington | | |-----------------------------------------|-----------------------------------| |I have done NO carving or subtracting. I | | |was designing some stairs similar to | | |e2m2's lift/stairs out of the GoldKey | | |water area. One of the stair pieces was | | |there and not there, it was visible, but | | |when I walked on it, I would go into it |OK, here's a solution for a | |and get stuck in it. I deleted the |problem I've had, and I'm sure | |offending brush, started the compile and |others have. It seems that after | |got this tidbit: |working on a map for a while, | | |some of the brushes became | |!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! |non-solid, allowing the player | |reached occupant at: ( 0, 624,-144) |to pass through. It was like a | |no filling performed |func_illusionary wall, but this | |leak file written to |was not the case. The brushes | |d:\quake\maps\tctf821.pts |were not snapped to the grid. | |!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! |After snapping all of them to | |3876 outleafs |the grid, the problem went away. | |--- MergeAll --- |Before snapping the brushes, | |4826 mergefaces |QBSP would report a leaf, and | |writing d:\quake\maps\tctf821.prt |the pointfile would pass right | |************ ERROR ************ |through some of the offending | |CanonicalVector: degenerate |brushes. - Don Riegel | |* QBSP did not build the BSP file - | | |can't continue. | | | | | |The *.pts file was blank. - Anthony | | |Distler | | |-----------------------------------------|-----------------------------------| | |All I can tell you for certain | | |is that it means exactly what it | | |says. You have some entity that | | |has a movetype of MOVETYPE_PUSH | | |that is not a BSP model (so it | | |is either an Alias model or a | | |sprite); this is not allowed. | |MOVETYPE_PUSH on a non-bsp model | | | |Unfortunately since the game | |That's the error that the server |exits, you can not see which | |sometimes (really ocasionally) crashes |entity it is. You need to look | |with on my map. I'm using the Worldcraft |for anywhere you are setting | |QBSP, LIGHT & VIS. It occured a lot more |movetype. Maybe in a touch or | |when the level had 8 buttons but i |use function? Sorry I can not be | |removed them & the doors they linked to |of more help. | |because | | |1. they slowed down DEATHMATCH play, |John Cash | |2. they crashed the map more often. - | | |Brendan Ragan |So basically, there really is no | | |solution other than to remove | | |the entities and/or any moving | | |brushes that are causing the | | |crash, which is what I had to | | |do. Quake simply does not allow | | |that in a game. - Reid Kimball | | |after e-mailing J. Cash at id | | |Software | |-----------------------------------------|-----------------------------------| |I have an extremely large 32 player map | | |that has what I call phantom brushes. | | |These are brushes that WorldCraft seems | | |to create whenever the hell it likes. I | | |thought it was a carving error but maybe | | |not. The brushes it creates go on | | |forever in the game world. Previously I | | |had one that went vertical and I had to |And here is the corresponding | |move an area of mine so it wouldnt get |solution... | |in the way. Now I have one that runs | | |horizontally thru a lot of rooms I have. | | |In WC I can select the brush but it is | | |only shown with the bounding box of red | | |lines, no white lines (white on black | | |background). The brush simply can NOT be | | |deleted from WC. - Reid Kimball | | |-----------------------------------------|-----------------------------------| |NO FREE EDICTS |Jack Perdue's No Free Edicts | | |mini-website | |-----------------------------------------|-----------------------------------| |My map is deathmatch only. | | | | | |The error on the server reads : | | | | | |Find Intermission: no spot | | |Object error in find intermission | | | | | |It occurs when the map times out. |And here is the corresponding | |It seems like it is looking for the next |solution... | |level and | | |gets screwed up on the change_level | | |routine. | | | | | |This was my first map. | | |Did I leave something out? | | |-----------------------------------------|-----------------------------------| |error: ifnot 4100(???) branch | | |23 client.qc : selectspawnpoint | | |client.qc : putclientinserver | | |(no function) | | |Runaway loop error |And here is the corresponding | |host_error: Program error |solution... | | | | |-when trying to play my level after | | |compiling in Qbsp | | |-----------------------------------------|-----------------------------------|