Problems Compiling

(sacrosanct) #1

I’m trying to compile my map on the GtkRadiant, version 1.3.8-ET and I get the following error message:

************ ERROR ************
Error opening C:/Program Files/Wolfenstein - Enemy Territory/etmain/maps/test1.bsp: No such file or directory

I can’t figure out what the problem is, the directory does exist and there is indeed the aforementioned map there.

Can someone help me?

(MuffinMan) #2

this means that there was never created a .bsp so the compile failed somewhere, would be helpful if you posted your compiler log here

(sacrosanct) #3

The complete compiling message is as follows:

Saving map to C:/Program Files/Wolfenstein - Enemy Territory/etmain/maps/test1.map
Map_SaveFile: C:/Program Files/Wolfenstein - Enemy Territory/etmain/maps/test1.map
Saved in 0.02 second(s).
Setting up
Listening…
=== running BSP command ===
“C:/Program Files/GtkRadiant-ET-1.3/q3map2” -connect 127.0.0.1:39000 -game et -fs_basepath “C:/Program Files/Wolfenstein - Enemy Territory//” -vis -fast “C:/Program Files/Wolfenstein - Enemy Territory/etmain/maps/test1.map”
Connected.

1 threads
Q3Map - v1.0r © 1999 Id Software Inc.
Q3Map (ydnar) - v2.5.5-SD
GtkRadiant - v1.3.8-ET Jul 8 2003 13:02:47
After further consideration, I retract her earlier statement
VFS Init: C:/Program Files/Wolfenstein - Enemy Territory///etmain/

— Vis —
fastvis = true
Loading C:/Program Files/Wolfenstein - Enemy Territory/etmain/maps/test1.bsp
************ ERROR ************
Error opening C:/Program Files/Wolfenstein - Enemy Territory/etmain/maps/test1.bsp: No such file or directory

(chavo_one) #4

Are you running q3map2 through the radiant GUI?

I’ve run into a bug where radiant spawns the vis stage before the bsp stage finishes…hence the “can not find map.bsp” error.

2 solutions.

  1. use a batch script instead (good idea anyway, look in the ET reference guide for an example)
  2. Just run q3map2 from the GUI again. This time the bsp file will already exist and it won’t complain. (bad idea, because once the bsp exists and the vis stage starts prematurely again, it will re-vis your old bsp, leaving you feeling like you lost your last changes.)
(Hewster) #5

looks like he’s doing it through radiant,
It also looks like you have not done a BSP pahse, before you ran a VIS
phase

You must create your bsp before you can vis or light it.
eg for a full test compile, choose:
Q3map2: (test) BSP -meta, -vis, -light -fast -filter

Hewster

(chavo_one) #6

This is all true Hewster, but I have seen the bug I described before happen to myself with the new tools. I’m guessing that it’s not him forgetting to run the bsp phase, but the tools messing up the threads.

(Hewster) #7

I never use radiant to compile myself, so cannot say I’ve ever had
this issue. (Q3map2Toolz for me :D)

I’m simply assuming sacrosanct is new to mapping RtCW / ET, and not
selecting the correct compile option is easy to-do when your new to the
ways of mapping :wink:

Hewster

(sacrosanct) #8

I am running the compiler through the Radiant GUI and yes I am new to mapping. So I need to run a BSP phase before a Vis phase? What exactly does that mean? What does it mean to create a BSP? How can I know in the future which compile option to chose?

Help much appreciated.

(sacrosanct) #9

I’ve tried using the compile option, Q3map2: (test) BSP -meta, -vis, -light -fast -filter, Hewster suggested and now Radiant is crashing on me.

What on earth is going on?

(MuffinMan) #10

try bsp only without lights

(sacrosanct) #11

Hey, it works! If I try the BSP only compile option it seems to run! My problems might be solved.

Thanks guys. (I’ll try to find out what the difference between the compile options are, that probably was my problem)

(Hewster) #12

Glad you got the BSP to compile … I bet your well chuffed :slight_smile:

To fully compile a map, you first do a BSP, which basically converts your
geomary into the format the ET engine understands, then you do a VIS,
this computes the visibility of the map, which, in short, keeps the engine
from drawing areas not in view when playing, then you do a LIGHT
compile… this is the bit that will probably take the longest, the light
compile works out where all your shadows should be, and makes your
map look pretty :slight_smile:

Are you sure radiant crashed? compiling a map can take a VERY long time
Large maps, with lots of lights, and badly built on a slow machine with
not much ram could take days !! seriously

if you have less than 512Mb ram, you may run into problems, as if
q3map2 starts swapping to disk, cos its run out of ram, this will increase
you compile time 100X fold.

Hewster

btw, the latest q3map2 has a -lomem switch, which helps a lot on the light
compile.

(rgoer) #13

What would you say the cuttoff point for users using the -lomem switch would be? I’ve got 512 MB of RAM, but every now and then (usually when compiling a _lightmapscale’ed map) I get that “Windows is doing its damnedest to create more virtual memory, but please, if you want nothing to crash, don’t touch your computer” alert message.

(G0-Gerbil) #14

If I’m doing a full compile on a large map, I quite often restart my computer and then compile straight away - it’s the simplest method of freeing up as much ram as possible for the compile.

A minute’s delay while restarting will usually easily save that much and many times more time during the actual compilation stage.