is there a way where u can compile faster
fx make two computers compile a map ??
or something else
because it take bloody long time then u reach that time there ur
map is half finished

dont know about 2 comps, but heres a small tip:

if your trying out entitiys or something small then u can compile using fast Vis and No Rad, it will compile much faster.

If you not using Batch Compiler, Id suggest doing so.

It uses batch files to compile your map quickly and efficiently, while allowing you to both multi-task and abort the process easily.

All in a easy to use interface.

/points up at sticky at top of page :)

A good map = a fast compilation and good r_speeds

Before compile:
- snap everything on large grid. Engine does not appreciate grid 1 unit
- reduce amount of world brushes, and brush-based entities (a func_wall used to clip a model is very time consuming)
- reduce the amount of visible faces (merging angles at 45°)

For compile:
- Compile on a faster computer. Or add RAM.
- Compile via a network using a few computers. I know guys at Valve do compile like this (particulary for big HL2 maps). The more brandwith, the more speed for compile. 10/100Mb LAN is ok but a higher network will extremely reduce the compile time. Can't remember how to do it :(
- Using the Hammer original compile tools will make it 5x faster but the quality is different (no NULL texture, etc...)

Hope this helps.

i have the Batch Compiler
and u can use something called NETVIS
which i guess allowes u to compile vis over the net

thx for all the response

@ Mythic_Kruger
if its very time consuming to use func_wall as clip for models
wot should i use then ??? :confused:

EDIT: i have a: P4 2.4 MHz cpu, 2x256 ddr ram.

I don't get why you ask for netvis if you know the answer :rolleyes: All is explained in the help files.
http://www.planethalflife.com/eddi/

For clip, yes it makes it longer (the test map i made for the uboot is 6x longer to compile with the clip)... Removing the clips can make gain time on a compile of a map in progress...use -noclip for HLCSG.

but u can shoot through noclip ...
u cant shoot through a func_wall

For me, I do almost all of my test compiles without vis. Only when I have added a lot of new architecture do I compile with vis to see if it helped or harmed my r_speeds.

Basically compiling is divided into 3 parts:
1) CSG- generationg world geometry
2) VIS - calculation of visibility
3) RAD - calculations of lighting

During creation of layout/r_speeds compile without lighting (but with full vis). In this case you will see right r_speeds values and quick compiling.
Use full options of RAD only for final compiling.

RAD takes 50-75% of compile time, but you can reduce it only by setting lower options (not much you can do with map design).

For speeding time:
Ad. 1) Don't use world brushes penetrating each other, this may also give many other really bad troubles. Try to avoid this too with brushes in same entity
Ad 2) Use all r_speeds optimizations strategies every time (since beginning of creation), this will speed RAD calculations also. Use "Full" option, it won't increase compiling time too much, but help heavily with RAD calculations. Use NULL texture everywhere you can.
Ad 3) Scale up non visible but opaque brushes (like material clips). On every brush entity you can (like some clips and brushes in shadows) set "Light flags (zhlt)" to "Normal", not "Opaque".
For non final compiling use this options:
- No "Extra" lighting
- Use "Sparse" only if you have to (low ram available, too many patches).
- Use "No opaque entities". With many clip brushes it will heavily speed up time.

Having func_walls doesn't make any difference, it's the opacity that hurts, if it's even somewhat less than a cube it can exponentially add time to your compile. Keep stuff to right or 15 degree angles, if there are too many odd angles and complex areas the compile will take a long time. Make sure you've always got "snap to grid" set on as well.