[MAP]x2: Blastedlands & Canoplasmaspam

[Rad Ished]

Divs patch deals with holes in models, such as .ase and .md3 which i assumed you'd used in this map. My map Ito took six hours or so to compile, and if that isn't a space map
Who told you a week is normal?
Is this true?
Ima not going to live that long!!!
Were you playing NEx the whole time on max settings? |>) <cylon
I go look at the .map now, thanaks four zaht

Ahh i see, a lot off overlapping things going on there, specifically I've been told that any overlapping involving caulk the best way to compile for a really really long time.
Here's the map i made that is very similar, it's no great shakes just an experiment, i made the .ase in blender (how i love/hate that app).
It has holes in it which I'm trying to get sorted before i give it a proper go.
http://www.mediafire.com/?2ymkojmzezn

[Revenant]

A week is ridiculous. You must have all of your terrain structural. Try making it all detail, and then placing structural caulk blocks underneath it to block vis.

[Rad Ished]

I just played blastedlands for a bit, ctf , I really like this map. Instant classic newb.

[TVR]

Since the terrain is a patch mesh, you should have entirely forwent the Visibility phase, as patches [1] do not block vis, [2] can add hundreds of unnecessary vis clusters if compiled solid with -patchmeta, [3] are buggy above 512 qu, [4] do not cast shadows

As a result, performance is horrid, compile times are extremely long because of the fact that visibility is virtually level-wide, lighting is not sharp, et cetera.

To alleviate some of the problems, specifically vis, caulk-hull the map [hiding a simplified version of the map as caulk brushes inside the patch mesh terrain, however do not overlap caulk brushes! As doing so will actually lengthen the compilation phase].

To fix the root cause for your next attempt, use a height map generation program ->
Easygen for smoothing -> GTKGensurf for auto-hinting and possibly grid decimation -> GTKRadiant for adding interiors and such.

Div0's patch deals with rotation of objects in GTKRadiant, [the patch] rotates based on a vertex, while GTKRadiant rotates based on origin.

[Rad Ished]

Divo has made three different patches, this sounds like it might be a different one.

[Clueless Newbie]

Were you playing NEx the whole time on max settings? |>) <cylon

Yeah. My mac has four cores, so VIS could compile on one core while the other three were available for other things. Only problem was that VIS kept hogging the HD, so everything involving the HD got bogged down. Next time I do something like this, I'll try moving everything - map and compiler to the other HD and see if that helps.

Ahh i see, a lot off overlapping things going on there, specifically I've been told that any overlapping involving caulk the best way to compile for a really really long time.

Overlapping? Where? I took care to make sure there wouldn't be any overlapping at all - other than of the meshes... hmm...

Here's the map i made that is very similar, it's no great shakes just an experiment, i made the .ase in blender (how i love/hate that app).
It has holes in it which I'm trying to get sorted before i give it a proper go.
http://www.mediafire.com/?2ymkojmzezn

I'll take a look at it later.

[Clueless Newbie]

A week is ridiculous. You must have all of your terrain structural. Try making it all detail, and then placing structural caulk blocks underneath it to block vis.

I thought about doing that, but then I decided I needed all the view blocking I could get. I'll give it a shot next time I make a map like this though.

[Clueless Newbie]

Since the terrain is a patch mesh, you should have entirely forwent the Visibility phase, as patches [1] do not block vis, [2] can add hundreds of unnecessary vis clusters if compiled solid with -patchmeta, [3] are buggy above 512 qu, [4] do not cast shadows

Are you talking about skipping the VIS phase?

As a result, performance is horrid, compile times are extremely long because of the fact that visibility is virtually level-wide, lighting is not sharp, et cetera.

Well, I'm kinda happy with the way the map turned out. Could do without the long compile time though.

To alleviate some of the problems, specifically vis, caulk-hull the map [hiding a simplified version of the map as caulk brushes inside the patch mesh terrain, however do not overlap caulk brushes! As doing so will actually lengthen the compilation phase].

I did a cheap-o version of the caulk-hull in Blasted. If you look beneath the meshes, you'll see blocks of caulk beneath all the boulders. They blocked the view and helped increase the fps for players in the valleys.

To fix the root cause for your next attempt, use a height map generation program ->
Easygen for smoothing -> GTKGensurf for auto-hinting and possibly grid decimation -> GTKRadiant for adding interiors and such.

You mean tri-soups and such?

Div0's patch deals with rotation of objects in GTKRadiant, [the patch] rotates based on a vertex, while GTKRadiant rotates based on origin.

(eyes glaze over) Huh?

[C.Brutail]

Blasted lands is pretty neat Good place for soma really mindless deathmatch action