I found a bug. CreateSingleSimpleEntityFromTable() works as expected when called from Update() or InterceptChat(), but the moment you throw it into a class, it fails to work. I already tried to explicitly access it from the global table with the :: operator with no luck. Example:


Maybe I'm just missing something here. I figured the Squirrel VM would check the root table (where I assume the function is located) by default. Is it even in the root table?

First post re-updated after latest updates. Thanks Valve for the hard work!

Woot!! Thanks Kerry! Thanks for the hard work!

Originally posted by Rectus:

I don't think there is any way to respawn dead survivors either, except through rescue closets.

Yep, gonna add to my list of needed functions, ha.

Originally posted by Rectus:

Bots wouldn't work very well, so you'd need to find a way to disable them.

There's a director option like cm_NoSurvivorBots.

It's possible. I've made Capture the Flag and Team Deathmatch before:

http://www.youtube.com/watch?v=6pm7HWv2k7c

http://www.youtube.com/watch?v=1fBPZxKkTCw

Just keep in mind that:
1) Without netprops, you cannot set custom glow colors, although there may be a workaround.
2) Without detouring several c++ functions, you cannot increase the player count (hence it's 1v1 or 2v2).

The rest should be possible in EMS.

Haha, guess so :P

----------------------------------------------------------------------------

I coded a VScript library that I call VSLib. You simply include 1 line in your nut file to use the library. I'll release it when I finish all the VSLib classes. Examples:

http://pastebin.com/XWmedB5B

When it's done, it'll be released under MIT license just to cover my butt :P

First post updated.

Thanks for the hard work Valve!

Netprops more so than keyvalues. Every keyvalue has a netprop associated with it, and there are way more netprops than key fields.

As for normalizing, what I did was divide by the length + epsilon just in case the length ended up being zero or null (for whatever reason).

A huge limitation is lacking the ability to get eye angles and position. At least for eye position, we can estimate it by getting the origin and adding on the vec maxes. Need a method to get eye angles so the trace ray feature is more helpful.

2 things I noticed thus far:

- Vector.Norm() appears to return a scalar floating pt value (which I believe is the old vector length) instead of returning a normalized vector. I think it is normalizing the vector I pass in itself and returning the length instead of normalizing a copy of the vector I pass in and returning the normalized copy.

- GetForwardVector appears to be using the base angles. There doesn't appear to be a way to get the client's or entity's "eye" position and "eye" angles nor a way to get the forward vector from the eye angles instead of the base angles. I can't even set up a good trace ray.

The second problem practically killed my port of Left 4 Amnesia to VScripts.

To try and compensate, I created my own function to calculate the forward vector from any given angles (given +z is up, the cross product of the up vector and right vector: x -> cos(yaw + pi / 2.0) and z -> sin(yaw + pi / 2.0) and y = 0), but what stinks is that I have no way to get the angles in the first place!

My suggestions:

1. Add a method to get the player's eye angles and eye position.
2. Add a method to get the forward vector from any given angle.
3. Maybe a method to trace from one entity infinitely in a given direction vector (an "infinite" raytrace).

Does anyone know if there are any updates coming to the beta? Haven't had one in a while...

Thanks