InsideQC Forums

Any reason to do them planar? I guess it saves a crapload of fillrate, but I assume it's pretty involved to get them robust enough? Looks great though!

Yes! I am not smart enough to implement shadow volumes :lol: These are _reasonably_ robust, not great, but they're cheap and they look good most ...

Dr. Shadowborg wrote: This is primarily targeted at FTEQW, not Quake2 proper.

Oh! So.. why Q2 BSP then? What's the advantage, besides colored lighting (which is possible without moving to an entirely new BSP version)?

Currently setting up Jackhammer config and converting id1 textures to .wal format, with the endgoal of converting episode 1 into Q2BSP format. Jackhammer config testmap looks coool so far with rad. :D

Learning a lot about Q2 mapping in the process, probably post a guide or something later once I ...

Yea, because your engine is open source (it is, right?), this would simply never work. The artist needs to get over his fears (and himself, from the sound of it). Despite his paranoia, "hackers" aren't waiting eagerly to leak or repurpose models from free Quake mods. Either he wants to help you out ...

Personally I'd just use C++, picking and choosing the bits of it I like, or that make sense to use, or that don't have a syntax that makes Brainfuck look sensible. There seems a huge element of "Not Invented Here" involved in designing your own OO framework on top of C.

No, it's not like that. I ...

Plain C with OO and a nice collection lib ? Woah.
Still reading the documentation, but so far I like what I see.
I'll download and play a bit with this later.
Impressive work. Thanks for sharing.

Right on! Looking forward to your feedback. I'm on a Mac, but I've been compiling it on Debian ...

Recently I had the idea of building an MVC framework atop SDL2 and OpenGL for (primarily) games. I'm fond of Apple's Foundation and UIKit frameworks, and thought modeling mine after those might be wise. To do that faithfully, retaining most of the idioms that make those frameworks sing, I needed OO ...

Uh, what difference does it make? Once your implementation hits the wire, HTTP 1.1 takes over. It's a spec. Makes zero difference what version of CURL (or any other client library) you use :P

Edit: I see -- you're hoping to basically copy/paste his implementation into your project. Well, in that ...

Hm, hopefully I'm wrong, but I think you'll run into overflows on connecting if you add a buttload more entities. Un-vis'ed or poorly-vis'ed maps with only a few hundred entities already have this problem. No?

Weird! Postimg.org must randomly move images from server to server. Updated the URL; maybe it'll stick for a while.

Yea, I would actually consider cross-compiling for Windows with Fedora. While I'm generally not a fan of Fedora, it seems to have the best cross-compiling support. This is what we use for Quake2World -- we have a Fedora box running Jenkins and cross-compiling for Windows. Here are our build scripts ...

I started using glib for collections management in Quake2World. It provides single and double linked lists, hash tables, and pointer arrays, all with a very complete API for allocating, extending, removing and freeing elements. Glib's memory allocator also uses slice management by default, which can ...

I've wrapped up my refactoring of Quake2's collision code. I've broken it into logical units, like Quake3, and I've ported over Quake3's "offsets" array that leverages planes' sign_bits to speed up testing against non-axial planes. I've also cleaned the code up and applied c99 standards. The full ...

As I've mentioned in other threads, I'm using LordHavoc's matrix library, which is heavily optimized and handles these kinds of transformations for points and planes really well. Here's the updated code which fully transforms the intersected plane for translated and rotated inline models ...

I noticed the other day that while Cm_TransformedBoxTrace rotates plane normals for inline BSP models, it does not adjust plane distance for them:


/*
* @brief Collision detection for inline BSP models. Rotates the specified end
* points into the model's space, and traces down the relevant ...