Game engines

[goldenboy]

There is nothing unknown about it. There is the BSP2 format used in RMQ/Schism maps and the slightly changed version in Darkplaces that was Lord Havoc's requirement for supporting the format in DP at all.

Personally I think q3bsp (and its variants) is superior in every way, not to mention q3map2 is a far superior tool. The only real advantage of BSP2 is that it requires no .map format change, and hence Worldcraft users don't have to learn a new editor, which was actually the main reason it was created, practically speaking.

It was a bad compromise. RMQ mappers wanted to make limit breaking maps, but at the same time keep Worldcraft. To be honest, that was stupid from the start. Refusal to learn a superior toolchain that could allow you to have the cake and eat it too, is one of the top 10 stupid things to do in game design and should immediately raise the warning flag. When someone says in 2012 "but it has to work with Worldcraft(!)" the right thing to do is to question their sanity, not to make a foul compromise.

Those are my 2 cents on that matter. If you make your own game, and you claim to possess sanity, use the right tools for the job. Worldcraft, QME, AdQEdit and TxQBSP are most certainly not the right tools unless you make a vanilla Quake map or a 100% "faithful" Quake mod. I can't stress enough how stupid it is to let yourself be limited by your own comfort zone. If your game can be better using something that requires learning, you should get out of your comfort zone and invest the time to learn it because it will pay off tenfold.

[Spike]

that said, BSP2 is much easier to implement in an engine than q3bsp (plus its derivatives), especially in a software rendering engine. An hour tops (assuming you already have raised limits - beware of the clipnode limit value if your engine still internally stores it as a ushort, you'll need to move those if statements out of your trace+pointcontents functions and in to the loader).

[r00k]

:X

[taniwha]

Yeah, I had bsp2 implemented in qf in just a couple of hours, and that included the engine itself (four renderers and two servers), qfbsp, qflight and qfvis, and then some testing and fixing misc bugs the support uncovered (yay shared bsp file support and already raised limits). As a comparison, IQM support took about a week for glsl renderer support, and then another week or so to get the other three renderers going.

q3bsp, on the other hand, would be a nightmare. I'd have to understand the format (if I don't, I can't debug it), write shared reading and writing support, debug that, create a super-set format in the engine and make q1bsp/q3bsp loaders convert to that, modify 4 renderers and a heap of physics code. Weeks to months of work (and no, just lifting code from another project is not an option: hasn't been for a very long time as QF is just too different) that I would much rather spend hacking qfcc. However, if someone were to provide a suitable patch...

[ajay]

While it's hard, though not impossible, to argue with the logic of what you said, gb, I think you're over-simplifying the circumstances and reasons that drive why people make mods. For most, I'm guessing, even those currently professionally employed in game making, it's a hobby; a creative outlet and/or distraction. Therefore - and I realise I'm dangerously meandering into making the same mistake you did, by imposing my drives, reasons and decisions on to others - people will make their mods in the ways that they find both most rewarding and that fit into their life; that give them creatively the enjoyment they seek within the time and energy they have to devote to it.
It's easy, I guess, to see the reasons that someone like myself sticks with worldcraft for instance, as a reluctance to move from my comfort zone, and yes there's a degree of truth in that, but that hardly makes up the whole picture. Sticking with what you know isn't always a negative; there's an enjoyment in using familiar things; a positive effort-reward in getting tools to do new and challenging things. I can clearly see the sense in using different tools to make my mod, indeed I should be using external textures too, making my own models - in fact thinking about what I want the mod to be I should probably be using UE3 or CryEngine, I should definitely form a team, then maybe it might be finished before I'm 50...
My laboured point it is that logic and sanity doesn't really come in to it; what I like about Quake modding is that even with my megre skills and with some old, some new and some evolved tools and engines, I can still aspire to make the ideas I have in my head - that I could make maybe make it easier, or more sophisticated or "more like my vision" (ha!) may be rationale, but isn't necessarily a priority for me or something I find motivational

[goldenboy]

I guess it simply depends what one's ambition is. I see that now.

In the case of RMQ, we spent 5 years on the mod with a team that was at times large, so it seemed to me like there was naturally an ambition that was greater than a hobbyist's. So it is my perspective that makes me say "BSP2 was a foul compromise considering what we were doing".

Making Scout's Journey is definitely more than a hobby for me, I see it more in line with indie games and things like Black Mesa, the Dark Mod, or Mount & Blade. Other projects, I guess, might have a different outlook.

You have some good points, but I think even for a hobbyist it pays off to force oneself to learn new stuff. It widens your horizon. Learning Crysis Sandbox taught me a lot about creating terrain and outdoor scene composition. Learning idtech4 taught me about materials, normal and specular maps, sound design, and what modelling props and textures can do. The knowledge from both of these now helps me make SJ a better game.

This question - should I stretch and learn better tech/software, and what is actually my ambition? - has a lot to do with choosing a game engine for one's project, actually.

One of the major fallacies of RMQ was probably that we never actually defined what our ambition was.

Good discussion, very helpful.

[Baker]

Black Mesa is technically superior and yet somehow is a very pale doppleganger of the original Half-Life.

Some of the things aggravating about Black Mesa are as simple as having to wait 45 seconds to 2 minutes when going from "one area to another while everything loads" versus Half-Life that did this in a couple of seconds. That kind of interruption in game flow is very annoying.

[goldenboy]

Luckily I'm not doing a remake.

[ajay]

What version of hmap2 is compatible? The one on the site is 2010, so I'm guessing not that. Got the latest DP...

[goldenboy]

http://icculus.org/twilight/darkplaces/ ... 0beta1.zip

Try this, from the files directory.

http://icculus.org/twilight/darkplaces/files/

[ajay]

[goldenboy]

Could you post the compiler output - the part where it fails?

[ajay]

It doesn't appear to be outputting then .bsp

[mh]

Space in the path?

[leileilol]

which makes less sense when quotes are being used.