Gamedev temptations

[mankrip]

What are yours, and how do you deal with them?

I'm tempted to buy the Substance Indie Pack. It should be great for creating textures for maps; probably the best solution available. Substance Painter also seems to be awesome for texturing models.
The problem is, I can't model for crap, specially realistic models. So, it would be pretty useless to have maps with completely realistic textures, if the characters are cartoony.
However, I still love the idea of trying to make such great-looking textures for maps.

I'm also very tempted to buy this 3D Action Adventure Kit for Torque 3D, even though I never used that engine before. Torque 3D seems really interesting and seems to offer a lot more freedom than Unity, which I've never used either. Unity seems to be quite bloated and its usage seems to be heavy-handed, but its support seems amazing, and it has thousands and thousands of content in its assets store.

[leileilol]

I don't believe in "make art" buttons.

[qbism]

I have a .plan, toolchain, qc codebase, and some testing ready in preparation for a Quake1 project. But every time I actually start to work on it I am thinking about how some feature or other would be easier/better/faster in Doom, Q2, or Q3 basis, and then I start up the search engine...

[goldenboy]

Mine are usually like, "oh, let's wander off and do a Doom 3 mod." "Oh, why not remake Hexen 2's Blackmarsh." "Oh hey, why not map for STALKER/Crysis/someotherprettygame." "Oh, why not mod Mount & Blade."

Basically, I get tempted into making some easy, quick pretty stuff. Instead of, you know, coding an RPG from scratch and slogging away on giant levels etc.

Also: "Oh hey, let's make something with Unity. It must be good right?" And then a couple days later, after realizing all the things I cannot do with that, I sit and softly shake my head.

It has become less and less though, since I can do pretty much everything I could do in those other games in my own game now. So I might as well put the time into that.

To be honest I also usually run into limitations in other games pretty quickly. I spent 10 hours one day trying to trigger a chaingunner to walk through a door in Doom 3, and said "screw that." I noticed that no Hexen 2 map compilers could compile my Hexen 2 map anymore, and people tried to blame me. I said "screw that." I noticed that Unity can't reliably make holes in terrain, and that the free version doesn't have occlusion culling or pretty water. I said "screw that, too."

I'm actually pretty comfy with my stuff atm. My recent temptation is to get SJ running in Darkplaces one day, but to be honest DP is just different enough from FTE that there is actual porting required in some cases, and that means extra effort.

I'm not terribly tempted by anything out there anymore. I do think "oh it would be nice to mod this or that" but it goes away after a while. Especially once you see that other games/engines/software are just cooking with water, too. We have particles in FTE. We have terrain, physics, CSQC, ragdolls, and all the other crap. I don't need some other software for that. We have GLSL in FTE so if I want depth of field, I damn well better sit down and code it. And one day I will. I hope.

I have Blender. I don't need any magical content generation software, because what I have lets me do what I want. Yes, it takes lots of effort. Anything worthwhile does. No magic bullets out there.

[gnounc]

So how long before you start modding SJ gb?

You can do everything you want to do in your own game, so when you see a mod you'd like to make for STALKER,
how long before you decide to just mod it into SJ instead?

Would be interesting. Especially since if it doesnt fit the main theme and system of SJ, it would make sense to
seperate it out into a proper mod and not implement it into the core mechanics.

[goldenboy]

So how long before you start modding SJ gb?

Well, SJ is already the game that I would want to play. It already includes exploration, outdoor areas, RPG elements, medieval stuff, high tech stuff... So I could imagine adding on to it in the future - more items, enemies, areas, quests, skills, perhaps unlockable characters. It has 3 main chars, not counting the coop guy, so the other 2 could be unlocked or be played in patches, sequels or mission packs.

With Mount & Blade, it is the core gameplay that's great. SJ does have quite a few melee weapons, but I'm not sure if I will go to the trouble of implementing manual blocking etc. Without per poly collision, that might also be somewhat difficult unless one were to abuse skeletal objects for this. One day, this might be a worthy mod.

Apart from that, SJ will be very moddable anyway because it's open source and the pipeline is readily available (and there is a heap of technical documentation.)

[SusanMDK]

I'm always tempted with making my stuff in Doom. Now I'm thinking I'd make my game on DP, but also make a model/resource pack for Doom later... or something like that. Maybe that will solve my gamedev temptations/issues.

[jitspoe]

Hm... where to begin...

Lately, I've been tempted to try out some of those engines as a service deals (Unreal 4, Crytek, etc.), but they tend to cater to more of a high-end gamer, and that market tends to be kind of small (not to mention very time consuming to build content for).

Then I get tempted to build something for procedural content generation... I think about all the things that would be necessary to procedurally build towns and characters, then realize there's probably not enough time in my life to dedicate to it to ever come close to finishing...

I hear stories of people making lots of money generating content to sell, so sometimes I'm tempted to stop making my own stuff and just try pumping out a bunch of generic stuff to sell... but what's the fun in that?

Occasionally I'm tempted to make a mobile game that's super addictive and gets people to throw away a bunch of money... but I think that would make me feel too guilty.
... also tempted to just try to make a mobile game that's actually good...

I'm constantly tempted to write new mapping/compiling tools/map format. I might do that some day, but my ambitions

But my greatest temptation is probably to not work on anything useful at all...

[ceriux]

my biggest problem is starting something, then jumping to something else, till i have no time to work on anything at all... it sucks -.-

[leileilol]

My usual temptations are actually data projects for GPLed games, however there's always a pipeline-based roadblock to end them as I discover, which crush all of my immediate urges to make art.

WhatCouldHaveBeen:

-QUAKE2: For Blender, a working MD2 exporter doesn't exist. Anything that promises to be an MD2 exporter didn't export a proper MD2 that works with normals and all, as if they just wrote it for a NeHe renderer tutorial predating the Quake2 source release. Some have even accused me of spreading FUD about it, and a screenshot of a broken MD2 in Quake2's software renderer (compared with a non-corrupted Noesis MD3->MD2 result) is the silencer.
-HEXEN2: The pipeline for converting palettes, importing skins and setting flags is too excruciating as it still relies on a registered qME 3.1. Blender MDL exporter also flips skins and is off by an offset. It would've been bright (Heretic-ish palette) and anime styled to avoid confusion. A shame, as I was on a roll in Nov '08 with two completed player models.
-ROTT: WAD tools don't write ROTT sprites properly. 3DRealms is a dick.
-DUKE3D: I can write GRPs, ARTs and MAPS from scratch. Build license conflict. 3DRealms is a dick.
-SW: Ditto. Same reasons. Would have been anime-styled, rule63'd and humorless to avoid confusion. Also, 3DRealms is a dick.
-BSTONE: 3DRealms is a dick.
-RTCW: There is no interest in any Blender pipeline for MDS. I kept getting suggestions to use IQM instead, but that never ever worked. Also it's impossible to use MD3 with RTCW's animation system. Also, I would have taken the whole WW2 theme out with for something far more fantastical.
-JK2/JKA: The prospect of creating new GhouL2 compatible animations from scratch would have been overwhelming. The Jedi direction would've been discarded for a Heretic-ish fantasy universe though, when I was working with it, I called my data "Shaftpole Squires"
-Doom3: No scripts. It's so bare you can only do new from-scratch games on it. I got a lot of suggestions to switch OA to this after it came out. My S, T, F and U keys were wearing out.

What I kind of want to look at:

-A&A: Great first person RPG from 1996 that had a recent little-reported GPL release (take a guess why it was never reported among all those Free Gaming sites). The data is already Free as in beer, but it's rough, Poser-abusing mid-90's shareware developer data that hasn't aged particularly well. It could also use a refactoring, like low poly models replacing player armor sprites....in the software renderer. Although it's not a free roaming world game, it sure plays better than anything from Bethesda, new and old. Currently there are no GPL tools that can create the .PIC files and other lumps inside them (like weapon sprites) since it's its own data formats and such isn't compatible with any of the existing tools out there - except for the maps, since that's largely the same old Doom map format.

[qbism]

Quark is the only opensource program I know of that will save mdl normals properly, besides the command line tool to build from frames. For md2, maybe the command line tool qdata. Haven't tried it but I've used the mdl one a while back modified for lwo (or obj?) frames and it worked well.

[leileilol]

Those 'take a bunch of files into frames' programs are not an option. If you're an artist that wants to get their animation previewed in the game as soon as possible, they're a complete joke.

[qbism]

So what quake-ish formats, if any, can Blender export without a post-process? md3? Quakeforge's mdl script is good except for the normals, although I might be doing something wrong. It must be dipped in Quark for normals.

This is definitely a crux for OS game or mod planning on new models. Engine features and mapping tools might be awesome but if the modeling pipeline is hell the modelers aren't going to be crawling out of the woodwork to join up.

[goldenboy]

So what quake-ish formats, if any, can Blender export without a post-process?

IQM (for creatures etc), ASE, OBJ (for mapmodels; ASE requires changing a path with a text editor afterwards, which is so easy that I wouldn't call it post-process.)

[Spike]

I believe electo favoured PSK for ease of export. It can result in some excessive load times so don't use it as part of some final release, but it does tend to be fairly widely supported.