About free game development

[bzt]

Hi,

I've seen some interesting topics here.

1.) Could it be that we free game developers need more tools before we can concentrate on developing? Proprietary developers can buy a license of a ready-made engine or throw "human capital" at the development of the engine in-house. We have Godot and other engines but we still have less tools than the game industry, haven't we?

Fair question, but I think this is needless to ask. It's fairly common knowledge that proprietary game devs have more and better tools. I think a better question would be, are the tools we have enough? I mean, is all of the game creation process covered by at least one FOSS alternative, or is there a step lacking a proper tool?

2.) And we have less people but have to do the same job. And also important, we have less capable people I guess. Sure we have coding and art geniusses. But less than the industry has.

Yes, I agree, we definitely have less manpower. I'm not sure if they are less capable, or just less motivated though.

3.) Maybe we behave less professional than the industry pros? See old forum. (EDIT2: This is not pointing a finger at someone. Notice that I said "we". That includes me. And maybe being unprofessional is part of the fun in open source developing?

Good question. I don't know if this is true or not, less motivated sure. (BTW by definition "professional" means you get paid for your work, this obviously isn't the case here, and I don't think you meant that in its original meaning, rather you meant "experienced" or "capable".)

I think the biggest difference is, a professional game developer has to finish the game no matter what (boss says, "I don't care if you're capable or not, you'll get fired if this isn't ready by tomorrow"), while a FOSS developer says "give me a break, I'm doing this in my freetime, I'll finish it when I want it to".

Conclusion? Professional game gets done and released (even with bugs), while on the other hand free games are left unfinished, and sadly over time, abandoned.

Looping back to this, I think today most places of discussion are sort of agnostic towards libre/proprietary. The majority of folks you might see in an engine's community, or itch, or various social medias will be doing proprietary dev because it's the default, but there's nothing stopping FOSS folks from joining in--and I believe some do.

I think this isn't so. The reason why proprietary dev is the default and what's stopping devs from joining FOSS is the same IMHO: newcomers simply do not know about FOSS. If you don't know there's an alternative, then you can't join in, and if few people are joining in, then it's never going to be the default, and if it isn't the default then newbies won't try it out. This is a vicious circle.

Also very few devs are doing game dev for fun, their goal is mostly not creating a good and enjoyable game, but about 99% of them are greedy and just want a piece of the gaming market. And if they are choosing FOSS alternatives they choose that because they want to do it without any investment, and not because of the freedom. (There are exceptions, of course, 1%.)

There's opengameart.org about assets.

OGA is a perfect example of that latter statement. If you take a closer look at the "free" assets there, you'll realize that they are unusable in a game, and pixel artist do that deliberately. They are not thinking about creating some great free assets that FOSS game developers can use (save 1%), most of them are thinking about OGA as just a free advertisement platform, where they release something half-ready and they never say, but they expect you to pay for the missing parts. (And they get mad at you if you point out that their submission is lacking something.) Well, this isn't the libre / free software development spirit for sure.

EDIT: changed url because topic was splitted.

Cheers,
bzt

[PeterX]

I think the biggest difference is, a professional game developer has to finish the game no matter what (boss says, "I don't care if you're capable or not, you'll get fired if this isn't ready by tomorrow"), while a FOSS developer says "give me a break, I'm doing this in my freetime, I'll finish it when I want it to".

I think that's the reason why we will never be like the industry or like the indie developers. Be it the eagerness to complete a project or be it our way of communicating.

I think we have some advantages, too, but that's another story.

Greetings
Peter

[bzt]

I think that's the reason why we will never be like the industry...

Frankly I don't mind that at all. Industry games are looking good, but save a few exceptions are terrible as a game (I mean no original idea, no genuine enjoy, pretty limited by genres; just small additions and a few tweaks to the good ol' and endlessly repeated recipes).

...or like the indie developers.

On the other hand I think we should make games like the indie scene. Does not always looking as good as an AAA game, but quite often built around some new idea, wide spectrum of genres, and you can just feel there's lot more love and enthusiasm put into those games than money. I think that could be (and should be) achieved by FOSS games too.

Be it the eagerness to complete a project or be it our way of communicating.

Yeah, I agree. The question is, what makes indie devs finish their products (not money nor a terrible boss for sure), that FOSS game devs lack. We should adapt that mentality too, and there would be less unfinished FOSS games.

Maybe improve regular game jams? I mean most game jams expect you to create something in a week or two. That's impossible, no good game can be made in that short time. I'm thinking something more along the line of Indie Boundle, a yearly announced competition where any games can be submitted, no matter how long it took to make, the only criteria would be it has to be finished and playable. Is there anything like that for FOSS games?

I think we have some advantages, too, but that's another story.

I'm really interested about the FOSS tools you've mentioned. Maybe it would be a good idea to create a page on LibreGameWiki about a list of curated FOSS tools? @Julius what do you say? I mean we should surely do better than this (nice start, but only 3 tools listed?).

And if there's really a lack in tools, I'm willing to implement some of those to help out the community. (For example people on OGA were asking about how to replace palettes, so I've created pngpal, here someone was asking for an easy to use atlas packer, so I've created spratlas, I wanted to preview and convert Minetest schematics outside of the Minetest game so I've created mtsedit, etc. I can write such little useful tools endlessly :-) )

On that note, my hands are full now, so I haven't started this project yet, but I had this idea of creating some FOSS 3D modelling tool for quite some time now. I think existing ones (like Goxel for example) are lacking way too many features to be called an usable tool, some (like netradiant) way too buggy and difficult to compile (and hence has way too many incompatible forks), while others (like Blender) went down the hill ever since company money started to pouring in. Currently there's no good, featureful yet easy to use FOSS 3D modeller that I know of. Do you (plural) know any perhaps?

Cheers,
bzt

[PeterX]

Industry games are looking good, but save a few exceptions are terrible as a game (I mean no original idea, no genuine enjoy, pretty limited by genres; just small additions and a few tweaks to the good ol' and endlessly repeated recipes).

I think that's true. That's one of the advantages of open source/free games.

Maybe it would be a good idea to create a page on LibreGameWiki about a list of curated FOSS tools?

Great idea. Do tools include engines? And libraries? Then I know several. And it's a wiki so we can add the tools we want.
I know:
Godot
Castle Game Engine
AGS
raylib.
GB-Studio (it's on Github so I assume open source)

Notice that in the new forum a list of libs already exists.
https://freegamedev.net/d/128-foss-libr ... evelopment
We can transfer them easily to the wiki.

I think a Godot-like engine would be cool that uses a different design/philosophy. For example Godot is meant only for high-powered, really new PCs. People don't notice that because they have a powerful PC.

Greetings
Peter

[bzt]

Great idea. Do tools include engines? And libraries?

I was thinking more in the line of GIMP, Inkscape, Blender, MagicaVoxel, Tiled etc., literally tools that you use to create games.
But engines and libraries are fine too, as long as they are FOSS. FYI the wiki already has quite some of those: see engines and libraries.

Notice that in the new forum a list of libs already exists.
https://freegamedev.net/d/128-foss-libr ... evelopment
We can transfer them easily to the wiki.

Yeah, that list has lot of entries not on the wiki, it would be worth adding them all!

I think a Godot-like engine would be cool that uses a different design/philosophy. For example Godot is meant only for high-powered, really new PCs. People don't notice that because they have a powerful PC.

Present! I always have issues with Godot games (and with most of the Unity games too, however interestingly some runs pretty fluently.) And it's not just the performance. Some games (for example this indie game) trigger a pretty annoying rendering bug in the Godot engine, and there's nothing I could do about it, see attachment. The noise gets a little bit less "crowded" if I set the gamma to a non-zero value, but never goes away entirely. Makes the game totally unusable.

Cheers,
bzt

[Technopeasant]

Present! I always have issues with Godot games (and with most of the Unity games too, however interestingly some runs pretty fluently.) And it's not just the performance. Some games (for example this indie game) trigger a pretty annoying rendering bug in the Godot engine, and there's nothing I could do about it, see attachment. The noise gets a little bit less "crowded" if I set the gamma to a non-zero value, but never goes away entirely. Makes the game totally unusable.

I agree with this. Even on my new rig a lot of Unity and Godot engine games are still kinda shaky. For simple 2D games I personally think GDevelop is a better fit, and I think is the best competitor we have on the FLOSS side for GameMaker or Construct.

I do not not however think the problem is (just) tools. While a lot of indie games do indeed use Unity or GameMaker or whatever, there are plenty that don't. Stardew Valley and Minecraft are some of the most successful indie games ever made and were both self-made engines.

I think there is some other cultural issue at play here. Part of that is I suspect that "open source" philosophy is based on continual development without an end goal. This is wonderful for applications, but for an artistic work this can lead to endless feature creep and an unfocused design that turns off players (I realize this is true also of early access games, but they have a reputation for a reason). My brother still refuses to play Freedoom "until it is done".

I have however been giving this article a facelift:

https://libregamewiki.org/List_of_free_ ... pment_kits

[bzt]

I do not not however think the problem is (just) tools.

Yes, we agree, there are certainly other factors too. However the lack of appropriate FOSS tools would be a major road-block (also a thing I could do something about), that's why I'm focused on those primarily. BTW, I did not necessarily meant just engines by "tools", anything used for creating a game goes (pixel editors, sprite animators, map editors, atlas packers, audio editors, music composers, ... whatever).

I think there is some other cultural issue at play here. Part of that is I suspect that "open source" philosophy is based on continual development without an end goal.

It is a common misconception that Open Source means "continual development without an end goal", this isn't true at all. There are plenty of actively used, yet hasn't-been-updated-for-years Open Source projects (think about libvorbis, libpng for example), also proprietary software are plagued by this "continual development" disease as well, so this is by no means limited to FOSS.

I have however been giving this article a facelift:

https://libregamewiki.org/List_of_free_ ... pment_kits

Nice! Thank you very much! Could you please add TirNanoG as well?

Cheers,
bzt

[Technopeasant]

Misconception or not, the premise often sticks. I certainly agree that libre software does not require continual development, but I am less convinced that open source does not, at least as far as Raymond's Cathedral and the Bazaar goes. Story or campaign focused single-player games, the kind I personally enjoy, essentially have to be designed as cathedrals in order for there to be a coherent artistic vision at play.

I note the examples you cited are libraries intended for a very specific purpose. I am not sure I would count those as using open source methodology, regardless of being free software.

TirNanoG is now included, though I am not sure the CC-BY-NC-SA file type thing is fully kosher with the wiki's libre requirements.

[bzt]

I am less convinced that open source does not

Let's just say people are tempted to change it when it's Open Source, and when it's proprietary then the marketing department is the culprit.

I note the examples you cited are libraries intended for a very specific purpose. I am not sure I would count those as using open source methodology, regardless of being free software.

Okay, you're right, my mistake to pick libraries only. Here are some classic UNIX tools: bc, cal, find, cut etc. All Open Source, yet no need to change them.

TirNanoG is now included

Thank you very much! One thing, could you please add "Android" and "web browser" to the platforms? Thanks!

though I am not sure the CC-BY-NC-SA file type thing is fully kosher with the wiki's libre requirements.

Well, a) the software is GPL licensed, b) CC-BY-NC-SA means truly free (as in free beer not just in free speech), and c) that's just one of the licensing options for the games, the default one (there are others too, because you can create even proprietary games). If you want, you can create for example a GPL licensed game with it, although I wouldn't recommend that in particular, but possible.

Cheers,
bzt

[Technopeasant]

I changed the article title to:

https://libregamewiki.org/List_of_free_ ... on_systems

As it was not exactly shaping into an article on engines in general, and that is kind of way too broad a topic on its own.

I added the Solarus Quest Editor.

GB Studio was removed because of an issue with a back-end, but I would not mind additional research on this.

https://libregamewiki.org/Talk:List_of_ ... bdk_engine

[freem]

However the lack of appropriate FOSS tools would be a major road-block

Well, I do not think FOSS lack tooling.
Blender, gimp, openscenegraph (and it's vulkan counterpart even if moderns GPU still don't have vulkan acceleration on debian...), are pretty good even if not dedicated to games.
Godot engine, castle engine, all the quake engines and ecosystem, tiled, cube2 engine...

What it lacks it documented tooling, with minimal, independent, examples. And to have that stuff up-to-date, put in a single place.

This works for code libraries (I spend quite the time on 3D AIs... there's only recastnavigation as FOSS lib around for navigating, and while it's features are good enough for any 3D game which have rather common movement means, it's documentation is... not nice. There's a reference for the API, and an "example" which is far from being minimal, but that's all, no high-level documentation, no hello world) but it also works for tooling: blender, for example, which, since it constantly changes the UI (for good reasons, I'm certain) makes users constantly stumble on deprecated youtube videos (maybe this changed, though, last time I checked things were much better than few years before, and last time was already more than few months ago).
I can tell you that daemon-engine for example is close to absence of doc, even openmw (not a game but is certainly a tool to create games, with it's map editor) is not clear (while being infinitely better) but at least they know that fact and are trying to build a minimap game with only FOSS assets and going to de-hardcode the engine after 1.0.0, which for once is a rather stable line.

So, that tooling can really be used only by specialists, and I doubt FOSS games often can afford specialists of every domain to create a game: AIs, gamelogic, rendering, sound, music, mapping, networking...
This may be partially normal, though, considering the fact I'm talking about rather big tools or big games/engines, several of the engines I'm talking about are based one way or another on previously AAA non-FOSS games, after all (quake3-derivatives engines, openmw...)

I think there is some other cultural issue at play here. Part of that is I suspect that "open source" philosophy is based on continual development without an end goal.

It is a common misconception that Open Source means "continual development without an end goal", this isn't true at all.[/quote]

I think it's true. non-FOSS games will be happy to put the 1.0.0 version on unfinished game, so it looks good enough to players, but FOSS projects will be very hesitant to do that, even if the game is very stable and mature.
It's like there's expectations for it to be complete as in, noting else will ever be added after 1.0.0, and thus the 1.0.0 target constantly grows and is never reached.
Examples: freedroid, tremulous (and it's derivatives), xonotic, openclonk ... of course, some project could move on 1.0.0 and more, too, but I feel there's less: red-eclipse, wesnoth, flare-rpg (empyrean campaign)... note I'm only mentioning rather big projects here, and that I'm forgetting a lot for obvious reasons (can't know all, and can't list all games neither).

[bzt]

However the lack of appropriate FOSS tools would be a major road-block

Well, I do not think FOSS lack tooling.
Blender, gimp, openscenegraph (and it's vulkan counterpart even if moderns GPU still don't have vulkan acceleration on debian...), are pretty good even if not dedicated to games.
Godot engine, castle engine, all the quake engines and ecosystem, tiled, cube2 engine...

Well, I don't know all of these, and I wasn't referring to engines in the first place, but here is my personal experience:

• Blender: sucks big time. Extremely difficult to use, rarely used features put in front, while frequently needed functions buried deep down in the menus. And pretty often crashes my entire X11 session on exit, which is extremely annoying. Took a deep dive in bloatedness ever since big money corps started to put money in it.
• GIMP: perfectly incapable as a pixel editor. I'm using it for more than 20 years, but there are still simple tasks that I couldn't figure out how to do with it, so I ended up writing pixel manipulator tools myself... Same issue as with Blender, hard to use menus (I hate that one of the feature I use the most, "Merge visible layers" is not accessible from an icon nor with a keyboard shortcut.)
• Tiled: is a perfect example of a bad tool. I could go on all day what's wrong with it. Seriously, after 7 years, still no tile atlas support? In a tiled map editor?
• Godot: never used myself to create a game, but I constantly have problems with playing Godot games. Just take a quick look at the screenshot above...

What it lacks it documented tooling, with minimal, independent, examples. And to have that stuff up-to-date, put in a single place.

Yeah, I agree. GIMP for example used to have a very good website, collecting all the know-how and plugins, but it was shut down for some reason unknown to me. Plugins and tutorials are now spread wide all over the internet, but luckily you can find quite a lot (usually outdated) with a little web-search.

it also works for tooling: blender, for example, which, since it constantly changes the UI (for good reasons, I'm certain) makes users constantly stumble on deprecated youtube videos (maybe this changed, though, last time I checked things were much better than few years before, and last time was already more than few months ago).

Nope, it's still the same mess. I don't watch Blender tutorials, but I'm an author and maintainer of a Blender plugin, and yeah, it's not just the UI that always changing, but the API as well... Blender's doc looks like a pretty good one at first, but when you start to look deeper, you'll realize it's good for nothing, only explains trivial things that you already knew... with out-dated screenshots and examples ;-) Yeah, I know your pain too well!

even openmw (not a game but is certainly a tool to create games, with it's map editor) is not clear

Oh, yes! You have no idea how much working hours I put into openmw just to get it working with translation!!! And there are still some windows where letters with acutes (like á, é) shows up as ';' or '<'... I totally agree about you saying doc is close to non-existent. Never could get the map editor working either. It starts, but in lack of any doc hard to figure out how to use it. Let's just say, not very intuitive.

So, that tooling can really be used only by specialists, and I doubt FOSS games often can afford specialists of every domain to create a game: AIs, gamelogic, rendering, sound, music, mapping, networking...

I know it's a big project, but that's exactly what I'm now trying to do with TirNanoG. Solve as much of the specialized requirements as possible, and let end users just pour in assets and press "Play!". It's supposed to be a click-click-finish game creator. (Oh and yes, I know the pain of lacking the manual too well, so I put extra effort in creating one, as well as documenting the file formats, even the most trivial ones to the last bit.)

I think it's true. non-FOSS games will be happy to put the 1.0.0 version on unfinished game, so it looks good enough to players, but FOSS projects will be very hesitant to do that, even if the game is very stable and mature.

I wasn't talking about the version per se, but the goal. I think non-FOSS games are also plagued by this "there must be continuous development" bullshit, not just FOSS games. Minecraft comes to mind for example, which is constantly changing and getting new versions, and for what? Is it really getting better?

Cheers,
bzt