https://en.wikipedia.org/wiki/List_of_commercial_video_games_with_available_source_code

These days I've been seeing more open-source engine projects for commercial games rather than fully open source games (as in SuperTuxKart or Battle for Wesnoth).

I think the bar for a "good" game is set rather high, especially with regards to new games. Outside of large or legacy projects, it's probably outside the bounds of what many volunteer coders/artists can do in their spare time.

That said, 0.a.d comes to mind is a newer and high quality offering. And you'll never find a shortage of new open source roguelikes around either!

Quoting: utxIt seems the open source games available are older games that have been around for years. Does anyone still make new open source games?

Yes. Mindustry and Remnants of the Precursors are good examples that are decently polished and brand new (well, RotP is very inspired by Master of Orion but so are 99% of space 4X games). There are also FOSS game jams, clones of not-so-old games like Unciv (clone of Civilization 5), and games that are in the style of older games but are still new, even actively developed (like say Wyrmsun or Zero-K).

But it is a niche. There is little to be gained by developers, and it is a significant commitment. And by little to be gained, I mean that often no one will even want your source code; rare are the projects that lots of people will contribute to. Of course the big publishers and studios go for copyright maximalism anyway, but even say developers that care deeply about Linux might not feel particularly motivated to publish their source under a FOSS license - it can have significant downsides, and if no one cares anyway...

Many FOSS games are engine re-implementations or clones, made by passionate fans that are moved by their love of the original. Also perhaps by teams of software developers and (sometimes) artists that aren't necessarily game designers, or at least not designers able to create games that stand to these "classics".

Then there are the commercial games that at some point release their source, either with a real FOSS license or just source-available. That is really really cool, but it rides on the back of the commercial success while proprietary.

And of course, many FOSS games are small, short projects because of course that is easier to justify for a game distributed at no charge - the kind of game into which people pour years of work and often a lot of money has to consider those risks and opportunity costs very carefully.

I would like to live in a world where making FOSS games was easy and safe. But indie dev is already hard and risky regardless, and AAA publishers already do so much worse, that I'm just happy with what we have.

As an open source game developer, yes new open source games are a thing. Keep in mind that a game takes a long time to make (mostly) so by the time you play a polished open source game it probably isn't new. Some relatively new games of mine:

Mutant Road - Playable but only a single level. Still need to add jump kick.
Shallow Stone Solar - MVP state. Playable but missing a lot of features.
Learning BASIC on Mars - Missing lots of content to really teach you BASIC.
Pinball Disc Room - By design only a single room for the Disc Room Game Jam.

If you want to hear about new games people are developing you can checkout freegamedev.net.

Personally I enjoy finding old open source games which are fun to play. I spend a lot of time restoring them. My gamerzilla achievement system shows some games I've played recently. Although I really need to get back to adding support for more games. Mojotron: Robot Wars definitely needs support.

https://veloren.net is also a relatively new open-source game.

In the end Open Source games take a long time to be developed as they are usually just hobby projects, but it is also usually worth checking out older projects again for new features and polish.