The AI is surprisingly good, and the character models are as good as Age of Empires. Bonus points that it is in all major Linux distro repositories if you're one of those kinds of people (like me!).
Honest question... did it get better recently? I tried it years ago... and the AI was non existent...as in, I built up an army and went to explore, only to find my enemy had like 4 peasants.
If you watch their svn history, they consistently are chipping away at it. I simply wait (to upgrade to a new 0ad) with the latest version in each new Fedora release. It gets progressively better from an AI, music, and artwork standpoint.
0AD is definitely not the same game now as it was years ago, it is massively improved. These were both pretty big releases:
I find the AI to be challenging to beat on hard, but not impossible. I'll play as the Ptolemies, because I know that they were the pinnacle of the Greek Phalanx lead by none other than Alexander the Great. So I was trying to take a large group of spearman up to take over an enemy town center filled with archers but wasn't having a ton of luck. Every time my infantry would retreat, the 20 archers would come out and chase them down. When I'd turn back, the archers would head back to the town center and shred through them. My winning strategy involved staging 10 companion cavalry to the east and have the infantry attack the town center and then retreat south, drawing the archers out to try to take out the infantry. The companion cavalry rode north of the archers cutting off their retreat to the town center and let the infantry cut them down.
The newer AI will also scout out your location and attack weak spots. They'll purposefully avoid attacking through walls and lots of towers if you mistakenly leave an unprotected rear or flank that they can use to march right up and destroy your town. It is certainly more difficult now and the new graphics make it a lot of fun to play.
Speaking of AI, there's pretty interesting project which aims at creating 0AD bot that uses hierarchical planning, not just some random heuristics: https://github.com/agentx-cgn/Hannibal
Another stupid question: did performance improve in the last ~two or three years? Even on the kinda small maps back then it quickly ran into CPU limits.
The Open AI guys made a very impressive Dota 2 AI that trained itself through self play (and against pros), there is a pretty interesting write-up about it:
To shed some light on why they won, they chose a character (Shadow Fiend) that heavily relies on tight timings to execute moves. Naturally, AIs beat humans in events where timing is the key differentiator. Even so it’s quite impressive
Watching the video sure brings back memories. I spent many an hour as a young lad playing Age of Empires, I always loved the sounds. It looks like they did a good job of recreating AOE audio.
Sprites in a 3D game look horrible for most objects: buildings, characters, terrain. Sprites are typically used for particle effects, grass, or objects that are far away. The latter is often used a performance optimization.
AoE 1 and 2, yes. AoE 3 is 3D. True 3D works fine for RTS games and generally looks better, and can apply dynamic shadows, physics, etc. If you are simply saying that you prefer the old-style graphics, then okay. But your initial statement wasn't worded well if that is what you meant.
I've made my own custom map of Taiwan with the correct locations for gold, stone, and types of fish in the area. I plan to screen-record some famous historical events and make a video, but I'm still figuring out how to make such a video interesting and short enough to actually watch.
For what its worth I would watch just based on your description alone. May be better to just make one and get feedback than focus on making it interesting. Also, please let me know once you do. :-)
AoE2 is the one game I play regularly. It's great that this project was started, but is it still active? Did anyone manage to compile and run it, and if so, how does it feel?
Not to belittle the nice work, but I feel there was a better reason for making this clone a few years ago. There has been a "HD" version on Steam for a couple of years, which finally has good network support and most of the bugs seem ironed out, after almost 20 years! The only thing I still miss is to be able to run it on Linux...
The link shows that the last commit was a week ago... so, yes, it's still active. It compiles and runs, provided you have access to the media assets. It still needs work though.
Having a FLOSS engine for this game is more desirable than having to rely on a proprietary mess like Steam to play AOE2. So I hope this project is eventually successful.
I had a look at some .NET code integrating steam once and the integration itself looked quite clean and lightweight.
I'd be interested if people got banned or something like that for modifying games when they were clearly not hacking/cheating/cracking/impacting the experience of others.
From DSfix for Dark Souls I know that it can be fine with them if you hook function calls to improve rendering for example.
I imagine you also have to distinguish between Steam and VAC (anti-cheat) integration.
This project has nothing to do with Steam, other than perhaps (optionally) using the assets from the steam version of AOE2 HD, but it doesn't interface or interact with Steam at all AFAIK.
Sure, I know, but the parent poster complained about how Steam was a terrible thing about the HD version. I wanted to know if it creates practical problems if you want to mess with a game that has Steam integration.
Citation needed. In my experience, running AoE2HD in Wine results in strange mouse errors at the start of a match which can ruin a game (where the screen is stuck scrolling to one side).
The mouse bug seems to be gone for me (it was mainly triggered by alt-tabbing out of the window). It still has an issue with text not being rendered properly while typing.
> We have an integrated terminal emulator supporting ecma-48. You can run vim or anything else within openage. This is neat to interactively edit scripts.
The approaches are dramatically different, though - this implementation seeks to discover and re-write the game logic from the outside in (black box), using reverse-engineering to understand things like the assets file formats. OpenRCT2 started as a direct translation of RCT from the original assembler into C (via Hex-Rays) and builds on the original game engine verbatim.
What I'm always surprised is the lack of tooling and test code and I'm of the belief that in the war of the clones, the ones that chooses tooling and a stack designed for larger teams, and a focus on BDD will get your project where it needs to go.
Instead of building a game, focus on building a team.
Great work on this clone so far, I was just talking top of mind.
I did not downvote you but I suspect you might have been downvoted because you "stole" someone else's thunder so I would suggest you post your clone in a, perhaps, Show HN post.
Interestingly, it's pretty common to also mention your own project in a comment, and how it's received seems to differ quite a bit based on how it was presented. If they start out with relevant comments about the current submission, a segue-way into a somewhat related project is usually taken well. If you start off with what can be taken as self promotion and take a while to get to relevant comments about the current submission (or never do), it's not received well.
In this case, I'm actually surprised that it was downvoted. After the initial line with a link, it's all fairly general and about clone projects in general, which makes that line read as presenting bon fides and not self promotion when the comment is interpreted as a whole.
To be honest I can't even tell that I've been down voted. I never thought to bother posting my clone to hackernews and likely wont unless I do a postmortem when the game has been released and if I have something worth reading.
My market isn't really here on Hacker News, there's already a built in community on Discord which is where its worth link baiting so I don't even have to promote.
https://discord.gg/bpk6ue
There are sum of 100 hard code people or less and thats all I'm likely to reach. My only motivation for my clone is so the community I already participate in has something for competitive play.
I could certainly say I could have written something with more breath to it and that might merit a downvote.
I remember I once wanted to hack on the open-source CIV4 game to give it a friendly UI game but found the codebase to be as unfriendly as the one linked above.
Contributing:
Being typical computer science students, we hate people.
Please don't contact us.
Nobody likes Age of Empires anyway.
None of you is interested in making openage more awesome anyway.
We don't want a community.
Don't even think about trying to help.
Guidelines:
No bug reports or feature requests, the game is perfect as is.
Don't try to fix any bugs, see above.
Don't implement any features, your code is crap.
Don't even think about sending a pull request.
Please ignore the easy tasks that could just be done.
Absolutely never ever participate in this boring community.
Don't note the irony, you idiot.
To prevent accidental violation of one of those guidelines, you should never
Why C++ instead of a higher level language like Python? The developers could focus more on game functionality than C++ bugs and cross platform issues. Performance should not be a big concern for running an old game on modern hardware.
Rule of thumb suggests that implementing AoE with Python today probably means PyAge2 runs about as well on a 4.5 GHz CPU as AoE2 did on a 266 MHz Pentium II. (not that well)
http://play0ad.com/
The AI is surprisingly good, and the character models are as good as Age of Empires. Bonus points that it is in all major Linux distro repositories if you're one of those kinds of people (like me!).