In case you haven't heard the news already, development on the popular 2D block mining game Terraria has ceased. A lot of people are disappointed to hear this, since at this point it would appear that a lot of features that people want will never materialize. What I'm here to talk about today are the inevitable open source clones that will no doubt pop up in Terraria's wake. There's a power vacuum right now, and it would be nice to see a FOSS game capture the imaginations of a massive audience now left clamoring for developer support. For the people considering starting up a clone project, here are some things you ought to take into account:
First and foremost, don't just make a clone.
If you want people to actually play your game in large numbers, you need to avoid the feeling of it being a knock-off. If you're thinking about naming it OpenTerraria or Freeraria or something like that, stop right now and consider the reasons why an experienced Terraria player, who has their own Terraria server and friends they like to play the game with, would jump ship to come and play your game. (There's nothing wrong with this, per se, but understand that you'll be fighting an uphill battle in terms of differentiating yourself from a product with an entrenched user base.)
Instead, take a look at Starbound and all the sweet new features it's bringing to the table and how much excitement it's generated in the few weeks since it was announced. Starbound isn't Terraria. It's inspired by Terraria, certainly, but it's already its own distinct game with an interesting setting and features that differentiate it from its predecessor.
Think about what Terraria is at its essence -- a 2D, action-oriented, block mining, item collecting platformer -- and consider how you might take that concept and build something unique with it, rather than just making a game that isn't quite Terraria. This means coming up with cool ideas that people didn't even know they wanted -- not just some more types of blocks or weapons or bosses, but entirely new and different concepts that raise your game from clone status to the much more desirable status of spiritual successor.
Check for similar projects before you start one.
There might already be someone else working on an spiritual successor of Terraria. If that's the case, help them. (I'm not aware of any currently in progress, but I'd be happy to link to any promising ones if people tell me about them.)
Design it with multi-player in mind from the ground up.
Yes, this takes the game into ambitious territory, but let's be honest here. One of the big draws of these sorts of games is the multiplayer experience. Miss out on that and you aren't going to get much interest at all. Terraria's multiplayer support is fairly weak, and as such people will expect yours to be better. Fail them on this point and you've lost most of your audience. Do significantly better and you've got a chance to win people over.
Make it moddable...
"Is it moddable?" is one of the first questions people ask about games nowadays, especially games of the block mining sort. The answer to this question should be a resounding yes. Not only should it be moddable, optimally you should provide tools to make it easy to mod, which means something more than just XML and a text editor.
I've advocated for easy content creation before, and one thing people always come back and say is that if you can't work with XML and a text editor, then there's no way you could possibly make anything worthwhile. Not only is that opinion wrong, it's also a turn off to the bread and butter members of your community, who may be perfectly adept at using GUI applications, but probably don't know much about editing XML.
Of course, if you want to make a game that nobody pays attention to, by all means force people to use XML and call them idiots when they ask for a graphical tool. :)
...but make sure there's compelling content.
This is an easy one. If people don't want to play your game, they won't want to mod your game. If there's no initial content, other people aren't going to come in and add it for you.
Make it easy to install and run.
Don't make me install a ton of dependencies. If I'm running a modern Linux box, I shouldn't need to install a bajillion packages just to make your game work. If I'm running Windows, I probably don't want to muck around with downloading the latest version of your favorite bytecode interpreter to run your game. I want to run the installer and have it appear on my start menu. If I'm running a Macintosh, I want to run it... well, however Mac people usually run their games (*ahem* perhaps dual-booted into Windows? -- kidding).
Point is, don't make it a pain in the ass to set up. Make it a couple clicks.
Make it run on Linux.
This is another no-brainer. If you want the FOSS community to be interested, make it run on a free OS, and make sure it works on a platform that isn't a patent trap. This will also have the benefit of making it easier to...
Port it to Android.
Mobile platforms are where it's at nowadays. Support Android and you'll make a lot of people very happy. My understanding is that there are some license issues with the iPhone app store and FOSS, but if you can at least get it to run on jailbroken iPhones, all the better.
Make it look nice.
I saved this one for last. People care a lot about graphics. If you can make a complete game with placeholder art that meets the above criteria, I'll personally pay to commission art for you.
So there you have it. If you're still thinking about making a Terraria successor, you've got a long project ahead of you. Better get coding. :)