Jump to content
Double Fine Action Forums
Sign in to follow this  
DF Oliver

Programming Update #2

Recommended Posts

I see a lot of talk about helping open source and whatnot, so I’ll just ask this to clarify:

How much of the game will be made opensource if any? Will you release the whole engine with improvements you’ve made?

This has been asked and answered a few times already, but the short version is that we'll definitely be contributing bug fixes and improvements to Moai back to the main git repository. Fully open sourcing the entire DFA project is something that's still TBD and not something we're going to think much about until the game is close to done.

Share this post


Link to post
Share on other sites
Does this system mean we might be able to play earlier builds of the game, rather than just a final Beta towards the end? Like a monthly build, perhaps? It’d be interesting for (many of) us, and if you can watch usage it may return quick cheap bug data?

Building DFA on an open source engine doesn't make it any easier (or harder) to distribute builds. Offering builds before beta is something the team is discussing and I suspect that if they do release earlier builds, that'll be posted separately and prominently.

Share this post


Link to post
Share on other sites
But it’s a real shame that there hasn’t been any movement in opening up the UbiArt framework.

Agreed. It's great tech and a great game. Sadly, which I heard some rumblings a while ago about opening Ubi-Art up, but it's been quiet for more than a year now. I seriously doubt that Ubi ever will open it up, which is a shame, but not surprising I guess.

Share this post


Link to post
Share on other sites
and it allows us to give something back to the community by contributing bug fixes for example.

This is, in my opinion, the best part of the whole post. :)

Share this post


Link to post
Share on other sites
But it’s a real shame that there hasn’t been any movement in opening up the UbiArt framework.

Agreed. It's great tech and a great game. Sadly, which I heard some rumblings a while ago about opening Ubi-Art up, but it's been quiet for more than a year now. I seriously doubt that Ubi ever will open it up, which is a shame, but not surprising I guess.

I've been discussing at the Toulouse Game Show last fall with M. Ancel about the status of the UBIArt framework. It was developed to build Rayman Origins with opening it in mind, but now, in order to do it for good, they are looking for a team inside Ubi to maintain, upgrade, and support it as the core teams wants to focus on creating games and not the engine.

So it's still on it's way, but not tomorrow sadly.

Share this post


Link to post
Share on other sites

It's nice that you're using this engine, but what about the tools for making scenes and that stuff?

I'd love to see some open-source tools for adventure game rapid development.

Share this post


Link to post
Share on other sites
It's nice that you're using this engine, but what about the tools for making scenes and that stuff?

I'd love to see some open-source tools for adventure game rapid development.

I'm going to write a programming update specifically about our toolchain. Look for it soon!

Share this post


Link to post
Share on other sites

What other adventure engine did you consider?

A lot of german adventures uses Visionaire Studio, but I'm not sure if it supports mobile platforms. Well, I only have a very old mobile phone and no ipad or whatsoever, so I'm just rooting for the pc version and hope every platform get the attention it needs. ^^

Share this post


Link to post
Share on other sites

We looked mostly at Adventure Game Studio and Wintermute for "off the shelf" adventure game solutions. The main downsides were platforms (only Wintermute currently has a viable mobile implementation and even that is a bit rough) and both of them had source code access issues (either totally or partially closed source plus legal encumbrances even on the "open" source bits).

Visionaire Studio looks like a great framework, but it has all of the above problems and then some. It's entirely closed source, only works on PC (Mac is coming "soon" and mobile "eventually" but "not soon"). Those are deal breakers for us and DFA. Their product is also designed to let non-programmers make adventure games, which isn't the right tech philosophy for DF and DFA.

Share this post


Link to post
Share on other sites
Visionaire Studio looks like a great framework, but it has all of the above problems and then some. It's entirely closed source, only works on PC (Mac is coming "soon" and mobile "eventually" but "not soon"). Those are deal breakers for us and DFA. Their product is also designed to let non-programmers make adventure games, which isn't the right tech philosophy for DF and DFA.

Visionaire probably would have been a good option if this was still a sub-$500,000 game and PC exclusive, though, for all these reasons. Save money and overhead on coding.

Using Moai seems like it will free you guys up a lot though. I hope you can take advantage of that and do something interesting in terms of presentation.

Share this post


Link to post
Share on other sites

MMMMMMM I like it, I like it allot. I love seeing and hearing about the development, what a great bonus!!!

Share this post


Link to post
Share on other sites

Could you please tell a bit more about this Moai framework? Is it essentially a 3D framework with textured objects as the main building blocks which are you going to tailor for your 2D needs? (like every 2D plane being a giant texture). If so, what kind of tailoring does the Moai graphic engine need in order to be suitable for DFA?

Also, with an open source project there is always a risk of some parts of the project getting less love from the devs than others. This is because some features are fun to work on and tend to attract a lot of people and some are just boring. If it becomes apparent that Moai support for, say, Linux or Mac isn't as great as for Windows, do you have necessary programming resources to fix this situation? Or are Linux and Mac users at the mercy of Moai devs?

Share this post


Link to post
Share on other sites
Could you please tell a bit more about this Moai framework?

Sure!

Is it essentially a 3D framework with textured objects as the main building blocks which are you going to tailor for your 2D needs? (like every 2D plane being a giant texture).

Moai's graphics run on top of OpenGL ES, which is the API used for pretty much all mobile devices and very similar to the standard OpenGL that runs on desktop computers. Sprites are represented as textured geometry (usually quads). We'll be targeting ES 2.0 which basically adds vertex and pixel shaders (similar to SM3/DX9). The quads can have "real depth" or they can simply be drawn at a fixed depth in a particular order. Things like z-buffers, stencil, etc. are all optional.

If so, what kind of tailoring does the Moai graphic engine need in order to be to be suitable for DFA?

We expect to write new shaders, do some optimizations to the core render loop, and possibly add support for per-vertex "skinned" animation (think Ubi Art). Of course there will also be game and platform specific bugs to fix. Basically, it's the usual sort of work you do when shipping a new game on an existing renderer.

Also, with an open source project there is always a risk of some parts of the project getting less love from the devs than others. This is because some features are fun to work on and tend to attract a lot of people and some are just boring. If it becomes apparent that Moai support for, say, Linux or Mac isn’t as great as for Windows, do you have necessary programming resources to fix this situation? Or are Linux and Mac users at the mercy of Moai devs?

We picked Moai because it is, today, really solid and pretty much ready for us to use (as you can see in the original Red Robot prototype). There are no features that don't exist today that we're hoping the community will create or else we're in trouble. We're well staffed for the work we need to do, and any work that the community does in the mean time (like the Linux port for example) just means we can put more of that engineering time into making the game more awesome.

Share this post


Link to post
Share on other sites

We picked Moai because it is, today, really solid and pretty much ready for us to use (as you can see in the original Red Robot prototype). There are no features that don't exist today that we're hoping the community will create or else we're in trouble. We're well staffed for the work we need to do, and any work that the community does in the mean time (like the Linux port for example) just means we can put more of that engineering time into making the game more awesome.

You hear that, programmy-linux-types? Get working on Moai's Linux support if you want to help the project get even better!

Share this post


Link to post
Share on other sites

Never heard of Moai but it seems like a good option. I'm glad you didn't go for a more closed framework

Share this post


Link to post
Share on other sites

Coupling the portability of C+ and Lua with the mobile compatability tools of Moai looks like a great option for rapid development on this project. I'm really impressed with how agile (and informative) your team has been so far. I haven't been following the forums much since they initially openned up to backers; but, looking back over these posts between updates 8 and 9 really makes me interested in seeing what the programming team gets up to.

I'm more of a GIGO failure than a programming wizard (dropped out of more programming classes than I finished). Still, seeing how coders work is kind of like getting a peek at something orgasmic and mystifying.

Thanks for showing me [us] a tiny sliver of "your special place."

Share this post


Link to post
Share on other sites

Thanks for the great post, Moai looks like an interesting SDK to be working with and I'll definitely have to read up more about it and it's uses. You mentioned that you use git, do you guys have a preferred windows git client as I've always found them to be as great as the ones that can be found on Mac OSX for example.

Keep up the good work and I look forward to the next post!

Share this post


Link to post
Share on other sites
Moai's graphics run on top of OpenGL ES

We'll be targeting ES 2.0

AFAIK on Linux OpenGL ES 2.0 is available for Mesa FOSS drivers for Intel, AMD and nVidia hardware (through EGL path) and nVidia proprietary driver (via GLX_EXT_create_context_es2_profile) but for Catalyst situation looks a little different. According to this AMD article you need to have working EGL setup to use OpenGL ES 2.0 with Catalyst. The problem is - Catalyst just doesn't have libEGL and libGLESv2 libs, this libs available in separate download. Not only in Linux version of the driver, bug also in Windows version too. User need to download and install it separately, if he want to use OpenGL ES 2.0.

I understand situation with Catalyst right? If yes, do you have plans to contact with AMD and ask them resolve this issue in future builds of the driver? I mean provide EGL and OpenGL ES 2.0 support out-of-the box with installed Catalyst driver.

Share this post


Link to post
Share on other sites
I understand situation with Catalyst right? If yes, do you have plans to contact with AMD and ask them resolve this issue in future builds of the driver? I mean provide EGL and OpenGL ES 2.0 support out-of-the box with installed Catalyst driver.

We'll likely use regular OpenGL on Win/Mac/Linux, just sticking to the feature set that's common to the two of them, perhaps with a small number of very common extensions here and there.

That said, we know some good folks at AMD, so it's probably worth reaching out to them and make sure we're on the best path for all plats.

Share this post


Link to post
Share on other sites

I like the fact that the game will use an open source game engine, imp lamented with C++, and also developed specifically for 2D games. I think that will offer the programs a lot of freedom and, as was mentioned, the ability to have bug fixes corrected easily, with the freedom of open source.

Share this post


Link to post
Share on other sites

Thanks Oliver and Brandon for the very interesting posts! And welcome to Anna too!

Me too I'm very happy with the FOSS Moai choice.

Share this post


Link to post
Share on other sites

There's never enough insight given to the technical and creative aspects of game development. We're usually just subjected to the same old PR and marketing crap that has an ESRB notice attached to a two minute glossy non-informative fluff-piece. As a software engineer in the enterprise world, I'm always very curious about what it's like on this side of the fence. Frankly, this is all part of the reason I backed the game. I think there is a much larger community out there - with more or even far less experience than myself - that finds the behind-the-scenes nuts-and-bolts development fascinating.

Share this post


Link to post
Share on other sites
I prefer Mode 7 myself. :)

The Moai SDK looks interesting. Hopefully I'll have time to browse their github repository tomorrow.

I fancied more Mode X. Those were the times ...

I´ll also will have to find time to take a lot at Moai (Lua is cool but I was never a fan of C++, probably learned it too early ...).

Share this post


Link to post
Share on other sites

Okay, but open-source here does not mean that I can easily access the source in this case.

Why do I have to sign up for that "cloud" to "get started"?

And are you really sure that you (DF) want to pay several hundred dollars each month while developing the game?

An old-school adventure should look IMO like old-school from the outside and the inside (code)!

Would you've used such an engine back in the days of DOTT? I guess you had not even had the possibility..

Share this post


Link to post
Share on other sites
Okay, but open-source here does not mean that I can easily access the source in this case.

Why do I have to sign up for that "cloud" to "get started"?

And are you really sure that you (DF) want to pay several hundred dollars each month while developing the game?

An old-school adventure should look IMO like old-school from the outside and the inside (code)!

Would you've used such an engine back in the days of DOTT? I guess you had not even had the possibility..

It's true that I can't find a link to the source anywhere on their site, but the source is available here. You don't have to sign up for anything. moai-dev has the latest version and moa-beta has a more stable version. You can download it all as a zip file.

https://github.com/moai

I think the dollars you're thinking of are for the cloud services, which are different from the SDK, the thing they're talking about in this update. I suppose it's possible that they're also paying for the cloud services, but they never said that.

As for coding it like an old-school adventure game... I see absolutely no reason for them to do a thing like that. They're not making an exact duplicate of an old-school adventure as we can see from the lack of jagged pixellated edges in the art updates, so why not take advantage of the latest programming technology as well?

Even if they were creating a game in that style, players just playing the game won't be looking at the code. It would only be relevant if they were creating this as some sort of case study or lesson plan about old school adventure games for other game development students to look at, and that's not what they're doing.

Share this post


Link to post
Share on other sites
Sign in to follow this  

×
×
  • Create New...