The Open Source Thread

[Velociryx]

Solver asked an important question over in the second journal entry thread, and I thought it deserved its own thread, so I'm starting one here.

The question he asked was "is it still in the works for us to release the source code for this project?"

and that's an excellent question.

That was one of our (numerous!) stated goals at the outset of the project.

Time, circumstance, learning and drift over the life of the project have been slowly pulling us away from that initial "open source" direction.

When I look back on the path we've traveled so far, so many of my assumptions were so dramatically off base that it's almost funny now, to think about how I viewed this project at its inception.

I still don't know everything there is to know....hell, not even close, but I'm a bit better off than I was before.

Anyway, before I ramble down that road and firmly off the topic at hand, the short answer to that question is...not at this time.

I'm still an avid fan of and supporter of Open Source. I still firmly believe in the methodology, however, there are a couple of compelling reasons for me to say "not at this time" at this particular time...

First and foremost is that this game is not going to be a "publish and forget" type thing. If it were...if we had no intention of providing ongoing support and development, then absolutely and without hesitation there would be an immediate source code release so that the game could live on, long after we were done with it.

However, that's not in the cards. I have every intention of continuing development on this project, long after its initial release, and the problem I see with that long-term support focus and open source release is this....well, for lack of a better term, I'll call it the "Unix Effect."

How many flavors of Unix are there?

Too many.

It's mind-numbing.

I don't want to see that happen to Candle'Bre, and by keeping the source here, we set ourselves up as the final arbiters for what's offical, and what's not, where the game is concerned, and I think that's an important bit. Pivotal actually, given my plans for the game.

Secondly, it's about as difficult to get a job with us as it is to fall off a log...shoot me a mail, and you're in, so it's not going to be hard for anybody who wants to see the source to get at it....it's just simply that, by keeping it under a singular umbrella, we maintain better control over the flow and direction of the ongoing development.

After....what's it been...two years now working on it, we have just barely begun to scratch the surface of this vast world we're creating, and I fully intend that everybody who wants to help us with those explorations will have the opportunity to do so. But like so many of my beliefs early on, this one too, has undergone some modification as the reality of Candle'Bre begins to settle over us.

-=Vel=-

[Kuciwalker]

Yeah, plus, weren't you going to sell it? That'd cause slight problems with any plans for open-source

[Velociryx]

yes and no. I mean, Linux is open source, and for sale. There are other examples of this as well.

But no...our plans where that is concerned is to simply give the game away.

Where the magic stems from though is the continual updates.

The exact methodology isn't worked out yet, cos we don't need to focus on that just yet (keeping the focus on getting the game out first), but ultimately, we'll have a secure forum set up and some means of downloading the incremental upgrades to the game (NOT the same as a patch...we don't intend to call the game "done" until we're happy with it, which means that there will be no further need for patching and such nonsense). By incremental upgrades, I mean robust additions TO the game itself, as we continue to explore the combinatronic and other possibilities.

We'll make that aspect of it nominally fee-based so we can eat, but more on that as it gets closer.

-=Vel=-

[Solver]

Vel,

I would like to make several comments on the issue. One solution would always be to release the engine source only, not the entire source - allowing other potential game makers to base their games on the Candle'Bre engine. For examples, the Unreal game series have been making their engines available to developers, and there are actually many games based on Unreal engines.

However, I'd like to say that it's very unlikely for a UNIX-effect, as you call it, to form with release of the source. The reasoning is as follows: clearly enough, any change to the source that has some significance will require good knowledge from those who make the changes. I don't know how big the source code is now, but it's definitely big enough not to be easily re-made by one man. Therefore, the only possible variations of modified Candle'Bre code could emerge after quite some time, AND from organized groups of people.

That does in no way interfere with your (great) decision to support the game in the future. The vast majority of players who will be doing any modifications whatsoever, will do it using whatever editing/modding tools are included with the game, as they're obviously easier than the source modification. And, based on these mods, you always have an easy way of determening what's "official", issuing updates to the game, etc. There will be official updates, and there will be player mods - while the first real modification of the source code into something that has any but the slightest difference from the original game would pop up like two years after the game's release.

Also, the release of the source has its benefits to the actual gamers - I have stated the same in regards to Civ 3, and will say the same about Candle'Bre.

1) Game Understanding - in a game as complex as Candle'Bre, Civ, or any similar, it's great to have access to the underlying code for players to figure out the formulas that drive the gameplay, etc., which in the end helps faciliate a much better strategy community.
2) Mods - yes, there are mod tools provided with the game, and yes source won't be changed heavily to mod the game. However, open source might give the opportunity to make small yet good changes to that structure, as in adding some additional function to the editor - which is easy enough to be accomplished in a reasonable time.
3) Open Source style debugging - don't let your users just report the bugs! If you have the source open, then the technically skilled users will also provide you with information relating to what causes the bug in the source, or even ready fixes it! No matter how determined you are to support the game, eyes of dozens of players looking for solutions to those bugs will make it all a game with even better patching experience.
4) AI - I'd be quite tempted to put this as number 1, as this issue has been brought up repeatedly with other games. Provided that the game has a sizeable enough fanbase (heck it will!), a group of gamers with programming abilities will form that would love to improve the AI. AI is, after all, a major factor in how entertaining TBS games are. You may remember, Vel, that you yourself mentioned to me that AI is the weakest point of SMAC. Now, whether Candle'Bre AI will be originally more competetive or not (I hope it will be good, of course), there's always room for improvement. And again, some users making smaller and maybe not-so-small sometimes changes to the AI, will result in much better experience overall.

As for selling the game & being open source - having the source doesn't mean you can compile it and play the game, you still need the data (graphics, etc.). So there's no overlap of problems....

I hope this post raises some valid points, and will be glad to answer further about possible concers on source release .

[Dale]

Ummm...... just to point something out here, the way I'm designing the code, modability will be so open that you can completely change it.

For instance:
- All graphics are simple 24-bit bitmaps. Anyone with a windows computer can redraw all the graphics of the game (windows paint).
- All the game engine data, including messages, events, configuration, etc are all simple ASCII text files. Notepad is what I used, and I'm sure everyone can find that on their PC.
- Scenarios are written in XML. This is essentially another simple ASCII text file, with some special flags to seperate content. Once again, notepad can be used.

So, for the advanced modder, there is so much opertunity to mess with the files. In fact, I can quite easily see how to convert the Candle-Bre medieval basin into a space game with solar systems (instead of provinces). Oh, was that Hex having a graphical heart-ache?

But for the part-time modder, someone who wants to write some fate cards for the game, or create their own faction to load in, there will be an editor.

Personally, I wouldn't like to see the code go public. The great game we're writing shouldn't need re-writing, and if it does, then tell us and we'll do it! As for the modding arguement, well it'll be open enough for any CTP2 modder (like me) to drool on. For releasing the engine? Not really any point to be honest. The core engine is pretty essentially based on a DirectX8 (yes I upgraded it from 7) standard game_loop cycle. There truelly is no secret code in the game (yet) you couldn't create yourself from lessons learnt at gamedev.net, gametutorials.com or any number of C++ books.

But it's also way to early to start talking about this with any definite certainty. At this point, if you want to see the code, join the team!

Dale

[Solver]

Dale, then why exactly don't you want the code to go public ? Does it have some parts so ugly ?

Rest assured, if I had more time, I would be on the team since the last year, to be sure!

[Solver]

Aww, cmon, Dale, Vel, Hex, Busta, has all been said on this subject?

[Velociryx]

Hey man....apologies for the delay....was so focused on getting the fate stuffs up that I sorta forgot about this discussion for a spell!

As fond as I am of open source (and I've done a fair bit where that is concerned...not with programming of course, since this is our first time thru, but with the SMAX guide, releasing all the older versions to the public domain), I gotta say that it doesn't appear to be in the cards for the foreseeable future for Candle'Bre.

So many things have changed since we started this big adventure, so I'd be remiss if I said that this was the final set-in-stone-answer to the question, but at least for the time being....

-=Vel=-

[Solver]

Well, thanks for the clear answer on that. I will continue to hope, though, that some time after the game is released you will follow the excellent example of some developers and release the code .

[Velociryx]

I'm countin' on you to keep reminding me now and again, too!

-=Vel=-

[ApolytonGuest]

There are many variants of "open source" that don't mean you need to give away your code immediately. For instance, there's "donation-ware" and "hostage-ware."

Basically, you set a numerical figure in monetary terms or in game unit terms. And when the donations or purchases reach the target amount of money or units sold, the source code is released. Perhaps this could be used for the fee-based portions.

Also, although your fears about the flavors of unix are valid, I would refer you to the game Nethack. Allthough you can certainly mod it, and it's opensource, there's a recognized and respected authority "The DevTeam" who is in charge of the official updates. And the variants are all based on the official updates.

It's kind of like how there are many flavors of linux from the various distributors, yet ALL of them are based on the official releases of the linux kernel by Linus Torvald. By making it open source, you allow others to modify and make their variants of the INTERFACE system, but not the core game.

For instance, Falcon's Eye is a isometric graphical interface for Nethack, yet it doesn't modify the underlying game of Nethack in any way from the official release. The key is trust and competence. When the general audience believes in the skill of the DevTeam and trust them, you won't see a lot of them changing the way the game works. They'll trust the DevTeam to work out the kinks in the gameplay and balance. But they will put effort into making programs that interface and improve yours, like the Hearse program that exchanges dead characters between Nethack players.

But things like Falcon's Eye, Hearse, etc, wouldn't be possible if the Nethack DevTeam didn't make their releases opensource. And yet there is still a central, widely acknowledged and respected core DevTeam who is the only authority in making changes to the game, and has managed to avoid the "Unix effect" thus far. The game has been around for over a decade, still has regular updates and revisions like new rules and character classes, a usenet group and fanatic fanbase, and avoided the "Unix Effect."

I would love to see Candle'Bre be to TBS what Nethack is to Roguelike RPGs.

[Leland]

Yup, I agree that the fear of losing control over the project is not a valid reason not to go open source. Making decisions is surprisingly hard, and if there is one central body working on the design (e.g the nethack devteam, or linus torvalds, or "velocigames") people will gladly let them take the lead.

Originally posted by ApolytonGuest I would love to see Candle'Bre be to TBS what Nethack is to Roguelike RPGs.

Turn-based strategies are a lot larger category than roguelikes. One could go as far as to argue that roguelikes are a subset of TBS's.