Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Zarat]

I believe the games source will be released only when its done. Otherwise it will go the way of infiniminer.

Given the other references to Infiniminer, I decided to check it out in more detail.

First of all, the creator released the source only when he decided it was done. So what you are recommending is that DF do the same thing Infiniminer did. After he released the source, the creator doesn't seem to have actually made any new updates himself, or at least significant ones, and going by the repository information he stopped even accepting patches about a month after he decided to open source it. He also made a point of encouraging forks.

Secondly, Infiniminer was a much smaller game with a much smaller community. In other words, it didn't have staying power to begin with. And a large part of the players at one point, at least, apparently came from 4chan.

Thirdly, since mainline development halted, a few forks apparently came about, but other games (Minecraft) came up and outshined Infiniminer. Browsing through the Infiniminer forums, pretty much all of them had moved to Minecraft. This is exactly what I was talking about, that you have to stay ahead of your perceived competition. Infiniminer didn't do that, since the creator basically abandoned it because he considered it done.

Fourthly, according to the forums, the game reached its peak popularity and use after it was open sourced.

Fifthly, yes, griefers were apparently a problem, and using "hacked" clients enabled some of them. Not only does this obviously not apply to a single-player game like Dwarf Fortress, but it doesn't apply to any multiplayer game with a sanely designed client-server protocol. Apparently, virtually everything was offloaded to the Infiniminer client, and the server didn't even do sanity checking to make sure the client was playing by the rules. There's a reason nobody does this - you don't need to see the source code to cheat, it just makes it a little easier. Anybody could and probably would have done the same things without the source code.
[EDIT: I had read the game's forums and news posts/comments before I posted this, but I neglected to check Wikipedia. The hacked clients actually came before the code was open sourced, because people decompiled the binary and figured it out for themselves. Yes, he could have obfuscated the code, and C# is "easier" to decompile than say C++, but it would only change the degree of difficultly, not make it much harder or impossible. But in short: releasing the source code changed nothing.]

So can we stop comparing apples to oranges?

That comparison doesn't make sense.  Goblin Camp doesn't have a tenth of DF's feature set.  An actual fork that built on DF's existing feature set would be an entirely different story.  Also, I explicitly said that it was a guess, and that's kind of the point -- there is no good basis for predictions of this kind, and why would he gamble his livelihood on unknown odds when he wants to work alone anyway?

In my opinion, there is a pretty good basis for predictions of this kind - the history of every other open source project.

Firstly, Toady is the founder, creator, etc. and he doesn't have to allow forks to begin with. As I said before, very few projects are ever forked to begin with. So if forking were not allowed to begin with, we could expect even fewer.

Maintaining a fork is a lot of work. Toady is the main force behind DF, he knows the codebase better than anyone ever will, he knows where he's going with it, and so forth. He is the maintainer of the DF. When someone mentions DF, they will always go to him. The vast majority of developers will always coalesce around him.

Yes, if forks were allowed, then every now and then some disgruntled individual might create a fork because he thinks DF should have Feature X. He then has to figure out how everything works and interconnects, implement his feature, and then spend all his time backporting all the changes from the new DFs into his fork just to stay up to date, which is a huge amount of work. And he will always be at least one step behind. And he is very unlikely to have anything like the dedication that Toady is displayed.

I would argue that the odds are pretty well known, given how it's played out for everyone else. But my point that I think you're missing is that Toady is playing against bad odds right now, we just tend to ignore that because, as I said, it's easier mentally to stay the course. Merely maintaining DF, let alone adding to it, is going to become harder and harder and harder until it is an exercise in frustration to make the smallest changes. Meanwhile, if serious competitors do emerge, they will be working from a clean slate and will easily be able to overtake DF. This is virtually an inevitability - maybe not the competitors, you guys know better than I would how likely they are to come about - but the development hell Toady is working himself into.

Maybe the best analogy would be a fey mood dwarf working in a workshop that becomes increasingly cluttered.

[ayoriceball]

Maintaining a fork is a lot of work. Toady is the main force behind DF, he knows the codebase better than anyone ever will, he knows where he's going with it, and so forth. He is the maintainer of the DF. When someone mentions DF, they will always go to him. The vast majority of developers will always coalesce around him.

Yes, if forks were allowed, then every now and then some disgruntled individual might create a fork because he thinks DF should have Feature X. He then has to figure out how everything works and interconnects, implement his feature, and then spend all his time backporting all the changes from the new DFs into his fork just to stay up to date, which is a huge amount of work. And he will always be at least one step behind. And he is very unlikely to have anything like the dedication that Toady is displayed.

To add onto this point, I doubt that many of the current supporters of Dwarf Fortress will move onto some other fork and start supporting that instead. A lot of the people who helped to make this game popular will still stay with the Dwarf Fortress they know and love. I'm not sure why some posters here doubt the strength of the game's community.

[Footkerchief]

In my opinion, there is a pretty good basis for predictions of this kind - the history of every other open source project. [...] I would argue that the odds are pretty well known, given how it's played out for everyone else.

This is implicitly numerical claim that doesn't have numbers to back it up.  What percentage of open source projects get forked?  What about the ones that have elicited public reverse-engineering attempts before going open source?  Without numerical justification, I'm not ready to believe that statements about the average open-source project or even the average open-source roguelike apply to DF.  If this was a typical project, I wouldn't be here, and probably neither would you.

Firstly, Toady is the founder, creator, etc. and he doesn't have to allow forks to begin with. As I said before, very few projects are ever forked to begin with. So if forking were not allowed to begin with, we could expect even fewer.

I don't think Toady trusts a license to prevent forks.  Enforcing the terms of a license requires lawyers and money.

To add onto this point, I doubt that many of the current supporters of Dwarf Fortress will move onto some other fork and start supporting that instead. A lot of the people who helped to make this game popular will still stay with the Dwarf Fortress they know and love. I'm not sure why some posters here doubt the strength of the game's community.

Two important points here:
- this community has already spawned disgruntled fans who are attempting to make their own DF, so it's not like the community is an indivisible monolith of support
- potential audience, as I keep emphasizing, is as important as the current community, and they don't know who Toady is

[ayoriceball]

Firstly, Toady is the founder, creator, etc. and he doesn't have to allow forks to begin with. As I said before, very few projects are ever forked to begin with. So if forking were not allowed to begin with, we could expect even fewer.

I don't think Toady trusts a license to prevent forks.  Enforcing the terms of a license requires lawyers and money.

Who says you need lawyers to prevent forks?

[Zarat]

In my opinion, there is a pretty good basis for predictions of this kind - the history of every other open source project. [...] I would argue that the odds are pretty well known, given how it's played out for everyone else.

This is implicitly numerical claim that doesn't have numbers to back it up.  What percentage of open source projects get forked?  What about the ones that have elicited public reverse-engineering attempts before going open source?  Without numerical justification, I'm not ready to believe that statements about the average open-source project or even the average open-source roguelike apply to DF.  If this was a typical project, I wouldn't be here, and probably neither would you.

No, I don't think there have been any studies done on this sort of thing. But it's pretty clear from a long observation of open source projects that forks are rare, sustainable forks are rarer, and forks that weren't made for a pretty good reason (project being abandoned) are rarest. There is a KDE "fork" to continue maintaining the old version of KDE, but it remains to be seen whether that's sustainable, even though there's a pretty major demand for that sort of thing. When Oracle bought Sun, OpenOffice and and MySQL were forked, but because there was and is major uncertainty about Oracle's plans for the future - they are basically have nothing to do with Oracle's business model. There are a few Firefox and Chrome forks, but neither of them have achieved any mindshare worth mentioning. Apache, Linux, SMF, etc etc have no forks or no forks worth mentioning. These are products with tens or hundreds of millions of users, in some cases a large majority of which are developers. If they can almost never sustain a fork, why would the much smaller DF community be able to?

As for reverse-engineering, well, that happens to everyone who does anything remotely interesting.

DF is not a typical game, but it is not a special software project.

Firstly, Toady is the founder, creator, etc. and he doesn't have to allow forks to begin with. As I said before, very few projects are ever forked to begin with. So if forking were not allowed to begin with, we could expect even fewer.

I don't think Toady trusts a license to prevent forks.  Enforcing the terms of a license requires lawyers and money.

What stops someone from simply hosting the DF binary on their own website and charging for it/pretending to be Toady/asking for donations? They could even make it a minor sort of fork by editing the raws. They could with relative ease remove the Bay12 strings.

The same thing. But the only way you can stop people from doing that is with lawyers and money.

And again, that's pretending that maintaining a fork isn't immensely difficult and not really worthwhile.

Most people operate with good faith. If you ask them not to do something, or it's obviously inherently scummy to do it, they won't do it. And the ones that do anyway  aren't going to get anywhere.

Two important points here:
- this community has already spawned disgruntled fans who are attempting to make their own DF, so it's not like the community is an indivisible monolith of support
- potential audience, as I keep emphasizing, is as important as the current community, and they don't know who Toady is

This is kind of baffling to me. How are these people going to find these almost-impossible forks and yet never ever hear of Dwarf Fortress and Toady? Are there people who download Pale Moon who don't know what Firefox is? Why aren't people doing this already if this can be done - they don't even need the source code! And then assuming they operate in complete secrecy so that somehow people who find them have never heard of DF and Toady, and they don't acknowledge it on their website, are they going to disable forums and comments so people can't bring it up?

This is just kind of a bizarre argument I don't understand. Nothing like this has ever happened to anyone else, so far as I'm aware. Can you elaborate more on how this might happen?

[ayoriceball]

Toady One has a source of income, working on DF. Anyone who attempts to create a fork will not have this source of money, and will therefor have less time to work on their version of DF. This means their version will be lacking and less likely to succeed.

[Footkerchief]

Who says you need lawyers to prevent forks?

How else are you going to do it?  People do in fact steal code and pass it off as their own, and this would be a huge pain in the ass for Toady regardless of whether they develop a successful following.  People even stole code to add to DF as part of the Kobold Quest open-source experiment:

Incidentally, in the proposed model, I still don't think I'd be able to bring code back in to the core project -- the first Kobold Quest port attempt showed that people bringing in code don't necessarily check or care about the licenses of the code they are swiping, and I can't assume that responsibility for them (it has stuff from Wine which shouldn't have been there).

When Oracle bought Sun, OpenOffice and and MySQL were forked, but because there was and is major uncertainty about Oracle's plans for the future - they are basically have nothing to do with Oracle's business model.

"Uncertainty about plans for the future" sounds a lot like the stance of those who have reverse-engineered DF and/or advocated for open-sourcing it.  You're asking someone to make a decision about their life's work and livelihood based on your personal observations.  It's not convincing.

Here's another noteworthy example of a project whose loyal following abandoned it for a community fork.

Reading between the lines, it seems like you're saying "Nobody will take control of the project away from Toady unless developers in the community decide that it should be taken away."  And given how developers in this community have behaved, that's not a fantasy.

Two important points here:
- this community has already spawned disgruntled fans who are attempting to make their own DF, so it's not like the community is an indivisible monolith of support
- potential audience, as I keep emphasizing, is as important as the current community, and they don't know who Toady is

This is kind of baffling to me. How are these people going to find these almost-impossible forks and yet never ever hear of Dwarf Fortress and Toady? Are there people who download Pale Moon who don't know what Firefox is? Why aren't people doing this already if this can be done - they don't even need the source code! And then assuming they operate in complete secrecy so that somehow people who find them have never heard of DF and Toady, and they don't acknowledge it on their website, are they going to disable forums and comments so people can't bring it up?

This is just kind of a bizarre argument I don't understand. Nothing like this has ever happened to anyone else, so far as I'm aware. Can you elaborate more on how this might happen?

When I said "[the] potential audience [...] don't know who Toady is," I mean that a fork that adds new features would attract players who do not care about the identity of the maintainer.  When I started playing DF, it was several months before I developed any sense of loyalty toward the developer.

[MrWiggles]

I believe the games source will be released only when its done. Otherwise it will go the way of infiniminer.

Picked up by Notch and through a flook make millions again, then flounder and lack direction with the game?

[Starver]

This is kind of baffling to me. How are these people going to find these almost-impossible forks and yet never ever hear of Dwarf Fortress and Toady?

Put it on Steam?  I don't know ('cos I don't use it) what sort of IP protection checks the Steam administration put new submissions to, but without DF currently on there it's possible that people who use Steam as their primary game-getting would be shown this 'new' game from "Cove13Entertainment"[1] or somesuch and believe it to be original.

(I think the repetition of the various little bugs in the game would probably act as a Mountweasel should there be any need to assert that the NotDwarfFortressHonest available on Steam is a direct rip-off (even if they go so far as to polymorphically obfuscate the bare source code, as well as redo the entire raws so that all entities are different, to claim that only the raw format is 'inspired' by the original).  Even if they're going to the trouble of hacking the code to get rid of/utterly change some of the currently hard-coded aspects.  But still someone might consider going some way down that path...  Made so much easier with source-code released, of course.)


But I still think this is a red herring.  Toady just appears not to want to do this.  I've already told you why if I had a Dwarf Fortress-like project/brainchild thing out in the public sphere[2] why I wouldn't do it, as well, but I'm not sure if this is the same as Toady's reasoning.


[1] I thought that on top of the "Cove" for "Bay" swap, "Cove" as a synonym for "Rogue" or not-honest person was actually a good pun.  But while I know a lot of you will appreciate this, I realise that it's not so obvious.  So I'm spelling it out.  If you have to explain a joke it's not worth it, of course, but I still like what I did there so much that I'm going to the trouble of this footnote.  Consider it as a small thing amusing a small mind, if you wish.

[2] Fat chance.  I spread myself and whatever inspirational coding I do too thinly to get any of my various projects in a "here World, have a look" state.

[ayoriceball]

Two important points here:
- this community has already spawned disgruntled fans who are attempting to make their own DF, so it's not like the community is an indivisible monolith of support
- potential audience, as I keep emphasizing, is as important as the current community, and they don't know who Toady is

This is kind of baffling to me. How are these people going to find these almost-impossible forks and yet never ever hear of Dwarf Fortress and Toady? Are there people who download Pale Moon who don't know what Firefox is? Why aren't people doing this already if this can be done - they don't even need the source code! And then assuming they operate in complete secrecy so that somehow people who find them have never heard of DF and Toady, and they don't acknowledge it on their website, are they going to disable forums and comments so people can't bring it up?

This is just kind of a bizarre argument I don't understand. Nothing like this has ever happened to anyone else, so far as I'm aware. Can you elaborate more on how this might happen?

When I said "[the] potential audience [...] don't know who Toady is," I mean that a fork that adds new features would attract players who do not care about the identity of the maintainer.  When I started playing DF, it was several months before I developed any sense of loyalty toward the developer.

How would those people be able to (somewhat successfully) sustain their own version of Dwarf Fortress? You're saying they could get the support they need from the community. Is this where they would get their new players? A lot of newcomers to Dwarf fortress have come from multiple sources on Youtube for example. They could not possibly make enough money to work on their on version full time to surpass Toady One's Dwarf fortress in quality. Toady's will (is) be the best because he can afford it.

[FallingWhale]

Infiniminer didn't do that, since the creator basically abandoned it because he considered it done.

Fourthly, according to the forums, the game reached its peak popularity and use after it was open sourced.

Okay, now I'm considering you stupid.

Barth left when the game shattered because of forks, he wasn't the first to post the source.

The forums peaked when people were trying to find safe harbor.

[ayoriceball]

Infiniminer didn't do that, since the creator basically abandoned it because he considered it done.

Fourthly, according to the forums, the game reached its peak popularity and use after it was open sourced.

Okay, now I'm considering you stupid.

Barth left when the game shattered because of forks, he wasn't the first to post the source.

The forums peaked when people were trying to find safe harbor.

The forks were incompatible with other forks. This messed up the multiplayer because version A would be incompatible with version B. On top of that, multiplayer was further ruined by hackers.

Zach did not publicly release the source, but it was leaked because he didn't properly "protect" the code.

[FallingWhale]

it was leaked because he didn't properly protect the code.

[ayoriceball]

it was leaked because he didn't properly protect the code.

He didn't compile it I think. So people took advantage of it. Dwarf fortress, on the other hand...

[Zarat]

Infiniminer didn't do that, since the creator basically abandoned it because he considered it done.

Fourthly, according to the forums, the game reached its peak popularity and use after it was open sourced.

Okay, now I'm considering you stupid.

Barth left when the game shattered because of forks, he wasn't the first to post the source.

The forums peaked when people were trying to find safe harbor.

The Infiniminer website has a post where he releases 1.5 with the announcement that "this will probably be the last major update." Previous posts said that he wouldn't release at all until he was done.

A post well after the "last update" post announces the open sourcing of the code.

I don't think there's anything stupid about coming to the conclusions I did.

He was the first to post the actual source code. The other source code that was used to make the other clients was derived from decompiled versions of the binary. This could happen to Dwarf Fortress right now. (C# is nicer to decompile than C++, yes, but that's it.) The forks that "shattered the game" - quite an unusual description - appear to have been limited to cheating in multiplayer until the actual source code was released, after which the creator quit work on it.

Also, I wasn't measuring the peak of the game based on the number forum threads. There was a convenient thread on "when was the big Infiniminer peak."

But either way, nothing about Infiniminer is an example of what would happen with Dwarf Fortress.

1. The game was already totally reverse engineered.
2. The creator encouraged forks - you had to have a public fork of the code to even have a chance of getting it into the main branch, which he quit working on anyway.
3. The creator quit, which essentially killed the project, although a couple of forks were apparently successful for a time, and some severely curtailed the possibility of cheating. In other words, opening the source saved what was left of the game.

Hyperbole like "forks shattered the game!" gets no-one anywhere.

Who says you need lawyers to prevent forks?

How else are you going to do it?  People do in fact steal code and pass it off as their own, and this would be a huge pain in the ass for Toady regardless of whether they develop a successful following.  People even stole code to add to DF as part of the Kobold Quest open-source experiment:

Incidentally, in the proposed model, I still don't think I'd be able to bring code back in to the core project -- the first Kobold Quest port attempt showed that people bringing in code don't necessarily check or care about the licenses of the code they are swiping, and I can't assume that responsibility for them (it has stuff from Wine which shouldn't have been there).

When Oracle bought Sun, OpenOffice and and MySQL were forked, but because there was and is major uncertainty about Oracle's plans for the future - they are basically have nothing to do with Oracle's business model.

"Uncertainty about plans for the future" sounds a lot like the stance of those who have reverse-engineered DF and/or advocated for open-sourcing it.  You're asking someone to make a decision about their life's work and livelihood based on your personal observations.  It's not convincing.

By "uncertainty about plans for the future" I mean that it was/is highly uncertain that they had a future, and Oracle was making clear moves against them.

A little bit of a different scenario, there.

Here's another noteworthy example of a project whose loyal following abandoned it for a community fork.

Uh.

The "loyal following" of XFree86 made the switch because of major legal issues stemming from the decision by the XFree86 president to change licenses to one that was not GPL-compatible. If you notice the list of "loyal following," it was a bunch of GPL-licensed projects/organizations that developed GPL-licensed projects (or even more permissive licensed projects.) So they had a choice:

1) Change the license of all their code, which for some projects, is not feasible, because every contributor owns their code in perpetuity.
2) Fork XFree86.

That's why "GPL compatibility" was such a big issue.

This is a totally disingenuous example. There's no analogy that could be made whatsoever with DF. This was almost purely a licensing/legal issue.

Reading between the lines, it seems like you're saying "Nobody will take control of the project away from Toady unless developers in the community decide that it should be taken away."  And given how developers in this community have behaved, that's not a fantasy.

No, what I was saying was that:

1) The correct license forbids forking, so this entire discussion is moot anyway. If your argument is that someone could steal DF illicitly anyway, they can do that right now, they don't need the source code. It would be perfectly trivial to do so.
2) Even pretending that (1) isn't true, history tells us that nobody would "take control of the project away from Toady" unless Toady quit working on DF.
3) Even pretending that (1) and (2) aren't true, history tells us that any fork would be a failure, because of the difficulty of maintaining a fork, writing your own updates, and backporting changes in the real DF.

And again, the crux of the problem is this is less risky than the alternative, which is to change nothing.

This is kind of baffling to me. How are these people going to find these almost-impossible forks and yet never ever hear of Dwarf Fortress and Toady?

Put it on Steam?  I don't know ('cos I don't use it) what sort of IP protection checks the Steam administration put new submissions to, but without DF currently on there it's possible that people who use Steam as their primary game-getting would be shown this 'new' game from "Cove13Entertainment"[1] or somesuch and believe it to be original.

If it were that easy, what stops them from doing that right now? Anyone can spend 10 minutes, remove the Bay12 strings and replace it with Cove13Entertainment, and post it anywhere. Maybe demand payment, donations, or whatever. Or heck, not even bother removing the strings, just pretend you're Toady and put it up for $10 in your account.

First of all, DF is a reasonably famous game. It was in the New York Times. Secondly, no, Steam isn't totally irresponsible, and even if they weren't, I am pretty sure someone there has heard of DF, even if only from the NYT.

When I said "[the] potential audience [...] don't know who Toady is," I mean that a fork that adds new features would attract players who do not care about the identity of the maintainer.  When I started playing DF, it was several months before I developed any sense of loyalty toward the developer.

You make it sound so very easy.

As I said, assuming this happened, such a fork would always be at least one step behind DF.

And they're going to have a hard time hiding the existence of the "real DF." If they call themselves DF, people are going to Google it and get the real version. If they don't, then short of disabling all third-party feedback on their website and the Internet as a whole, they're going to see comments like "hey this is a DF ripoff." And I don't think anyone is going to play any version of DF without googling for help.

So in short, we're relying on:
a) People willing to be scummy criminals, and not just players with a misguided passion for the game.
b) Those people have to be really stellar developers. Plural. They will need a team. A dedicated team. In which case, why didn't they just write it from scratch again? - almost certainly a lot easier than figuring out how things work.
c) People have to shut out all third-party comment, or they will just be sending people on the way to the "real DF."
d) Development on the real DF won't change at all, despite the open sourcing. Because a moving target is impossible to hit, let alone come out ahead of.

it was leaked because he didn't properly protect the code.

There is no protection he could have used that would have made a real difference. The code was gleaned from the binary. .NET and JVM-based languages make it a little easier, but the only way to prevent it is not to let anyone play your game. You can obfuscate all you want, it doesn't make that big a difference. Back when I was in grad school, that's actually what I worked on.

The SOURCE CODE was not leaked. The BINARY was decompiled and the source code gleaned from that. The ORIGINAL source code, actually written by the developer, was released later.

The same thing could happen to DF right now. Any developer with a brain could do it.