Wesnoth2, Wesnoth, Inc, and other things

[Dave]

Just for the record, and this may not be something you care about, but (mainline) art assets under CC-BY-SA-NC or with any other sort of non-commercial clause will also effectively prevent distribution by a bunch of Linux distributions (definitely Fedora), as non-commercial clauses are considered by {the FSF, the OSI, Fedora Legal, Debian Legal (I'm pretty sure anyway), etc.} to make a license nonfree.

As someone who enjoys getting Wesnoth from his system package manager, I would find this to be an unfortunate decision.

Actually this is something we definitely care about and we definitely want to see what we can do to ensure that it would be distributed in Linux distributions.

[TC01]

Just for the record, and this may not be something you care about, but (mainline) art assets under CC-BY-SA-NC or with any other sort of non-commercial clause will also effectively prevent distribution by a bunch of Linux distributions (definitely Fedora), as non-commercial clauses are considered by {the FSF, the OSI, Fedora Legal, Debian Legal (I'm pretty sure anyway), etc.} to make a license nonfree.

As someone who enjoys getting Wesnoth from his system package manager, I would find this to be an unfortunate decision.

Actually this is something we definitely care about and we definitely want to see what we can do to ensure that it would be distributed in Linux distributions.

Well, that's good to know.

I do agree that the GPL is a bad license for art and music, so one of the CCs or another content license would almost certainly be better. But the -NC licenses are rather problematic (for this reason and others that have been already mentioned above).

[Dixie]

First off, I haven't read everything and just glanced quickly over the github repository so maybe what I'm saying is already being considered. I remember an art contributor or somebody like this saying once that if they had something to change about Wesnoth, it would be the hex orientation. With the "flat" side up/down, you need 4 base frames (and associated animations): front, back, diagonal front, diagonal back. If the "flat" side of the hex was east/west, however, you could arguably pull off a single side-way baseframe (and associated animations), effectively reducing the workload by 4. Of course, it implies trashing most of the actual art assets, so it's a clear loss at this point, but it's also possibly an investment for the future. Something to consider if you are thinking about rebooting the whole thing, anyway. Also on the disadvantage side, playing from bottom to top (or top to bottom) and having units face you is potentially more engaging than a left-right dynamic with unit never facing you. I thought it was an interesting idea when I read about it, anyway, something to think about

[TheWhiteKnight]

I must say I am with mixed to bad feelings about that.. Basically, I believe this project will take time from the developers and all the responsible persons to maintain the original Wesnoth if the games go as a separate ones.. On the other side, if that version takes over, we may lost the current Wesnoth as we know it.. And for player like me, that is with the game almost from the start in 2003rd either way is not good..

While I understand your concerns, at the moment I don't believe there is any cause for concern, based on the attitudes of the current main developers (other than Vultraz).

I hope you are right, but lets see what will happen..

[johndh]

As an artist, I think NC seems even worse than a closed license. If you're planning on letting other devs make Wesnoth spin-offs with some kind of royalty payment using Wesnoth2 assets, then that doesn't require a CC license since you'd be setting up an explicit agreement. NC sits in a weird middle ground where it's not quite free, but it still propagates like free content, potentially infecting free projects and turning them non-free. I guess I'm just concerned with any propagation of NC assets at all, which I (and many/most free software advocates) see as a Bad Thing.

I think NC is better suited for stuff that you want people to be able to use for personal use, social media, etc. You can put an XKCD comic on your Tumblr without any copyright concerns (except maybe in Germany), but that's about all you can do with it. If someone uses an NC asset in their game, depending on jurisdiction and judicial rulings, they may not be able to accept donations, have ads on their site, and basically have to operate at a loss by paying for their hosting out of pocket with no compensation, which limits the size of the project and basically punishes the developer, so I'd hate to see valuable assets distributed in this way. A dev would be better off just hiring an artist or paying royalties.

If you're concerned that having a closed license would be bad because it is bad for contributors, just make it non-exclusive so the contributors can still use their work freely and distribute it under any free licenses they desire. If you're concerned that it wouldn't be "giving back" to the community, you could set some clause that says any asset that gets replaced in mainline becomes free, and if Wesnoth Inc ever dissolves then all assets are thereby released freely. Contributors win because they own their work, Wesnoth Inc wins because they can do what they want, spin-off devs win because they have great content to work with for a reasonable royalty, and other FOSS devs get something eventually too. Nobody loses, everybody wins.

[name]

- We want to drive more experimentation and rewarding of new ideas in Wesnoth2. We want to try new game mechanics. New approaches to presenting storyline. New graphical effects. In Wesnoth1 there has been a problem where the amount of prestige associated with a developer is how long they have been with the project (seniority/tenure) and how many commits they have. Newer contributors have struggled with not feeling like they are "core developers" or can effect change. I want to change this with Wesnoth2. I'm not sure we have all the answers on how to change this, but one component is building it using a technology where contributors can try things out without having to rely on a C++ developer to implement things for them, and where things can be modularized better.
- In reference to the previous point we want to have a culture of "If You Implement It, We'll Try It".

This sounds really exciting! Especially the notion of having most or all game logic exposed to or actually implemented in a proper scripting language.

We will have add-ons but with a new language/system for implementing them, which will be more flexible than WML.

As soon as I read you were working on something called "wesnoth2", my first thought was 'no more WML' and a great weight was lifted off my soul.

As someone interested in functional programming but with no actual experience with it, I am curious what considerations lead you to create FFL rather use an off the shelf procedural language like Lua?

Okay. So what licensing model do you think would work best?

Do you think we should have a licensing model where we can distribute our work on Steam/App Store/etc? (even for sale to generate a revenue stream). Or do you think this is a goal we should abandon and rely on direct download to PC's? What goals do you think we should have for our licensing and how should we achieve them?

How about CC0 for both code and content contributions? Would that work for these commercial distributors?

[Pentarctagon]

As soon as I read you were working on something called "wesnoth2", my first thought was 'no more WML' and a great weight was lifted off my soul.

As someone interested in functional programming but with no actual experience with it, I am curious what considerations lead you to create FFL rather use an off the shelf procedural language like Lua?

Personally, I actually kind of like WML. I'm also really glad that Lua was integrated to handle the "heavy lifting" and for the extra things it allows to be done, but despite any of WML's flaws, to me it is significantly more readable as well as being much easier for someone without any programming knowledge to get started in.

[Tad_Carlucci]

Random thoughts:

My only issue is calling it "Wesnoth2" because that sort of cuts off any hope of the current Wesnoth moving from 1.x to 2.x and not creating confusion.

The choice to do WML at all was, however, probably a wise choice, in general, given that there are some issues which it handles which would change a language like Lua into something which is certainly "not Lua" if they were added to the language.

Like all languages, WML is a terrible language in some ways, and terrific in others. I've been thinking of a front-end pre-processor which fixes some of the problems with WML by 'compiling' to classic WML but that's a long-in-the future idea for now.

[name]

I'm also really glad that Lua was integrated to handle the "heavy lifting" and for the extra things it allows to be done, but despite any of WML's flaws, to me it is significantly more readable as well as being much easier for someone without any programming knowledge to get started in.

That's a bit of an apples-to-oranges comparison though, like comparing Lua to JSON. One is for doing just about everything from scratch while the other is focused on storing data in a structured way so that engine code can automagically do things with it for you.

A fairer comparison is WML versus JSON or YAML, for examples. And in that case I would say WML is definitely less readable or beginner friendly, since XML derived languages are too verbose to score highly on human readability and WML over-complicates itself by trying to handle logic itself to an extent.

I say the best of both (all) worlds is to have only the core, performance-hungry engine code written in a compiled language like C++ or Rust, have all the game logic written in Lua or a similarly beginner-friendly interpreted language and have all the game data broken out into JSON or YAML files. Beginners would start by toying with game data (which could actually be enormously powerful in an engine that isn't only written to support vanilla wesnoth). Then as intermediates they would start to write some of their own routines in Lua.

[Pentarctagon]

I'm also really glad that Lua was integrated to handle the "heavy lifting" and for the extra things it allows to be done, but despite any of WML's flaws, to me it is significantly more readable as well as being much easier for someone without any programming knowledge to get started in.

That's a bit of an apples-to-oranges comparison though, like comparing Lua to JSON. One is for doing just about everything from scratch while the other is focused on storing data in a structured way so that engine code can automagically do things with it for you.

A fairer comparison is WML versus JSON or YAML, for examples. And in that case I would say WML is definitely less readable or beginner friendly, since XML derived languages are too verbose to score highly on human readability and WML over-complicates itself by trying to handle logic itself to an extent.

I say the best of both (all) worlds is to have only the core, performance-hungry engine code written in a compiled language like C++ or Rust, have all the game logic written in Lua or a similarly beginner-friendly interpreted language and have all the game data broken out into JSON or YAML files. Beginners would start by toying with game data (which could actually be enormously powerful in an engine that isn't only written to support vanilla wesnoth). Then as intermediates they would start to write some of their own routines in Lua.

The way I usually think of WML, is that it's like basic HTML if someone had tried to fuse the logic into HTML rather than making javascript. Anyway, WML's verbosity is something I always thought was a plus, especially when I was trying to write my first add-on. I can read Lua relatively well now, but even today and especially when I had literally 0 programming experience, this WML is much easier to read than the comparable Lua.

WML:

Code: Select all

[if]
  [variable]
    name=x
    numerical_not_equals=5
  [/variable]
  [then]
    [message]
      speaker="narrator"
      message="Hello World!"
    [/message]
  [/then]
[/if]

Lua:

Code: Select all

if x ~= 5 then
  wesnoth.fire("message", { speaker="narrator", message= "Hello World!" })
end

The WML can be read almost like english: if the variable named x equals the number 5, then display the message.

The Lua statement on the other hand relies on knowledge of what things mean. ~= meaning "not equals" in particular used to throw me through a loop, which it did ironically enough because by the time Wesnoth supported Lua I had started learning a little C++ and Java.

[Tad_Carlucci]

There are a lot of things which could be improved upon in WML without simply going fully over to Lua. Actually, there are things which are easier in WML which would be hard in Lua (such as defining event handlers).

The [if] example, just above, is one of the areas where WML could be improved to be more like a programming language and less like a markup language.

Hypothetical

Code: Select all

    if x ~= 5 then
        message {
            speaker=narrator
            message= _ "Hello world!"
        }
    end

Which is almost-Lua simply because it might be better to borrow as much syntax as possible. (It might even be possible to put enough shims into the Lua side to get this specific example working.)

The issue, though, is how to do events:

WML

Code: Select all

    [event]
        name=side 1 turn 4 end
        .
        .
        .
    [/event]

This is one of the places where WML works far better,

The issue, as I see it, is to get the best of both worlds, where it's easy to define event handlers, but we get a cleaner syntax for more programmatic things like it/then/else blocks with complex conditions. It's not that hard to write a pre-processor which converts a new language into 'classic' WML, which is how I'd suggest proceeding. The idea is to make the simple things simpler, reduce the likelihood of programmer error, and not break current content.

[Celtic_Minstrel]

Hypothetical

Code: Select all

    if x ~= 5 then
        message {
            speaker=narrator
            message= _ "Hello world!"
        }
    end

Which is almost-Lua simply because it might be better to borrow as much syntax as possible. (It might even be possible to put enough shims into the Lua side to get this specific example working.)

This example is actually valid Lua, and will even do exactly what you expect with a few additional lines:

Code: Select all

local _ = wesnoth.textdomain "test"
local message = wesnoth.wml_actions.message
local narrator = 'narrator'
local x = wesnoth.get_variable('x')

[Tad_Carlucci]

Ah, but with Lua we run into synchronization problems. I chose the example because it's so close to Lua it could be made to work. The real issue with Lua is it's far harder to set up an event listener. We could, of course, solve all that with a major re-work of the Lua sub-system .. and to make events easy someone would need to go into the Lua interpreter and add syntax (and do it in a way which merges easily with the upstream Lua modules else we'd be in for a maintenance nightmare).

It's all "doable" the question is the pain.

[iceiceice]

From my point of view, as a programmer, the arrangement you describe seems unattractive in that, I'm spending free time to help create a game / game engine technology, but at the end, basically you own the assets and I have only a limited right to them myself.

Okay. So what licensing model do you think would work best?

Do you think we should have a licensing model where we can distribute our work on Steam/App Store/etc? (even for sale to generate a revenue stream). Or do you think this is a goal we should abandon and rely on direct download to PC's? What goals do you think we should have for our licensing and how should we achieve them?

Has anyone actually ever attempted to make a "cheap commercial knock-off" of Wesnoth? If they call themselves "Battle for Wesnoth" and pose as the original developers, then you can fight it with trademarks I guess. But if they don't call themselves "Battle for Wesnoth" and just make some crappy look-alike, I mean who cares? Is it worth it to spend our time fighting such people, given that this is supposed to be free software anyways?

Yes, one fellow was trying to greenlight Wesnoth on Steam and sell it for himself. The trademark helps against this but is hardly a complete solution for a number of reasons.

And I'm fairly confident that the same would have happened on the app store except that a GPL project can be pretty easily forced off the app store by any contributor to that project.

A different way of arranging things would be, if there were something like a "Wesnoth Art Fund", which was strictly limited to, paying artists to create Wesnothian Art and release it under a CC-BY-SA license or similar. (Maybe both GPL and CC-BY-SA? Whatever.) Even if all Wesnoth activities at large wouldn't qualify as charitable, I think something like that might qualify as charitable anyways, and then people can donate to it. It also has a pretty limited scope and it's easy to understand what it is supposed to do.

Something like that also wouldn't be mutually exclusive with Wesnoth Inc. or other arrangements you want to make, so far as I can see, or even necessarily specific to Wesnoth 1 vs Wesnoth 2.

What would you think about something like that?

So, the major issue here is that for there to be a Wesnoth Art Fund in the first place we would need a revenue stream. The only realistic way I know of to have a chance at a revenue stream large enough for this would be to sell the games on the App Store or other marketplaces.

How would we have a mechanism to sell the game and generate revenue for the fund under this proposal?

Sorry for long delay, I got busy at work.

I guess I don't know what your monetary goals are for Wesnoth 2. But I don't think you should discount donations as far as Wesnoth 1 is concerned. I recall that some time ago there was a discussion on irc about donations, and AI0867 found that someone had started a Wesnoth-related flattr account, which looked like it had about $5000 in it at the time we looked at it.

I found this record by grepping through the logs:

https://www.wesnoth.org/irclogs/2014/05 ... -05-21.log

I did a little more research, I think Rhonda was responsible for this flattr and at some point deleted it. I remember dimly that we told Noy about it at some time and never heard back.

It was never clear to me how that works -- if wesnoth inc. does not accept donations then where does this flattr money go?

Also I don't know exactly over how many years that amount of money accumulated.

But regardless, from point of view of Wesnoth 1, $5000 is a ton of money. Afaik for instance when Lord Bob is commissioned to do a portrait, it's somewhere in the ballpark of $500 - $1000. If you look at how much new art there was in wesnoth 1.12, wesnoth 1.14, etc., it looks to me like $5000 could potentially be enough to do a whole new set of portraits for a campaign like SotBE or something like that. So it's potentially like, an art budget for Wesnoth 1.16.

I looked through the forums, I guess that earlier you accepted donations in like 2008: viewtopic.php?p=316929#p316929

So maybe you are judging how much money you got from that experience?

I think it's different now in that

• Wesnoth is a lot more famous now in 2016 than it was in 2008.
• Many of the people who played it for free years ago, have fond memories and significantly more disposable income nowadays. Many of them would probably be happy to donate if there were a way.
• It's weird that on the one hand, we can have these "Wesnoth is a sinking ship" banners "Call for help, anyone with C++ skills, we need you" etc. etc., but on the other hand we aren't interested in cash donations, despite art commissions grinding to a halt. Before, in this thread viewtopic.php?f=13&t=44185, you said that wesnoth didn't really make (any? as much?) money in recent years, and that was why there were fewer commissions. (I think you edited your post.)
• Even if you don't think it will significantly improve the budget for wesnoth compared to iphone sales, many people actually *like* donating to causes that they support. Unless it is like prohibitively complicated to accept donations, it seems like we should be taking anything we can get. It seems to me at least that you can't really know how much money is turned away by not creating a donation mechanism and advertising it.

The main drawback would be that there would be more financial entities I guess. But with respect, the truth is that it comes off as a bit weird and suspicious that the project like aggressively solicits people to volunteer their time, but can't accept monetary donations due to corporate registration. There were several discussions on wesnoth-dev irc in recent months about why we don't accept donations, and truth is that no one really seemed to understand it. It just boiled down to "ask Dave".

I don't think it's particularly important exactly how it is set up or who controls it, but at least in my mind, if there were a way to donate to the project, and at least the Art Fund part of it is completely transparent about it's operations, that carries some benefits over the current organizational arrangement and over what you proposed in this thread.

Edit: I'll say this also: There are a lot of organizations out there that say they are open source, but have little interest in actually sharing. They are willing to release their code under some CC-0 license if it means they can benefit from the free labor of open source contributors, but there's no meaningful support or documentation, and there's no intention that anyone could factor out large components or fork the project in any meaningful way. Wesnoth is not one of those projects -- it's legitimately a community hobby project, and the open source contributors are the heart of the whole thing. When you propose to make the content proprietary and the organization not to operate transparently, it may send a mixed message to potential contributors. So that may be something worth to consider.

So, the major issue here is that for there to be a Wesnoth Art Fund in the first place we would need a revenue stream. The only realistic way I know of to have a chance at a revenue stream large enough for this would be to sell the games on the App Store or other marketplaces.

How would we have a mechanism to sell the game and generate revenue for the fund under this proposal?

I don't think that proprietary ownership of the content is necessary to sell the game. I think copyleft is compatible with selling the game. You don't need to be the copyright holder to all the art and code to do that. If I didn't understand your point, please explain.

[Iris]

Hi,

I feel Dave hasn’t adequately expressed this, but the idea of assigning content and source code copyright to Wesnoth Inc. in addition to the actual contributors actually originated from me. I never proposed changing the code license in any way or moving away from the ideals of open source, quite the contrary. Open source is what makes this game and this community possible in the first place. Moving towards a proprietary model would not be just a step backwards, it would signify the death of Wesnoth.

So, let me explain my original motivation.

At the time (late 2014), my argument to Dave was that just as it was suggested that we could file a DMCA claim to take down a certain spurious Steam Greenlight submission, any single one of our contributors could file a DMCA claim to get anything we publish on any platform taken down (even if only temporarily) if they did not personally approve of us distributing Wesnoth on some particular channel, regardless of the legality of it. This was primarily based on an incident where GitHub took down a Minecraft-related project after a DMCA claim from one of its founding contributors who did not agree with said project having been acquired (in very loose terms) by Mojang at the time (before Microsoft’s acquisition of Mojang, even), and Mojang refusing to allow the project to be shut down per said founding contributor’s wishes. Said project was technically open source and licensed under the GNU GPL, even if some parts of it had very dubious legal standing to say the least.

The fact that Dave was somehow able to take down the spurious Steam Greenlight submission even though it wasn’t legally in violation of anything at the time (there wasn’t a Wesnoth trademark, and all the content in the submission itself was either licensed under the GNU GPL or counted as fair use) only served to prove my point that our tentative Steam Greenlight submission and the Steam app itself wouldn’t be safe against bogus content claims as long as there wasn’t a clear copyright holder for the whole package, and at the time I thought Wesnoth, Inc. would be the best entity for the job since the Battle for Wesnoth Project (the people who actually develop the game) has always been an informal ad-hoc organization without any legal standing.

Since then I’ve reconsidered my original stance and I don’t feel making Wesnoth, Inc. a definite copyright holder for the Battle for Wesnoth Project’s output is entirely necessary or ideal. Asking past contributors to assign copyright to another person or company is no small task, with most of them having moved on, in a couple of cases due to licensing disputes around Wesnoth’s distribution on Apple’s iOS app store. It certainly cannot and must not be done without their express consent. Those who do not agree with Wesnoth being distributed on certain channels will most certainly not be willing to share or transfer the distribution rights to Wesnoth, Inc., and the process will fail or the project will be forced to replace the affected code or assets with original substitutes, and the people in question might still file bogus content claims. People who would permit Wesnoth’s distribution on those channel may not necessarily want to assign those rights. Ultimately, the point that actually matters is that Wesnoth is not being distributed through channels that people perceive to be legally (not morally) in violation of our licenses of choice. As long as we all agree on that, there should be no need to reassign or share copyrights with a company with an uncertain future.

(Note that this is completely separate from the matter of relicensing assets under a license more suitable for visual and acoustic data files. When and if that is done, the copyright owners should remain the same as ever, and the license must be one compatible with the spirit of open source. But do note that even now it is unclear who owns what, legally or in spirit, and it often requires very careful examination of Git history or asking the right senior developers to obtain this information.)

I have had my share of concerns about Wesnoth, Inc.’s operations over the past few years, not helped by the fact that two of its founding members have apparently left the community until confirmed otherwise, and I wholeheartedly agree that there is a lot of room for improvement with regards to promoting more open and fluid communication between (and within) the Battle for Wesnoth Project and Wesnoth, Inc. However, I feel that it is also too easy to write words that could become detrimental to the game’s future if read by the wrong people at the wrong moment — especially outside people who have only just heard of the game by following a link to a post on the Internet. I think I’m not alone in feeling very strongly about Wesnoth and its community, having dedicated several years of my life to it. If it weren’t for Wesnoth I literally would not be here writing words on a computer screen, so I am thankful to the developers, artists, musicians, translators, and players who have made all of this possible, and also to Dave, Noy, and Turuk for putting up with the tedious task of dealing with the paperwork required to keep the project running and its infrastructure financed. Some mistakes have been made and we can all learn from them, and improve things for future generations of players and contributors to come (even if they tend to be notoriously short-lived). Communication is one of those things that are in most dire need of improvement, and it is my hope that both sides can find a way to make things happen without unnecessary delays and without feeling pressured or coerced into taking immediate action.

Regards