Commiting to svn & QuestMode discussion

[Braids]

@Hellfish: you posted while i was writing this Library of Leng. i have your back. here is some optimism. of course, i'm also repeating some of what you just said.

Forge will NEVER be a human vs human app.

Why? The code can be improved enoght to allow this kind of gameplay... it's a matter of efforts and developers' skills working on this project.
The wizards - they won't be happy, but who cares? They won't be able to persecute developers. That are players who use their copyrighted content, they are numberless and annonymous, they are legion.

Wizards of the Coast (WotC) can make life uncomfortable and annoying by serving legal documents to the sites that host Forge-related information. on two occasions, they have already done so. the first was resolved by a simple project name change. i do not remember the details of the second, but it was more substantive. the more Forge encroaches on their perceived intellectual property and profits, the more persistent they will be. yes, we can evade them by changing code repositories, going peer-to-peer, etc., but it is just plain annoying.

Rob is one of the "owners" of forge. although we function without an official hierarchy, he has some de facto privilege in saying what is Forge and what is not. because the code is GNU Public Licensed, you have many freedoms. you could make a fork of the entire thing and call it "Max Forge". better yet, you could create a "plug-in" that isn't part of the official forge code, but works with it to provide the functionality you want.

DotP has uncomfortable deck building UI and a limited set of decks, 2012 being even worse than the 2010 edition, it also lacks cards collection, (as well as experience, achievements and so on...) - it is not a competitor niether to Forge, nor MTGO.

all WotC has to argue is that if a single person plays Forge instead of purchasing Duels of the Planeswalkers 2012 or cards in Magic the Gathering Online, they have lost profit on their intellectual property.

What sort of changes would you like to make? And what sort of changes would you be capable of making? Play balance can be a tricky issue. Quest mode is fun to play and we all enjoy it ... granted, improvemnts to the quest mode can be a good thing

QuestMode first of all - better collection management, more reasonable rewards (ability to choose the prize booster edition or block at least) involving both UI and logics, while sticking to the MVC pattern to keep code like quest.addRares(10) away from Gui_WinLose classes. I would also love to make cards from different editions distinguishable, be stored as different cards, to make UNH-Plains a reward more rare than common lands from core sets.
I don't like booster generation rules and fantasy-quest-store contents being hardcoded for now - would like to store these settings in XMLs... that is while I've looked through quite a little part of code.

i will now share with you Braids's anarchaic development postulate. As a proposed change to the product approaches the end user's or another developer's experience, the more consensus you will need to garner among the developers for that change.

i would suggest opening topics on this forum, especially for changing the quest rewards and distinguishing cards by set. if you're going to make something less hard coded, we will need documentation for the XML for it to be mentally digestible.

you mentioned MVC. in the case of Forge, how would you separate the model from the controller? usually the model is a database or file format, right? i've been operating under the assumption that separating the model from the controller breaks object oriented encapsulation, so i have been keeping them together.

Later switch to duel UI, help separating logics from UI, develop a protocol so that two forge clients can communicate to each other directly or via a trusted "judge" forge-server.

implementing a minimax AI could help you indirectly with this. minimax needs to treat players more equally, without regard to species. in other words, it needs to be able to handle the human player and the computer player interchangeably. each could implement a more generic player interface. it is possible that someone later could implement some sort of remote player class that implements that interface, without having to change too much.

That is a big plan and I doubt if I'll be able to complete everything in practice, however these are my aspirations.

aspirations are good. but sadly, we have some technical debt. our large CheckStyle, FindBugs, and PMD outputs are clear evidence of this. Forge is a large system with some design missteps {and way too much static cling}. making small changes can affect this system in unexpected ways.

i would implore you not to lose sight of your dreams, and also not to take Forge and try to create a fork right away. if you work with us, i think Forge as a whole will be better off!

[Rob Cashwalker]

I've been involved with this project since before Rares opened the source code. I do not "own" anything other than the cardforge.org domain, which could be the next target....

That said, someone needs to be a "leader"....

Everytime someone brings up human vs human, we answer "no". Technical challenges are the best excuse, though not impossible, there are AI assumptions in every corner of the game. Who's going to run the judge server?

Seriously, almost every other magic project is designed for human vs human, our niche is AI and the quest mode thing.

[Max mtg]

They can, and in fact must do so periodically, in order to protect their trademark.You may not have noticed the recent move from google code svn->gitorious GIT->Slightlymagic svn. That was because of cease & desist letters from WOTC, the second time they have been sent, I believe.

I have noticed that. Google is not going to protect the project - they are not interested. It's easier for them to ban an account instead of checking if the files mentioned in letter are really voilating something. Slightlymagic.net is a much better hosting - I heard site admin lives in Russia and doesn't need to care about wizards' unhappiness.

Because when Forge starts to noticeably impact WOTC's profits (and even before that, see my first quote reply) they are within their full legal right to bring the proverbial hammer down, HARD.

Welp, I started writing this post with the intention of realism rather than pessimism.. Sorry about that.

1. Their "full legal rights" have no power in foreign countries.
2. An impact on profit cannot be proven or measured. How can they prove that all those people who downloaded and played Forge-Online for free, would otherwise - that is if forge online never existed - bring their money exactly to Wotc and measure that amount? Why at all are we speaking of any impact? Competitors don't accuse each other of lost profits, do they?
3. Which laws is Forge supposed to break? Patent 5,662,332? Valid only in US, host it abroad and the problem is gone. Call cards 'flat rectangular tokens' and this is no longer a card game. Copyrighted images of cards? Well, maybe... but users download card images by themselves and it is their personal responsibility. Anything else? Or is everyone just frightened by the almighty WotC?

[Rob Cashwalker]

The bulk of our copyright laws apply in most countries. If Hasbro has an office in a country, they can extend their legal impact. Notice Apple's lawsuit with Samsung is based in Germany, not the US. Gitorious.org wasn't US based, but their TOS included language for the DMCA, which is US-based legislation. About the only place we could truly operate risk-free would be China.... Next best option seems to be mother Russia.

I happen to like the game that they work their asses off to design... the cardboard version anyway. Ensuring that they can continue to do so by not promoting our project as a Magic Online replacement is a good start. If Wizards/Hasbro killed Magic then Forge would be a dead project anyway.

Competition to Magic are games like Pokemon and Yugioh, not Forge. The only thing somewhat unique to Forge is the quest mode. Otherwise, the only use for Forge is to play Magic. The original reason Rares wrote it was to play with power cards because he wouldn't buy them IRL, part because of money, mostly because of a lack of folks to play with in his remote location. Wizards hasn't made a dime directly from the power cards since they were printed 16 years ago. But M12 and Scars block are still being sold through Wizard's channels and Forge can draft from those sets instead of you going to your local game store to pick up 3 packs per player. So not only is Wizards losing money, the brick and mortar store loses money, hell, Magic Online itself has put some shops out of business.

We have various ideas planned to further remove various card information from the core of the game, into external files, which could be downloaded separate from the game itself. Not so much as legal appeasement, but because it would make the initial download package smaller and external files are easily updated by non-coders.

[Max mtg]

Wizards of the Coast (WotC) can make life uncomfortable and annoying by serving legal documents to the sites that host Forge-related information.

Use sites that are outside of US juristiction. I guess it's already done.

all WotC has to argue is that if a single person plays Forge instead of purchasing Duels of the Planeswalkers 2012 or cards in Magic the Gathering Online, they have lost profit on their intellectual property.

I don't get it. Maybe they would never get any profit from that particular person? Or what if that person _has_ bought DotP'12 or _has_ played several games in MTGO, but didn't like them and switched to Forge?

i would suggest opening topics on this forum, especially for changing the quest rewards and distinguishing cards by set. if you're going to make something less hard coded, we will need documentation for the XML for it to be mentally digestible.

Sure thing.

you mentioned MVC. in the case of Forge, how would you separate the model from the controller? usually the model is a database or file format, right? i've been operating under the assumption that separating the model from the controller breaks object oriented encapsulation, so i have been keeping them together.

Yes, I mean to store all that data that describes the game/quest/decks state separately from the code that manipulates it and away from GUI - just in different classes. For instance QuestData has method addRandomRare() - that's bad (model and ctrl are metled together). There should be a separate class for CardCollection, with methods for adding a definite card or a set of them, even if it is just a warper around List<String>. It is controller-class who should determine which card is that random rare and update both model (to ensure it is added to cardpool now) and the view (pop up a messagebox or show it in a list).

You are right, encapsulation as information hiding mechanism is gone once the data (model) is stored in one class and methods to process it are located in a different one (which I refer to as controller). On the other side it helps avoid having 5000-lines-of-code monster classes.

in other words, it needs to be able to handle the human player and the computer player interchangeably. each could implement a more generic player interface.

That would be very nice, so that players can duel, as well it might be like a unit test for AI - to pair two computers and be sure that a stronger deck wins.

we have some technical debt. our large CheckStyle, FindBugs, and PMD outputs are clear evidence of this.

I see. The bare refactoring (meant to improve overall code quality) is not a very interesting and rewarding task. So that I would make some changes and give the related files a better look on the go.

[Max mtg]

The bulk of our copyright laws apply in most countries. If Hasbro has an office in a country, they can extend their legal impact. Notice Apple's lawsuit with Samsung is based in Germany, not the US. Gitorious.org wasn't US based, but their TOS included language for the DMCA, which is US-based legislation. About the only place we could truly operate risk-free would be China.... Next best option seems to be mother Russia.

Gitourious has no reference to DMCA, but they agree to remove stolen copyrighted work. It's common practice.
As for me Russia seems way better than China

I happen to like the game that they work their asses off to design... the cardboard version anyway. Ensuring that they can continue to do so by not promoting our project as a Magic Online replacement is a good start. If Wizards/Hasbro killed Magic then Forge would be a dead project anyway.

They won't promote forge or kill Magic. For this to happen Forge has to become much better than Mtgo... how long will it take?
Why dead? Do you think people will forget magic if no more sets are printed? It will at all times have a group of fans, like old times games have now, and simulation programs, such as Forge will be appreciated.

Competition to Magic are games like Pokemon and Yugioh, not Forge. The only thing somewhat unique to Forge is the quest mode.

We are competing for human's needs satisfaction, though not in quality but in avaliability of product. Those who cannot or don't want to pay, will play Forge, those who can, will keep playing mtgo.
Quest mode is unique. It's was a great idea to make an rpg in mtg setting. Though people beg for Shandalaar clone, the quest mode Forge currently has is at least better than nothing.

Otherwise, the only use for Forge is to play Magic. The original reason Rares wrote it was to play with power cards because he wouldn't buy them IRL, part because of money, mostly because of a lack of folks to play with in his remote location.

Sure, it also helps one to get familiar with cards and perform better at tournaments or playing with friends.

Wizards hasn't made a dime directly from the power cards since they were printed 16 years ago. But M12 and Scars block are still being sold through Wizard's channels and Forge can draft from those sets instead of you going to your local game store to pick up 3 packs per player. So not only is Wizards losing money, the brick and mortar store loses money, hell, Magic Online itself has put some shops out of business.

When playing with friends offline or at a club, a choice for solid cards is out of question, so the retailes still have a chance to sell their SOM12 packs.

We have various ideas planned to further remove various card information from the core of the game, into external files, which could be downloaded separate from the game itself. Not so much as legal appeasement, but because it would make the initial download package smaller and external files are easily updated by non-coders.

That would be great.

[Braids]

I've been involved with this project since before Rares opened the source code. I do not "own" anything other than the cardforge.org domain, which could be the next target.... That said, someone needs to be a "leader"....

which is why i referred to you as a de facto owner. i do need to correct myself, though. technically, i believe mtgrares {Rares} is the original licensor of the code, so Rares is the one with actual say as to what is and is not Forge. but this person may as well be Rob, who seems to know Rares better than a lot of us.

Seriously, almost every other magic project is designed for human vs human, our niche is AI and the quest mode thing.

you are right. besides, the rules-enforcing games will always be many steps behind the virtual tabletops. Online Play Table for example supports just about every card except Falling Star and Chaos Orb, for Circu's sake.

We have various ideas planned to further remove various card information from the core of the game, into external files, which could be downloaded separate from the game itself. Not so much as legal appeasement, but because it would make the initial download package smaller and external files are easily updated by non-coders.

it's droll how Rob and i agree on the means but not the reason here. i suppose i'm more wary of lawyers.

. . . Slightlymagic.net is a much better hosting - I heard site admin lives in Russia and doesn't need to care about wizards' unhappiness.

i think you are right about Goblin Hero, the site admin. but last i checked, the actual slightlymagic.net site is in the Bavarian province {Bayern} of Germany.

Or is everyone just frightened by the almighty WotC?

i would describe our position as cautious and proactive. not fearful.

you mentioned MVC. in the case of Forge, how would you separate the model from the controller? . . . i've been operating under the assumption that separating the model from the controller breaks object oriented encapsulation, so i have been keeping them together.

Yes, I mean to store all that data that describes the game/quest/decks state separately from the code that manipulates it and away from GUI - just in different classes. For instance QuestData has method addRandomRare() - that's bad (model and ctrl are metled together). There should be a separate class for CardCollection, with methods for adding a definite card or a set of them, even if it is just a warper around List<String>. It is controller-class who should determine which card is that random rare and update both model (to ensure it is added to cardpool now) and the view (pop up a messagebox or show it in a list).

i'm not ready to say whether i think that's a bad idea, but i can say that i don't think the system is robust enough to handle the model-controller separation yet. i would prefer we pry the view apart from the model and controller, bundling the last two as the "model" for now. then decide where to go from there. i am talking about refactoring old code. if you want to introduce a controller separation in new code, i think i can stomach that. but please write clear javadoc.

You are right, encapsulation as information hiding mechanism is gone once the data (model) is stored in one class and methods to process it are located in a different one (which I refer to as controller). On the other side it helps avoid having 5000-lines-of-code monster classes.

well, there are ways of avoiding large classes classes with high cyclomatic complexities without a model-controller separation. i did this with the FModel class by using object oriented composition. Every FModel has a BuildInfo instance.

in other words, it needs to be able to handle the human player and the computer player interchangeably. each could implement a more generic player interface.

That would be very nice, so that players can duel, as well it might be like a unit test for AI - to pair two computers and be sure that a stronger deck wins.

AI versus AI for unit test sounds very nice.

we have some technical debt. . .

I see. The bare refactoring (meant to improve overall code quality) is not a very interesting and rewarding task. So that I would make some changes and give the related files a better look on the go.

are you trying to say you'll refactor classes directly related to the features and/or fixes you are implementing? that makes a lot of sense, since you'll be testing those classes more rigorously as well. i hope you won't run into any of Forge's fragility in the process. if i misunderstood you, could you clarify?