Dominion: Nocturne

[jeebus]

One thing is more or less obscure bugs (of which there seems to be quite a few). Another thing is basic stuff like player tokens not being displayed and the after-game log having wrong information. I don't even feel like starting to list all the other equally important stuff, it's been done before, but I guess I should, but I actually don't have time right now. Anyway, it's very strange to me that people would rather have new cards than basic flaws fixed. You can always play with the new cards by actually buying the physical set and playing.

I think you're in a very small minority on this one.

You're making the same mistake that has been pointed out before. You're confusing the vocal minority in this message board with the totality of the online player base. Most people who post here regularly, so the people you mostly hear from, are experienced players who by now already know all of the quirks of the implementation. For instance the fact that have to have to hover over the user name to see certain cards that are in play or set aside, which crazily is still not fixed, will totally confuse all new players. Even if we talk about all the users still actively playing, those are the people who have not been driven away by a confusing or faulty experience. Most people who have a bad experience, will not post about it here for you to read, they will simply not return.

[007Bistromath]

Ha. This is funny.

For a couple months now, I've been wanting to make a thread along the lines of "it's always bugged me that the only type with only one card in it is Curse." I didn't, because my experience with gaming fora is overwhelmingly that people hate kicking around unrealistic ideas for the fun of it. If the thing wouldn't ever be implemented in a million years, they don't want to talk about it, and always refuse to understand that you're not making a serious suggestion.

Today I log in and see this. I have no idea if there will be Curse cards other than Curse in Nocturne, but that's certainly in keeping with the theme it presents. That's some validation. :P

[Donald X.]

One thing is more or less obscure bugs (of which there seems to be quite a few). Another thing is basic stuff like player tokens not being displayed and the after-game log having wrong information. I don't even feel like starting to list all the other equally important stuff, it's been done before, but I guess I should, but I actually don't have time right now. Anyway, it's very strange to me that people would rather have new cards than basic flaws fixed. You can always play with the new cards by actually buying the physical set and playing.

I think you're in a very small minority on this one.

You're making the same mistake that has been pointed out before. You're confusing the vocal minority in this message board with the totality of the online player base. Most people who post here regularly, so the people you mostly hear from, are experienced players who by now already know all of the quirks of the implementation. For instance the fact that have to have to hover over the user name to see certain cards that are in play or set aside, which crazily is still not fixed, will totally confuse all new players. Even if we talk about all the users still actively playing, those are the people who have not been driven away by a confusing or faulty experience. Most people who have a bad experience, will not post about it here for you to read, they will simply not return.

You don't have enough information to know what jsh considered there; you just disagree with his conclusion.

It is easy to ignore the people who don't say anything, and thus have a less accurate picture of what fans want; that's straightforward. People who don't play Dominion at all are not going to get anything from yet another expansion appearing. People who play Dominion but not online may try the online version because of the new expansion, and you can argue that that's bad, that we want to save those people for when you can tell what Duration cards are out and never have to click on the log and there's an AI and stuff. You can also argue that it's good, that some of them will subscribe, and the success of the game affects how much money is spent on getting work done on it to make it more successful. I don't need to have those arguments, but recognize that they are there for the having.

There is not much data here, but I have seen the experiment play out once, of an expansion being delayed a year so that other stuff could be worked on instead. I was not impressed with how that went.

Again it's moot; however much jsh is ignoring non-posting online Dominion players, the expansion is coming.

[jeebus]

You don't have enough information to know what jsh considered there; you just disagree with his conclusion.

It is easy to ignore the people who don't say anything, and thus have a less accurate picture of what fans want; that's straightforward.

I'm pretty confident that jsh considered people who post. As you say, it's easy to do just that. At the very least, it's pretty certain that jsh did not consider the people who have stopped playing (or people who will try the client in the future and won't stay).

People who play Dominion but not online may try the online version because of the new expansion, and you can argue that that's bad, that we want to save those people for when you can tell what Duration cards are out and never have to click on the log and there's an AI and stuff. You can also argue that it's good, that some of them will subscribe, and the success of the game affects how much money is spent on getting work done on it to make it more successful. I don't need to have those arguments, but recognize that they are there for the having.

I definitely don't think it's bad to get more people to try the online version, now or ever. I'm just talking about allotment of resources: The more time you spend on new cards (and inevitable new bugs), the less time will be spent fixing existing problems.

Again it's moot; however much jsh is ignoring non-posting online Dominion players, the expansion is coming.

Okay.

[Ingix]

A new set is an especially easy task, as has been demonstrated in the past here. Aside from occasional tricky cards like Inheritance, or cards that require unique interface stuff like Stash, it's all trivial. At the same time the new set is a significant draw for not-new-to-Dominion players.

That's the usual 80/20 rule: The small 20% of 'occasional tricky' cards make up 80% of the required implementation time. Unlike in real life, where you can ignore a card like Inheritance, in a digitized version you have to include the full ramifications: That now a card can have changed and additional properties while in your possession. That the only card affected is the Estate doesn't make the programming any easier, because there is no special electronic card board only for Estates, the change must be made for *all* cards.

The same goes for any number of changes that happened over time: Alchemy/Empires introduced a new type of money, which makes cost comparison cards potentially problematic. Knights changed what supply piles could be made off, etc. The real problem aren't cards that does something new in isolation. The big problems are always cards that make something new that interacts with many card effects in a new way that may not have been anticipated when programming those cards.

[Donald X.]

A new set is an especially easy task, as has been demonstrated in the past here. Aside from occasional tricky cards like Inheritance, or cards that require unique interface stuff like Stash, it's all trivial. At the same time the new set is a significant draw for not-new-to-Dominion players.

That's the usual 80/20 rule: The small 20% of 'occasional tricky' cards make up 80% of the required implementation time. Unlike in real life, where you can ignore a card like Inheritance, in a digitized version you have to include the full ramifications: That now a card can have changed and additional properties while in your possession. That the only card affected is the Estate doesn't make the programming any easier, because there is no special electronic card board only for Estates, the change must be made for *all* cards.

The same goes for any number of changes that happened over time: Alchemy/Empires introduced a new type of money, which makes cost comparison cards potentially problematic. Knights changed what supply piles could be made off, etc. The real problem aren't cards that does something new in isolation. The big problems are always cards that make something new that interacts with many card effects in a new way that may not have been anticipated when programming those cards.

I'm not sure what your thesis statement is, if you just want to talk about programming or if you disagree with me strongly on some point.

The existing cards did not take Stef & Co. long to program, not counting the ones that still aren't done. I don't know exact numbers but really, that's a thing that happened and there's real data there and it did not take long. If past performance is indicative of future performance, the new set will not take long.

I do not think 20% of the cards take 80% of the time, and those numbers look suspicious since they add up to 100 but not in a meaningful way. It's more like 5% of the cards take, I dunno, 60% of the time? I wasn't paying that much attention to those numbers when I was programming cards a decade ago.

[Ingix]

See this (https://en.wikipedia.org/wiki/Pareto_principle) for the origin of the 80/20 "rule", and of course it is a generalization that does not always work with those number.

My main point is: It is much easier (though still a very complicated task) to program a complex rule system (as Dominion is), when you already know all the parts, then to start with a subset and then get 'surprised' by new effects when a new expansion comes around.

'Already knowing all the parts' is what happened initially, though I'm not sure if Empires was known to the Developers then. Now comes the 'surprises' part, where you may (or not) have introduced a mechanic that invalidates some assumptions made when writing the existing code base (like new money or non-uniform piles in previous expansions).

Depending on how those assumptions manifest themselves in the code, it may be very easy or really complicated to make the changes. If that new concept was known at initial design time, it may not have been so complicated to include it. But with an existing code base, design changes are hard.

[Donald X.]

See this (https://en.wikipedia.org/wiki/Pareto_principle) for the origin of the 80/20 "rule", and of course it is a generalization that does not always work with those number.

I see, interesting. Do you know this one? I measure something - let's say how tall a mountain is. I use some units I pick - whatever units. I report the number. What's the first digit? 30% of the time, it's 1 (and it's 9 only 5% of the time).

My main point is: It is much easier (though still a very complicated task) to program a complex rule system (as Dominion is), when you already know all the parts, then to start with a subset and then get 'surprised' by new effects when a new expansion comes around.

That's totally true, but I have evidence there too.

I programmed a crude home version of Dominion back when (I stopped updating it when Doug Z. provided isotropic, which was way better). I did not know what all the cards would be ahead of time. Sometimes there were things I had not expected, that required changing stuff (like Fortress) or a novel interface (like Black Market).

A typical expansion nevertheless took a weekend to program, including time spent playtesting. Maybe there would be a bug we didn't notice for a week, but you know.

Doug Z. didn't know what was coming up either, and did not have such a crude program, and he was fast too.

[jeebus]

I have a different view on the triviality of programming new cards. To me there are quite a few existing ones that are not yet fully implemented. As an example, when you can't see the contents of a split pile, I wouldn't call that card pile implemented. The same goes for tokens that are not shown. Then you can start getting into the card bugs, which show that they were not properly implemented.

[Jacob Marley]

I have a different view on the triviality of programming new cards. To me there are quite a few existing ones that are not yet fully implemented. As an example, when you can't see the contents of a split pile, I wouldn't call that card pile implemented. The same goes for tokens that are not shown. Then you can start getting into the card bugs, which show that they were not properly implemented.

Regarding Split piles and tokens, I think you are confusing interface issues with card issues.  To me, a card is fully implemented if it is programmed in and basically works almost all of the time.  I cannot think of any situations in my play where I encountered a card bug, so to me, all cards (except Stash) are basically fully implemented.  The interface problems remain, but that is because we are looking for an improved interface, not because the cards are not programmed correctly.

[jeebus]

Regarding Split piles and tokens, I think you are confusing interface issues with card issues.  To me, a card is fully implemented if it is programmed in and basically works almost all of the time.  I cannot think of any situations in my play where I encountered a card bug, so to me, all cards (except Stash) are basically fully implemented.  The interface problems remain, but that is because we are looking for an improved interface, not because the cards are not programmed correctly.

Let's say Nocturne has a new kind of card that can be placed sideways or not, giving you different effects, and another card performs differently depending on the number of sideways cards. If this is implemented such that the interface still shows no cards sideways, would you say that the cards are fully implemented? What about cards that just show the name of the card, no card text, cost, type or image? Are those fully implemented?

I mean, the interface could just tell you which cards you draw and which gets played and bought/gained, and you could look up all the cards elsewhere and track on paper how many are left in supply etc., but I wouldn't call that fully implemented Dominion.

[Jacob Marley]

Regarding Split piles and tokens, I think you are confusing interface issues with card issues.  To me, a card is fully implemented if it is programmed in and basically works almost all of the time.  I cannot think of any situations in my play where I encountered a card bug, so to me, all cards (except Stash) are basically fully implemented.  The interface problems remain, but that is because we are looking for an improved interface, not because the cards are not programmed correctly.

Let's say Nocturne has a new kind of card that can be placed sideways or not, giving you different effects, and another card performs differently depending on the number of sideways cards. If this is implemented such that the interface still shows no cards sideways, would you say that the cards are fully implemented?

Ok, first the sideways card idea is a manufactured example that is not worth considering for the purposes of what I'm trying to say.  Unless we see this in the set, it doesn's need to be discussed.

What about cards that just show the name of the card, no card text, cost, type or image? Are those fully implemented?

So, according to you, Isotropic was not fully implemented?  Since what you describe is basically Isotropic except that you could click to get more information.

I mean, the interface could just tell you which cards you draw and which gets played and bought/gained, and you could look up all the cards elsewhere and track on paper how many are left in supply etc., but I wouldn't call that fully implemented Dominion.

But you are not describing what actually exists on ShuffleIt, rather some theoretical implementation that no one would actually release for pay.

Mechanically, the cards are implemented in the sense that when you play them, the code correctly applies the effects and updates the game state in memory.  The main problem is that the interface doesn't do a good job of showing the results.  That is an interface issue, not an implementation issue.

So, my point is that as long as the system correctly records the effects of the cards played, they are fully implemented AS CARDS.  If it is hard to look at the screen and figure out what is going on, the problem in an interface problem.  I think the point is that doing the mechanical implementation is easy.  The hard part is the visual aspect, and that would be true regardless of whether the new cards are programmed or not.

[Donald X.]

I have a different view on the triviality of programming new cards. To me there are quite a few existing ones that are not yet fully implemented. As an example, when you can't see the contents of a split pile, I wouldn't call that card pile implemented. The same goes for tokens that are not shown. Then you can start getting into the card bugs, which show that they were not properly implemented.

For me, there's no program until you can play against reasonable bots. By this standard, currently nothing is implemented.

It turns out that's not relevant to this discussion. The people who currently aren't happy will not become happy if Nocturne is added; that's not news. The people who like the program will be happy to get Nocturne though. And the amount of programming needed to get them that happiness is not much programming, whether you personally would call those cards "implemented" or not.

[jeebus]

So, my point is that as long as the system correctly records the effects of the cards played, they are fully implemented AS CARDS.  If it is hard to look at the screen and figure out what is going on, the problem in an interface problem.

That's a completely made up distinction. The fact of the matter is that both the mechanics of the cards and the interface of the cards need to be implemented. To put it another way, Stef is going to add the full card image to all the new cards; he's not keep them blank with just the card name and then focus on undo features until that's complete.

(Btw, Isotropic had the card text when you hovered over the card. Shuffle IT is the same; you right-click to view it. I don't read the card text without right-clicking it. Type and cost was shown in Isotropic without hovering.)

[AdamH]

People have different things they care about for software, I can totally see how DXV would want to mostly play games against bots and how other players wouldn't. I care a lot more about certain animations and displays and other things that would help me keep track of my opponent's turn. Everyone's wants are valid here and the software should do all of those things.

It would have been nice to have them all done by release but that didn't happen, so now the devs have to prioritize them.

Obviously they should prioritize the things I want above everything else. Sure, DXV can have his stuff after all of my stuff is working perfectly.

(in case it is not obvious, I'm not serious about that last part, unless the devs suddenly want to do that I guess)