Ongoing Uru Review: Handling Player-Created Ages, Again

I just sent this message off to Cyan as a (bulky) feature request.

It is, to some extent, a recap of my forum post of July 2006. But then, I never filed that as a formal feature request. And this new post is informed by the past 18 months of Uru development, and the parallel growth of the Uru fan community. I figured it was time.

Note that "GoW" is the nascent Guild of Writers; "GoMa" is the Guild of Maintainers. ("GoMa" to distinguish it from "GoMe", the Guild of Messengers.)


It's the season break, and the status of player-created Ages is still "Cyan is thinking about how to handle it." Obviously, the players are thinking about it too, and there's plenty of disagreement among us.

I want to offer my thoughts. I do not speak for anybody but myself. (But I have posted much on the MOUL "Guilds" forum, and then on forum.guildofwriters.com, to argue for these ideas.) I started writing this post when I read the Oct. 31 interview on Gametap, which included this quote:

"We've been kicking around a lot of plans for season two, like ramping up the guilds and player created content," says Ryan [Miller]. "It won't be like say, Second Life, where players can just create anything. There will be some quality assurance involved."
Quality assurance is certainly required, and Second Life is an obvious bogeyman. (My capsule description of Second Life includes the phrase "storms of flying genitalia". I will take for granted that you don't want Uru to be like that. I don't either.)

However, saying "quality assurance" doesn't tell us who is assuring quality, or who defines quality.

I say: (1) QA of player Ages must be an open and transparent process, handled by players, not by Cyan. (2) It must be a process of finding and promoting the good, not locking out the bad. (3) Both of those statements must apply at every stage: story, design, modelling, bug-hunting. (4) Player Ages must not be judged by the standards of Cyan Ages.

In those four assertions, I've ruled out a whole pile of "obvious" plans which some players -- and maybe you folks -- have taken for granted. By way of supporting my claims, let me go through them and explain why they don't work.

(A): "Cyan acts as the QA team for player Ages." Doesn't work. Cyan isn't big enough. I'm assuming that Cyan's testing resources are redlined by your own Age production -- Season One made that pretty clear.

If player Ages are a success, there will be more of them than one or two per month. A lot more. Mostly tiny Ages, perhaps -- but the small ones still take a lot of testing work. (Jalak did, right?) If you plan to handle them in your spare time, you're planning for failure.

(B): "Cyan acts as the gatekeeper, deciding when an Age is good enough to be released." Doesn't work. You might think a go/no-go decision is quick, but you're wrong -- everything I just said about testing still applies.

Am I crazy to think you'll get so many Age submissions that you can't even thumbs-up them? Maybe; but players do crazy things. In particular, players will create small, simple Ages just to prove they can. And that's good. You want those experiments to exist in Uru, accessible to all. That's how you prove that the system is working.

Look: from the player point of view, Cyan is a black box which usually takes months to respond to any given request. (Sometimes it's much faster, sure; but equally often, requests and bug reports and etc just vanish into the murk forever.) If you try to take a gatekeeper role, maybe 80% of potential Ages (and Writers) will just never come out from under their rocks. Releasing an Age to other players is encouraging: players will come back with praise, suggestions, help, cheerleading. Submitting an Age to Cyan is scary: the best that you can hope for is a plain "yes", and we're not trained to expect that in a hurry.

Therefore, Ages must reach other players without going through Cyan. It's that simple.

(It's the exchange of Ages (including half-written and half-working ones) which will lead to a flood of successful Ages. This is such an important point that I should write a whole post devoted to it. Instead, I'm jamming it in between points (B) and (C). Please be generous to it. If you haven't got a penny then a ha'penny will do.)

(C): "A select, Cyan-blessed group of players can act as the testers and gatekeepers." Doesn't work. How would such a group act? For their blessedness to have any meaning, they'd have to operate in secrecy -- as a black box. And that makes them just as bad as Cyan. Pardon my French.

The process must not only work, it must be seen to work. If the player Age creation system takes months to produce its first result, it might as well not exist, for those months. Whereas if the process is transparent, the excitement starts on day one.

Besides, we've already got an Age creation pipeline that operates in secret and takes months to respond to player feedback. No offense. We don't need to clone that model for player Ages. It only dissipates our best advantage: speed and enthusiasm. (And it smashes us up against your best advantage, which is experience and professionalism. Cyan Ages and player Ages will have different standards... I will return to this point.)

(D): "The Guild of Maintainers can act as the testers and gatekeepers." Doesn't work... or rather, it does work, if you handle the Guild system properly. Which is, I say, openly -- with no attempt to define privileged access or privileged subgroups.

If the GoMa is a privileged group, to which only selected players are admitted, you're right back in case (C). You create tremendous pressure to get in; people want to see all the cool upcoming Ages. (Even the bad ones!) Membership becomes a reward, and a highly sought-after reward. And... rules condense around what people consider important. Your Guild of Maintainers will turn into a Guild of Keeping the Wrong People Out Of Our Precious Ages. With all the bad feeling that implies.

(All this logic applies to the Guild of Writers, too. If the GoW has special privileges to look at each others' in-progress Ages, then you create tremendous pressure to get in... blah, blah, bad feeling, train wreck.)

(E): "Experiments and sample Ages are fine between writers, but they shouldn't be part of Uru." Wrong. Did I mention the principle of exchanging ideas? You know what will arouse more excitement than anything else towards player-created Ages? The first time Joe Schmoe, logged into Uru Live, selects "Fred Schmedt's First Age!!" and links to it. Even if that first Age is a square stone platform hanging in grey mist. Heck, especially then. I can do better than this!

And Joe will.

(I fully intend to be Fred Schmedt, there, with my square stone platform. I'll have the Age model posted on the GoW forum, too. With as much documentation as I can muster about how to download it, add stuff to it, and then upload it into Uru Live.)

Player Ages are not Cyan Ages. You have a quality standard. That's fine. We will have lots of standards, because players never agree on anything. Some of us will be really interested in puzzles. Others, in visual technique. Others, in sound environments.

Some people (many people, I'm sure) will release single-aspect Ages. This Age exists to show off a single lamp or chair model; this one, to show off a few tracks of music; this one, a particularly nice flowing-water effect; this one, my poetry in a set of journals. All those Ages should be part of Uru.

(See why you can't possibly keep up as gatekeepers?) (See why no single gatekeeper group can be allowed to wield the magic "release this" wand?)

(F): "To prevent storms of genitalia, somebody must act as gatekeeper." This is wrong. Honest.

Of course you have to keep an eye on player Ages, to prevent decency and legal violations. You have to keep an eye on forum posts, KI messages, and in-game chatter, for exactly the same reasons. Do you have gatekeepers on any of those? Of course not. It would have a fatal chilling, paralyzing effect -- for all the reasons I've described.

You moderate, that's all. Same as you do on the forums and the KI-mail lines. If someone complains, you take a look, and you smack down violators. Yank the Age; yank the player account if necessary; problem solved.

You can have moderators who are player volunteers. (Just like on the forums.) This is a form of player privilege that doesn't lead to a train wreck -- because it's open and transparent. Anyone in the community who sees the wiener-storm Age knows it's against the community standards; they know a moderator will yank the Age. It won't happen inside a black box, so people will be satisfied with the result.

(And players won't come screaming to be moderators, because moderators have no special access to cool stuff. It's just a job -- work, responsibility. Sure, there's power, but most players are mature enough not to want the wand of smack-down for the pure sake of having it. And the forum system has proven that you can screen out the immature ones.)


Okay, enough of the "doesn't work". What do I think will work?

When I decide to release an Age, it becomes accessible to all players.

I've said plenty about privileged guardians and release gatekeepers. I'll repeat it if you poke me with a noodle. Twice for fettucine.

But every player must be able to choose his level of Age quality comfort.

Is a buggy, half-finished Age the same as a polished one? Of course not. This is the true role of the Maintainers: tagging. Looking at an Age and saying "Yes, everyone will want to see this!" Or, "This is a nice demo of alpha textures. Age modellers will want to look at this." Or, "This is full of bugs. Make sure people who link in are forewarned."

Yes, players will want to see buggy Ages. Some players, as I said, will want to see everything. You let them. Then you let them talk about the bugs. Then the original creator talks with them about the bugs.

Does that sound familiar? I've just described the Guild of Maintainers! If you're voluntarily entering a buggy Age, you're doing Maintainer work. Letting players into buggy Ages is the same as letting players join the GoMa. And, if the player community has agreed on anything, it's that the Guilds must be inclusive and open to everyone who wants to join.

(Again, the same logic applies to the Guild of Writers. Letting players enter unfinished Ages is the same as letting them join the GoW. The Guild is defined as everyone involved in the creation process. And part of the creation process is peer discussion.) (Part of the creation process is tagging, too. Authors tag their own work, before the Maintainers do -- that is to say, before other players arrive and judge. Everyone is a Maintainer, just as everyone is a Writer. Get used to this.)

See all, recommend the good stuff.

This is essentially the same as what I just wrote. In a volunteer economy, you don't create quality by blocking out the bad stuff. That only inhibits the volunteers. You encourage creation, and trust that some of it will be good. Then you encourage the good stuff to get better.

(And encourage the bad stuff to get better, too.)

Personal recommendations (and labels) are as important as "official" Maintainer ones.

Labelling is good, but letting one group monopolize the labels is bad. (Again, you're generating privilege, which generates pressure and tension.) I should be able to put together my own list of "great Ages" -- or "Ages with great sound effects", or "Ages for my writing club to discuss", or "you want to see my Ages huh huh do ya?" I should be able to share those lists with my friends. If my lists prove really popular, they should be able to spread through the community; the more popular they get, the easier they should be to find. Maybe they'll even come to outshine the GoMa lists.

There are a lot of ways to implement this, and (unfortunately) they all involve upgrades to the KI system. (This is a reflection of the general sad state of the KI system. Messages, marker games, etc should also be able to spread through the community -- and right now, they can't. Pretend I have a detailed and brilliant proof of this assertion.)

If you want a starting point, start with the Neighborhoods. They're the most powerful "reputation system" in Uru today. Popular neighborhoods rise on the list, and thus become more popular. But I can invite anyone to my neighborhood, which means that I can still share stuff with my friends even if I'm not popular. (Both of these properties are critical.)

Imagine that I could post one or more lists of Age links in my neighborhood. That accomplishes almost everything I want. The GoMa would have lists (and maybe separate 'hoods) for "bug testing", "samples", "demos", "puzzle ages", "story ages"... and those would be popular neighborhoods, but I'd have my own lists in my own 'hood. And maybe I'd attract attention too.

(I said "almost everything". What's missing? The ability to look at a posted link, and copy it into my own 'hood list. Age blogging! Hint: we need this ability for messages, KI photos, and marker missions too.)

Don't let the vault database be corrupted. Don't let avatars get wrecked.

I've left this for last because it's such a sticking point.

I realize that anything I've said in this post could run afoul of your database consistency requirements. I don't know your code. I don't know how much damage a buggy Age, or a malicious Age, could do.

But everything I've said remains true. If you have to restrict untested Ages to a separate server, then all players should have access to that server. Maybe you'll need a Cyan QA team to "promote" Ages to the "real" Live server -- but if player Ages are a success, I guarantee that the gatekeepers will be backlogged, and the life of the community will be on the testing server. Yes, even if it's wiped weekly. Players go where the action is; Until Uru proved that.

Good luck.

Out of words. I hope this makes sense. I feel passionately about it, or I wouldn't have written all this.


Last updated November 13, 2007.

More of the Ongoing Uru Review

Other Uru Stuff

Zarfhome