I'm beginning to envy operations like Reflexive or RealArcade, where everything works the same. Everything is a 60 minute unlockable demo, everything is under 20 megs, and everything is a $20 price point. (Actually Reflexive has some at $15 and some at $25.) And everything is either casual or arcade-y.

Part of the reason we've taken so long is the sheer variety of the stuff we'll be offering. Here are some examples:

Installer file sizes: From 1.1 MB to 3 GB (!).

Demo/full game types:
1. Unlockable demo (that is, enter an unlock code and the demo becomes unlocked into the full game).
a. Some are limited by time--typically 60 minutes, but sometimes as much as 5 hours.
b. Some are limited to a certain number of levels.
c. Most use our unlock system.
d. Some use the developer's unlock system
i. and they can provide us a list of unlock codes to get into our database, so it works much like our existing scheme, but we need to go back to them to get more codes if/when we start to run out
ii. and some generate unlock codes on the basis of a userID, or first and last name, or first and last name plus email address, some with a remote server we can query for an unlock code, some using an algorithm we have to implement on our servers to generate them.

2. Demo and full game are separate downloads, with only the full game requiring an unlock code; both can be hosted on Amazon S3, because we don't mind if random people download them, since they'll still need to purchase the full game from us if they want to use it.

3. Demo and full game are separate downloads, but the full game is unprotected, so it needs to be hosted on our own secure servers so it can't be downloaded without purchase.

4. The game provider prefers to host installers themselves, and we link to them as an affiliate.

5. The game is playable in the browser and there's no installer, but you have to subscribe to play or play is free but you can purchase a premium membership or items, and we link to them with some sort of affiliate code.

6. The client software is a free download, but connects to their server for MMO play with some sort of affiliate code.

7. The full game is so huge that providing any demo would be financially disastrous, so there is none, but we offer a money-back guarantee instead.

Price points: $0 to $40. $20 is typical, but bizarrely, the main driver of our costs for most games is going to be demo size. For instance, if the demo is 500 megs (as it is in a few cases), and we assume a 2% conversion rate, even at Amazon S3 prices, we wind up paying $5 is bandwidth fees per sale. Yet the larger games are mostly from people who publish mainly at retail, and are not used to any relationship between file size and cost... So we have 5MB games that cost $20, and 200MB games that cost $10... Not, mind you, that file size has any relationship to actual merit or level of fun, either. But trying to ensure that we have a reasonable margin on low price/high bandwidth games is causing me to lose some hair.

Platforms: PC, Mac, and Linux...

Game type: All over the place. Adventure games, RPGs (a surprisingly large number), sim/tycoon, shmups, RTS, third-person shooters, sports management, turn-based strategy....

Business model: Revenue share of application sale. Revenue share of application sale via an affiliate relationship with a third party. Bounty payment on MMO subscription. Lifetime share of item sales. Share of subscription revenue for first year of membership. In some cases, no business model--we link to some cool free games as a means of generating good will...

It's chaos. Managed chaos, hopefully. I don't even want to think about when I have to start writing checks to our developers... How exactly do you transfer money into Byelorussia cost-effectively? (Wire transfer, I suspect.)

I signed the contract for our hosting partner for the final deployment site today, btw, so in a week or so we'll move the development environment over there for testing in the actual deployment environment.
posted by Greg at 1:11 PM 8 comments

Damn.

I first encountered Jim Baen as a teenager, when I was a subscriber to the late, lamented Galaxy magazine, of which he was the last editor. After Galaxy disintegrated (in a clusterfuck typical of collapsing companies, with subscribers never getting paid for unpublished issues--not Baen's fault), he went to work for Tom Doherty at Ace, and left with Tom when he went off to found Tor. In 83, Pocket (foolishly) canned their SF line, and Jim left Tor to found Baen Books, an independent house but distributed by Pocket, a company he ran until the day he died.

I met Baen personally in 83, when he was in the process of launching a game publisher (with help from Jerry Pournelle, sf writer, famous computer pundit, and not incidentally an old-school board wargamer). IIRC correctly, Baen published only one game (based on Fred Saberhagen's Berserker series), but curiously, Moby Games makes no mention of it (Apple II title, I believe). At the time, I was working and living out of a roach-infested hovel on Amsterdam Avenue, doing a series of (non-digital) games for Avalon Hill, Victory, Steve Jackson Games, and West End, but had acquired an Apple II and knew enough about programming to think it would be cool to do a computer game. I pitched Baen on a game based on Poul Anderson's Flandry series, and he contacted Poul who had, however, given the computer game rights to the series to a friend. Jim commissioned a new novel from Poul called Game of Empire, set in the same universe with with Flandry's daughter as protagonist, and commissioned me to design a game based on the new work.

I still have the source code somewhere (8502 assembler, if you care), but to my regret never completed the thing, and shortly thereafter, in the wake of the videogame crash, Baen decided to pull the plug on his game publishing operations. C'est la guerre.

I don't have a lot of recent Baen books on my shelves; in later years, the company seemed to specialize in what I've elsewhere called "war bores," that is, military skiffy, something I read enthusiastically as an adolescent but have little patience for now. But though our tastes diverged, and despite his somewhat astringent personality, which made him some enemies in the field, I always thought of Jim as basically one of the good guys.

Ah, well.
posted by Greg at 5:04 PM 12 comments

Everything always takes longer than you'd like, confound it.

But we're getting there.

Critical path issues at the moment:

1. Implementing the Drupal theming (meaning transforming Karen's graphic design into the look for the site): under way, but at least a week to work out the kinks.

2. Moving from the development site to our ultimate hosting service: Calls with our two finalists tomorrow, after which we make a decision. Probably 2-3 weeks between prying a contract out of their legal department, getting everything moved over, and working out whatever kinks arise as a result.

3. A few issues with the end-to-end transaction system:
a. We handle okay the following situations: demo that unlocks into the full game; separate demo from full game installer, but the full game installer is not protected by DRM (which is the preference of some of our partners); no demo, only a full game installer for purchase (the case with only one game, which is so large that providing a demo is infeasible); situations where we act as an affiliate for someone else, and the demo and purchase links are offsite. Unfortunately, we still need code to handle two other situations:
b. Demo is a separate installation, but on purchase we need to deliver an unlock code as well as a full game installer. (Shouldn't be hard to do, but isn't implemented yet.)
c. We integrate with the developer's unlock solution, but their solution isn't designed so that they can simply generate a bunch of unlock codes and hand them to us. Typically, unlock codes are generated at runtime on the basis of userID and/or some other data. This is going to require separate custom coding for each such developer--and unfortunately, we may launch without it, and add it as feasible.

4. The last issue is prying files for installation out of all our partners, injecting the unlock solution when necessary, building installers, and getting everything up on the site. Currently 38 titles out of a projected 100 are ready to go. This is what I'm spending most of my time doing at the moment, but the bottleneck is less my time than getting all of our partners to send in their files. Just FYI, for those who commented previously, we're using Inno Setup rather than InstallShield, as it works well and produces smaller and quicker installers; and files are hosted on Amazon S3 for minimal bandwidth charges (except for unprotected installers, which are hosted locally with non-accessible URLs for security reasons).

Oh, yeah... Minor disaster of the week is that Aladdin has apparently stopped supporting OS X, so we can't use our unlock solution for Mac applications. As it happens, this affects only one game (so far) as the other partners who have Mac versions all want unprotected installers or provide their own unlock codes. And with that one partner, we have a work-around--essentially, we become an affiliate of another site that sells the Mac version, so we can still offer it. But a pain nonetheless.

Currently, our target is to launch simultaneously with the Texas Independent Games Conference. But it's certainly possible we won't hit that target.

Lot of balls in the air... but at least we're making progress, goddamn it.
posted by Greg at 6:08 PM 5 comments

"Electronic Arts Inc. (NASDAQ: ERTS) today announced
that it has entered into an agreement to acquire Virginia-based Mythic Entertainment®. Upon completion of the acquisition, Mythic Entertainment will become EA Mythic, a wholly-owned studio dedicated to developing Massively Multiplayer Online Role-Playing Games (MMORPGs). Mythic is recognized worldwide for revolutionizing the online gaming space with the awardwinning Dark Age of Camelot® and is currently developing Warhammer® Online: Age of Reckoning™ under license agreement with Games Workshop."

Rumored for some time, and now official.

Let's parse this a little.

Mark and crew get to cash out--a multidecade "instant" success, and good for them.

EA. Ah, yes. EA, the far-sighted MMO pioneer that had the field's first success with Ultima Online, and shortly after gutted Origin. That bought Kesmai, a company with longer experience in online games than any other, and shortly afterward, closed them down. The company that spent something like $150m on "EA.com," and failed to show anything for the investment, other than a handful of failed MMOs and the purchase of the Pogo.com audience. The company that a few years ago was claiming that subscription-based games were the wave of the future and would vastly increase revenues for the game industry--only their idea of a subscription-based game was Majestic. Ho ho. A company that, in other words, has corporately never understood online games and today displays no greater knowledge of it.

From this you can come to one of two conclusions: Either they desperately need to acquire a company like Mythic and tap its expertise and make Mark Jacobs leader of a new foray into MMOs (since he for damn sure does understand the field). Or else Mythic is doomed, doomed, I tell you.

With the acquisition of Jamdat, EA seems to have gone the first route; Mitch Lasky now runs EA Mobile, which is a good thing for EA, since he actually understands the mobile space, and no on else at EA really does. (Other than other Jamdat folks, of course.) But Mitch has the advantage of being in California, so he can more easily rub shoulders in Redwood Shores and keep on top of politics at the mothership--and is also ex-Activision, so has considerable experience with big game company politics.

Mark, by contrast, is on the East Coast, and has basically run his own operation forever. The snakepit that is EA corporate will doubtless be a novelty to him.

My guess is: Mythic gets left alone at least until they ship Warhammer. Shortly thereafter, they are gutted, with responsbility for the Mythic games transferred to California.

Welcome to the new boss, worse than the old boss.
posted by Greg at 5:45 PM 7 comments

From an interview with Henry Jenkins on GameSetWatch.

"It is also a fight to help game designers gain greater creative freedom from the marketing forces in their own companies, to gain a toehold for innovation within games. Players don’t have to care about whether games are art if they don’t care that every new game looks just like the games that were produced and sold to them last year. There has to be someone out there championing innovation and diversity within the games industry; otherwise, the economic forces will lead towards more and more franchise titles and more and more formulaic games. It doesn’t matter whether there are games in the Museum of Modern Art. It does matter whether the best game designers are given enough room to push the limits of games as a medium and whether or not there are people out there who are willing to support risk-taking and experimentation within the medium."

Precisely.

David Rodriguez, however, disagrees:

"But then, like a bolt from the blue, it hit me.

"I’m not an artist.

"Sure I work in a creative field. Sure many of the things I do are creative and I get to imagine things and attempt to put them into reality. But an artist gets to do what they want, how they want, when they want. That’s not what I do. Someone comes to my company with a contract. They give us money to make something. I make it. They take it and sell it. I don’t work in art.

"I work…in customer service."

But then to explain what he means, he tells a little parable in which he pretends to be an architect--and has a client from hell... but has ultimately to work to the client's demands.

Yes, okay... But you know what, Dave? Architecture is widely understood and regarded as... wait for it... art. Indeed, architecture has been considered a form of art back into the Greco-Roman era. Sure, no architect can ignore the concerns of his client, and many architects wind up doing what they consider uncreative and mind-numbing projects because they need the money--but so what? In his early career, Frank Gehry designed a whole serious of unambitious and run-of-the-mill shopping malls, but he's now famous and highly regarded enough that he can pretty much get his vision through on new projects. Not 100%; every design needs modification to deal with the realities of construction. But there's no mistaking his style, and clients come to him precisely because he'll give them something original and different from the mundane.

The fact that artists--all artists, even fine artists aiming at the gallery trade--need to be aware of their market and often modify what they do to conform to market demands does not alter the fact that they are involved in a creative discipline, the ultimate outcome of which is determined by creative vision. Your conception of art--"an artist gets to do what they want, how they want, when they want"--is foolish. That's not how the world works, in any medium, until you are as famous and as highly regarded as Picasso, Frank Gehry--or Will Wright.
posted by Greg at 8:30 AM 5 comments

Recent, Rich Carlson sent email to the usual list of suspects providing a link to Wikipedia's entry on the Whitespace programming language, a language that is completely encoded in white space (that is, spaces, tabs, linefeeds, and so on).

Actually, there's a multidecade tradition of designing completely impractical programming languages that is, um, hilarious to the right people and completely charactizable as, well, pointless intellectual masturbation to almost everyone else.

Vide:

Intercal
"The language designed to be Turing-complete but as fundamentally unlike any existing language as possible."

brainfuck
"An eight-instruction Turing-complete language."

Unlambda
"Your functional programming nightmares come true."

Funge-98
"Funges are programming languages whose programs are typically expressed in a given topological pattern and number of dimensions."

Haskell
"Haskell is is a polymorphicly typed, lazy, purely functional language..."

Malbolge
"Malbolge was designed to be difficult to use, and so it is. It is designed to be
incomprehensible, and so it is. So far, no Malbolge programs have been written. Thus, we cannot give an example."

I remember, years ago, esr telling me that he had a standing offer to build a compiler for any language anyone could define completely, but I don't see any reference to this on his various websites--guess he's too busy defeating the cathedral with open outcry in the agora, and waging the Long War against Islamofascism to, you know, do pointless cool stuff any more. But someone should take up that dropped gauntlet. Don't think there's a compiler for Malboge, for instance. Not that anyone would use it.
posted by Greg at 3:12 AM 6 comments