georgedorn

As Old Man Murray once said, "The problem with making a dark and disturbing version of Alice in Wonderland is that it's pretty dark and disturbing to begin with, which gives it little training wheels that help cultural firebrands ride it into geniusdom once every eighteen months or so."

Sentencing Mr Liddell is a brief retelling of Through the Looking Glass through the lens of a family tragedy. Or maybe vice-versa. In either case, neither tragedy nor dream were particularly deeply implemented. The potential is here for a really interesting world to explore with strange settings and weird characters. Instead, it feels more like a vigorous and repeated nod while somebody off-screen is reading Lewis Carrol - "Yeah! That! Yeah, and I want that in the game, too!" The Alice references vary from subtle and clever to subtle as a sledgehammer, but the silliness is offset significantly by dark undertones and a darker ending.

At the same time, the darkness of the ending really only worked by using cheap shots - the risk is to your own life and the life of your young daughter. No character is actually detailed enough for me to care about them (except maybe the wife, who I would have gladly tossed in the ocean, given the chance), but you can always grab a player's attention by threatening either them or babies. Puppies and kittens are not far behind.

Plot-wise, the "real" world is a railroad plot to the tragedy hook. This could have been a metaphor, but it wasn't. It was a slightly interactive plot dump to get you into the "fun" world, wherein there are puzzles and characters that hint at being more than cardboard cutouts. But between the "real" and "fun" worlds is an endless underwater maze with no room descriptions and the tiniest of cues to suggest you've done anything but gone in circles.

Back to the puzzles. The main puzzle, the collection of words to form the eponymous sentence, was intriguing, and >TIPS at the end suggested there were a few other ways of solving that puzzle. However, since it takes place throughout the entirety of the "fun" world, I ran out of time for replays. The only other noteworthy puzzle is the Tea Party scene, in which you must search for teacups to use as ammunition. Here, the sparseness of the room descriptions really shows through, as one of the nouns you need to search vanished from the description during a scene change and never reappears.

There's a dialog system (of the >ASK NPC ABOUT NOUN variety), and >TIPS assures me that the way I treat family members has an impact on their behavior, but between the sparseness of the implementation and the railroad-y scenes I couldn't tell if this was actually true.

The game also lists four beta testers. I feel like something went wrong there, if >CLIMB ON DESK doesn't work but >GET ON DESK does. Likewise, the number of items lacking descriptions and the limited number of topics NPCs respond to suggests either uninspired testers or the author gave up or ran out of time.

In short, a problematic but potentially good idea that fell short. 4.

The medieval city of Rylane is experiencing a crisis that threatens to destroy life as currently enjoyed by its residents. ( The local population of barrels and crates has skyrocketed... )

This year's crop of IF with religious overtones has been plentiful, but few have been so ( obsessed with vegetables. )

A brief trek through a collection of fey lore. ( Read more... )

So I hafta start this off with a big disclaimer that I'm on a pretty big anti-religion kick right now forever, so I'm sure this review ( is going to be a little colored by that. )

Pak'ma'ra and Ood.



Tentacle and, um, a thing from a space whale.

I'm starting this off with a spoiler alert - I'm not going to shy away from the big reveal because reviewing this without digging into the whole game is futile.

At first glance, this appears to be a fairly sparsely-implemented puzzle game, somewhat zork-like. You're on a quest, though you don't remember what the quest is for or why you're on it. There are some fairly standard puzzles, somewhat on the hard side (the hint system might help), given the brevity of the text. It took a while to get a feel for this - every step of the puzzle was a challenge and some were fairly brutal. This can't be solve by trying everything with everything else, you will need to do some verb guessing on everything.

So, first puzzle complete, you enter a dream sequence. This year must be the year of the dream sequences and flashbacks; I swear there are more of these this comp than the last three comps. Anyway, this one is fairly well-designed and implemented, but the point was lost on me.

Then, on to some more wilderness puzzles (again, somewhat hard due to the leaps of logic needed), interspersed with more flashbacks. While lengthy, they still don't actually tell you anything substantial. A careful reader might think this deliberate, but when this game is alongside a dozen games with no substance, it's easy to miss.

Finally, you reach your objective, read it, loop back (apparently) to the second day again. This is unnerving, particularly because this "start over in a similar place" has been used by games wanting to soften the blow for failure; some players may assume that they lost due to missing something about the book puzzle, >UNDO and try again until they quit in frustration.

But it turns out it's not a loop, you find a plane, and shortly thereafter you're in an airport lounge. That would be the reveal, there. This whole thing was the result of a hallucination/parable brought on by a weird crystal some dude shoved in your face. Now we're not in a quest game anymore, we're in a CYOA. And this is my biggest beef with the game, really - the parallels with a CYOA book are too pronounced. You're expected to make a decision with fairly limited information, some of which is misleading. For instance, the dream (within a dream) sequence and the hallucination suggest that it would be a bad idea to make one choice. Yet another choice, hinted at by the hallucination as potentially the right choice (the wolf is friendly, after all) is also wrong. To top it off, the "right" choice involves noticing an item that appears in your inventory and using it to defeat a supposed government agent who is armed with a gun.

Anyway, it's actually a fairly lengthy game, if somewhat sparse and linear (until the end). It's just that I spent the first 75% of it alternately stumped and bored and never really got into it before the genre-hopping started. The author limited himself to the Z-Code format, and in doing so had to make a lot of compromises. It's an interesting experiment, but given a choice between sparse and big versus deep and small, I'll take the latter any day.

7/10.

This was a series of rooms, each containing a puzzle. To wit:
- one riddle that I found entertaining
- four rooms where the solution is to do exactly as the disembodied voice tells you
- one puzzle where the solution is to not take the puzzle at face value
- one retelling of a classic - the key in the lock on the other side of the door
- one puzzle where the only way to win is not to play.
- one mind-bogglingly complex puzzle (at least, it seemed like it) where even the walkthrough requires 11 steps (or 33 total moves) to finish. Maybe there's a pattern, but this late in the game it didn't seem worthwhile.
- finally, a choose-your-own-ending puzzle, which, with >UNDO, shows you all three mostly pointless endings.

There's some vague fiction involved (you're seeking a goblet or something), but there's not much else here. Mostly, it's a programming exercise.

2/10 for actually having a sorta-kinda game.

(Disclaimer: There are times that I wonder if English is the author's second language; if so, let me apologize profusely in advance.)

This was the speedbump that derailed me for more than two weeks.

I first tried to play this via parchment on the iphone, which works surprisingly well. However, it was in fits and starts as I busywaited during shots on a movie set, which made this already confusing and experimental game impossible. It didn't help that I didn't have the walkthrough, which is needed.

You start off in a garage. Okay, points for not being prone in bed.

Now I'm assaulted by the prose, which is egregiously verbose, melodramatic and occasionally impenetrable.

Examples:

>x bicycle
It appears red under the light.

The bicycle is so beaten that it looks as if it was mercilessly pushed through the spinning blades of a large tree grinder. Its color is now the only deemable characteristic that can be extracted from your examination of the wreck-caused mass of metallic tubing.

Tree grinder? Deemable? Also, if the only characteristic I can tell is that it's red, why isn't it just called "red mass of metal tubes"?

However, there are a couple of objects of importance that lie to the north, one of which you will be shamefully acquainted with soon.

Well, if somebody had told me it was that kind of party...

Wait, was that really it?
You keep staring downwards.

I feel like I'm playing FF7, watching Cloud have a conversation with Some Guy On The Street - something is happening, I'm sure, but beats me what it is. However, Condemned's author missed out on several opportunities to force me to say "...", instead favoring several page bursts of dialog.

Feeling the inevitable, you slowly drop your hand, eventually exposing your view to an object leaning against the wall, clouded by a plastic cover.

>feel inevitable
You can't see any such thing.

>x object
Here IT stands, leaning slightly crooked against the wall in front of you. Its image looks a little clouded due to the plastic cover wrinkled over it, but its tall, huge, monstrous shape can be clearly seen. And unfortunately, you know exactly what object that shape forms.

It waits to be uncovered.

>uncover object
Your head starts to shake like it is undergoing a seizure, your fist drops downwards, your teeth clenches, and your hands grasp your thighs as you close your eyes and lower your head.

Dude. Either cut back on the meth or up your thorazine.

Anyway, onward to a flashback with a bit of a find-the-verb/noun puzzle that had me stumped until the walkthrough. This is largely because I wasn't in the groove of the game, which is:

1) read large text dump
2) spot the verb and noun at end of text dump
3) type in verb and noun verbatim
4) go to 1.

So when the game says:

However, lowering your feet or stepping onto those elevated rocks may prove fatal if you are not careful enough, especially if you have never tested the rocks before for their solidity.

This just means:
>TEST ROCKS

But I'd caved at this point, so I'm following the walkthrough to avoid having to read too carefully this tortured torturing prose. Later in this flashback, the walkthrough says, and I shit you not:
>talk to jill (do this 7x)

I didn't read any of that, but I'm sure the seven pages that follow can be summed up as: angst angst broken home angst bratty sister angst angst nobody loves me. I'm all for teen angst, but -- no, wait, I'm not.

Anyway, the IF proceeds like this, jumping between flashbacks while you search the garage for tools, like rope and a "wooden axe" and getting progressively less coherent and more desperate to shock and disturb the player. Any pretense of this being a game is lost; the flashback scenes involve no actual puzzles other than guessing the next step in the story (and hitting z repeatedly as things happen to you); you blame yourself for a horrible accident, you live in fear of an umbrella stand covered with a black potato sack, your mom pukes up a brain, you kill a dude with a seriously hard-packed snowball.

And tehn evarything got better and they all wented to heaven.

3/10. Author needs a proofreader, an editor armed with a tazer and a soul-crushing internship at a tax accountancy firm.

I predict this game wins the "WTF does that title even mean?" award.

This game is the result of the intersection of half-assed hippie pseudo-ecology and 80s-cartoon nostalgia.

The game is basically in two parts.

In part one, you're in a giant wilderness grid, somewhere around 30x30, with most of the rooms completely irrelevant to the game. You start on a tram (at least it's not in bed) which is surprisingly useful, since you'll be needing to zip around the wilderness on a goose (read: crow) chase for a while.

Are there crows in the everglades?

Anyway, the description of all of the wetland areas:

What commonly gets called a "swamp" is actually the most extraordinary ecosystem
on the planet. Water and grass extend as far you can see, filled with an
incredible variety of birds and fish and turtles and (of course!) alligators.

That's the description for probably around 400 rooms.

Anyway, for no good reason, you follow a beacon (I think you mean tracking device; the beacon is the thing you are looking for) to a hatch, track down a crow that stole your key, try to remember where you left the hatch anyway, go in the damn hatch.

Then:

>x chute
It was originally drilled years ago, by an oil company.

Duhn duhn duhn!

So, get in the giant robot alligator, meet your giant robot alligator brethren (one named, I'm not kidding, Osceoleta), form Voltron Gator-On and join the battle with some other robot.

And then:

Gator-Osceoleta explains all about how the Pyth-nor Consortium (consisting of reactionary sugar growers, building developers, feathered hat makers, and extremist elements of the United States Chamber of Commerce) intends to destroy Gator-on before it can begin its mission closing destructive canals, stopping illegal use of damaging fertilizers, and thwarting poachers

Okay, a) giant robotic exposition dump! and b) holy ambiguous run-on sentence, Batman Gatorman!

It's a quick fight, the game shoves you in the right direction, and now you, uh, fight crime? As a giant gator? Whatever.

HELP/Credits indicates the one beta tester was the author's girlfriend. It showed. There weren't any horrible bugs, but there also weren't enough testers to point out where the game was lacking, and there were a lot of areas like that. For example, the lack of descriptions lead to guessing nouns, not just verbs.

I can sorta cope with the giant grid of copy&paste rooms; I did play DIKUMuds for years. But if you're going to do that, make use of the rooms. I'm pretty sure the author randomly threw in a couple meaningless areas as a tribute to his (and I quote) "significant otter".

But the lack of descriptions in other areas is just plain lazy. Here, I've taken the form of a giant robotic alligator:

>x me
You look mighty.

That's it? Come on.

On the upside, the game gave me a lot to critique. Clearly, the author had some kind of revenge fantasy (for what?) while out hiking in the everglades and felt that only a giant robot (amusingly powered by fossil fuels) could deliver. I get the whimsical nature of it, but I don't get the shoddy, shallow and boring implementation of it.

3/10.

Postscript:
Games with an axe to grind immediately get my hackles up. You, the author, have a responsibility to engage the audience and get them to sympathize with your axe point of view before you start running with it. Show (don't tell!) me the horror of ecological destruction before assuming I want to live out your ecological-defender fantasies. You love the Everglades? Make me love them.

Yes, it's a loftier bar to reach than most genres, but failing to reach it reflects badly on your axe.

One of two games I've played so far via the rather intriguing parchment client.

Lotsa mixed feelings on this game, which, given the showings so far, is actually a good thing. For starters, ABOUT explains, in possibly too much detail, why the author wrote the game. I'm all for some childhood nostalgia, but pour that energy into the game development, not the plea for judges to like it. In other words, write the game for the world, not the comp.

I bought into the game, at first. The backstory was fantastical but I bought it. The immersion level was high, even with such a simple set of rooms, and there were a couple neat uses of flashbacks/alternative narration styles that deliberately jar you into the horror of your predicament. And the kid-in-over-his-head feel of the game reminded me of Arrival, though Interface was less campy and more of a puzzler.

The biggest gripes I have are:
- It was a bit sparse. As much depth was put into the red herring hints (which alternated between entertaining and abusing the player) as it was the game itself. It left me with questions. Why was your Uncle using you as a guinea pig? Why was his assistant such a jerk? Why didn't anybody call the cops? I'm okay with the answers only be alluded to (you're a kid, after all), but a little more depth would go a long way.
- It was really short. I agonized over one section for a while only to find that I didn't even need to solve the puzzle to get the best winning condition - you can win with only a handful of actions and a lot of waiting. For all that wait time, at least give me something to do, or make the puzzles a bit more complex.
- The puzzles were nearly trivial. Aside from a red herring (was the plywood actually necessary?), most puzzles were quite simple and just required you use the right thing with the other thing. The only wrinkle was the four-item inventory, but dropping things to retrieve later wasn't very taxing, given the small number of rooms and ample time limit.

However:
- It was a game, it wasn't especially buggy, it had multiple endings.
- It was believable and engaging, at least for the time it lasted.

6/10.

I wanted to like this game. The opening text quickly set the tone: humorous and hinting that this wouldn't be a normal IF. Throughout, the narrator's voice was timed with ample (verging on annoying) use of pauses. When it worked well, it heightened the effect of a joke. Sadly, it only worked occasionally and when the joke wasn't funny the effect fell very flat. Because the humor was poorly derived (ripped off? aped?) from Monty Python and Douglas Adams, there were a lot of flat jokes.

The plot, if you can call it one, is that you're a beta tester for a generic faceless corporation, testing a, um, thing? Toy? Virtual space? I dunno, it was lofty and never fully realized.

The game actually consists of:
- One complicated open-the-door puzzle. At first, I thought it would be a nifty homage to the babelfish puzzle in HHGTTG, but sadly the solution to this one made even less sense. The main failing was that you don't see all of the pieces to the puzzle until you've taken some arbitrary try-everything-with-everything-else actions, at which point more pieces literally poof into existence. The game even pokes fun at this notion (you're apparently allergic to poof dust) but never actually resolves this conflict. Perhaps upon "finishing" the "game" you should be able to provide feedback in some fashion, like a beta tester?
- Two arbitrary random "games" that you play against a very shallow NPC for no particular reason except to win things you can use in other rooms for no particular reason.
- One "pull the lever and win" room.
- Three rooms with apparently no purpose other than listening to a bad comedian tell jokes, order mugs of ale, beer or (teehee) lemonade, and maybe somehow get food (I never bothered to find out).

So, you solve one puzzle, screw around in a pointless "virtual space" for a while, and when you're finally sick of it, you leave and "win" the game. Bonus stupid for reporting that I obtained 0 out of 0 points at the end, instead of disabling the skill system.

2/10. There was a puzzle, it was bad, and there was some humor and most of it was bad, and there was a whole lot of pointless. I really wanted it to redeem itself at the end, but it failed.

Also, I sorta feel like Ifs should contain some, you know, fiction.

New rule: if a game does not have an ABOUT or HELP page that lists beta testers, the game is on notice. Every major bug or name frustration I encounter knocks two points off the max. If I hit 0, the game is a 1 and I move on to a game that isn't a waste of my time.
( the transcript of the damned )
This last one, though, was the clincher, the cake-taker and the money shot:
>put gizmo in receptacle
That can't contain things.

And with that, I save myself one hour and 45 minutes not playing this game any further. Maybe there was a game in here, maybe not, but come on, get a beta tester.

1/10 because I'm drawing a line in the goddamn sand here. I'm not even going to try to come up with a fitting song for this entry.

<rant>
There are cruel games and then there are Cruel games. Some people like Cruel games. I don't know who, but somebody must. I imagine these people also enjoy maximum-difficulty Sudoku puzzles and playing chess against Fritz. I am not one of these people.

Let's quibble over the difference between Nasty and Cruel for a moment. Nasty games tell you when you make the game unwinnable. Cruel games do not. In a Cruel game, you can do something apparently innocuous on turn 3 (like spending one extra turn examining something instead of taking a necessary-to-win action) that makes the game unwinnable and never learn the reason why, short of exhaustively trying every combination of possible moves.

Broken Legs takes this a step further. I would not call it a Cruel game (despite the game's ABOUT), as doing one thing in the wrong order doesn't put the game in an unwinnable state - it puts the game in an un-endable state, forever stuck while the player searches in vain for a trigger that isn't there. If this were unintentional, this would be called a deadlock and would require a bugfix. This isn't just Cruel, it's Cruel + lazy or mean-spirited.

Even worse, should you notice that the game is unwinnable, the undo command is limited to 8 turns preceding. I don't know if this is a Glulxe limitation or if the author is just that Evil.

Additionally, there were a lot more guess the verb/guess the topic puzzles than there needed to be, and a few instances where you are punished or misled for guessing the right phrase at the wrong time (sometimes by as little as one turn).

In any case, several points docked.
</rant>

Aside from the author being cruel, Broken Legs is just chock fulla cruel. You play a horrible ambitious teenager and because you can't win an audition legitimately must resort to systematically destroying your competitors. Every other character is also a terrible person (with the possible exception of the dorky "stage kid" character) so they basically deserve it. Even the protagonist's mother, who brilliantly serves as an in-game hint system, behaves atrociously, even towards you.

Despite leaving me feeling just a little ill, this all worked great. If I can forgive the suicidally-depressing Initial State, I can downright enjoy petty, vindictive dramas, particularly when the characters are this believable.

It's a shame I couldn't make myself go back and try to play through to the conclusion.

Tentatively 5 out of 10.

This is a sparse game with essentially one puzzle and 5ish scenes, if you count an empty wasteland populated only by you and an animate windmill to be a scene. There are two possible outcomes, determined by whether you succeed in predicting how to stop a bullet in a duel, so there's an element of learn-by-dying. There's some subplot about your wife having left you, but I lost interest because it seemed little more than background to set the scene. I gather during one of the flashback scenes you can watch her write a poem, and maybe this hints at why she left or how you can survive the inevitable duel scene, but again, no interest.

There were a handful of nice touches along the railroad, though. Clearly the main npc had a lot of thought put into him, the flickers of memory as you explore your house added some depth and I gather there's an easter egg relating to last year's Violet.

Overall, though, a lightweight effort without a lot to say. 4/10.

The game starts with a dense block of narrative not directly about the game. Likely it will make more sense in the future, when episodes 2-18 come out and I applaud the grand vision. It does foreshadow some small plot points, and the author makes good use of scene changes with other narratives sprinkled throughout.

Plot-wise, you're a space marine on a rescue mission that goes bad pretty quick. That was a relief, actually, as I was glad to ditch the few NPCs in some fairly dramatic scenes. Wounded and losing air, you take refuge in a restricted-access facility that's seen better days.

And that's the routine of the game - perform some minor puzzle tasks, get rewarded with another narration. It's an IF version of a FPS, complete with endlessly spawning enemies and a linear path. The problem is it's not especially interactive. Take the twitch gameplay out of a mediocre FPS and you're left with cutscenes interspersed with some jumping or find-the-key puzzles.

If that's harsh, I don't mean it to be. The narration is interesting, the pacing works well and the puzzles are fairly intuitive (frequently involving a pressing problem like bleeding to death or running out of air that you need to solve to move further). It's a solid effort.

7/10.

What, already?

Anyway, going to mostly follow the rating system from last year unless I notice a trend.

My list, updating links as I go:

The Duel That Spanned the Ages
The Duel in the Snow
Broken Legs
Star Hunter
Beta Tester
Interface
GATOR-ON, Friend to Wetlands!
Condemned
The Grand Quest
Snowquest

Not Reviewed due to lack of time:
Earl Grey
Gleaming the Verb
Grounded In Space
Yon Astounding Castle! of some sort
Trap Cave
Spelunker's Quest
The Believable Adventures of an Invisible Man
zork, buried chaos
The Ascot
Byzantine Perspective
Resonance
Eruption
Rover's Day Out
The Hangover

Actually, there have been several GDCU challenges, but since I didn't make the enrollment in time, I've just been picking and choosing. This one came out of some brainstorming that jearl and I did. Also, there were other names floated for this one, at least one of which was better than this, but I can't remember what they were.

The GDCU challenge was to come up with a game in which some of the rules are made up by the players. After some deliberation, we settled on some goals:

• There should be more strategy than randomness, even though some randomness is desirable.


• Everybody should be made to deal with the rules that they make up; it's much less fun if everybody makes rules that only impact other players.


• The game should provide enough structure and theme that coming up with new rules should be easy; this shouldn't be as inaccessible as a boardgame version of Nomic (not that I have anything against Nomic).

( Anyway, the rules and game board... )

The exercise was to make a silly little path game like Chutes & Ladders or Sorry. I wanted a game with a bit more mathematical strategy than your typical path game, so I made one where you get to choose from random options each turn. Took about 20 minutes to come up with the idea, 20 minutes to type it up, and a bit more than than to monkey around in photoshop for no good reason.

Click for big version
( The Rules )
So, nifty things about the exercise:
- It's pretty easy once you get started.
- It doesn't take much to add strategy to an otherwise random game that could play itself.
- Color (by which I mean flavor text) can really make a game evocative even when very simple. I probably should have fleshed out the plot just a little better. How did the players end up on a rock in the middle of a lava flow, anyway?
- Photoshop is pretty cool and a great way to waste a lot of valuable time.

(This is where I'm collection some notes and thoughts as I take the free online game design course here: http://gamedesignconcepts.wordpress.com/.
I will be cutting for length after a brief summary for each 'class').

Summary: Discussion on what constitutes a game.
( Read more... )

Incidentally, while I've been thinking about the subject matter of the previous post for a while now, I was really inspired to think on it after reading a new game from the author of Kill Puppies For Satan and Dogs In The Vineyard.

It's a high fantasy game that captures some of the feel of D&D, particularly by leaving so much open for interpretation (preferably while it is being played) that you can do almost anything with it.

It's an interesting read; the first book is here (warning, pdf). Even if I never get a chance to play it (though I would like to give it a try) there are quite a few elements of world design that I would incorporate into a game in the future.

There are two really good things that I like about 4e:

1. The magic ritual system is really freeing for spellcasters, who don't need to choose between utility and combat spells anymore.
2. Suddenly everybody is pulling out old D&D editions and going back to their favorites.

The second one isn't entirely because of 4e, but it certainly helped. There's also OSRIC which has gained sudden fame for publishing free versions of really, really old editions of the game.

These old editions all have their own look and feel; none have been quite perfect, each had some pretty major and sometimes embarrassing flaws, but all of them had certain qualities that I think the D&D community is generally longing for. Of course, nobody can agree on what those are.

If I were to make an amalgamated version of D&D with all the best qualities of all of the editions I've read or played, this is what it would contain.
( What I'd steal from each D&D to make the UberD&D )
What would be on your list? What did I miss? What am I totally wrong about? (Surely, there must be something else in 4e that is merely an improvement and not a complete departure from the original feel of the game...)

Edit: added initiative from 3e

These are from last night. They're very rough, I haven't edited them at all. A nice feature of Reaper/ninJAM is that is saves each track separately, so you can edit everybody's contribution independently. I just haven't gotten to it yet.

The first is a chill synth piece. I threw in some synths but mostly did drum loops.
The second was a fun, impromptu 8-bit noise jam. It drifts a lot, starting very chaotic but then drifting into basically two different almost-tracks. Again, I threw in some synth but mostly did (8-bit) drum loops.

Been having a lot of fun in the test servers lately. The styles are all over the place, from blues to folk to hard rock to metal to various electronica. At the moment, I'm limited to electronica - midi synths and drum loops. I program a pretty good drum loop and push audiomulch pretty far for bizarre sounds, but the hard part is finding a way to work that into whatever jam is going on at the moment.

Tonight I had two really good ones, though. A couple of guys were hopping from server to server, trying to get interest up in what eventually turned into an all 8-bit jam. It took me a bit to grab a free 8-bit synth VST and a handful of drum sounds sampled from the NES, but then we had a lot of fun.

Then I hopped onto a different server by myself and threw together a 145-bpm techno track and let it run for a while until somebody joined the server. It was another midi keyboardist, who swapped out some sounds and joined in.

Normally, the server automatically cuts tracks out of stuff that happens, but the autosong server it down at the moment so I'm not sure I'll be able to link to it.

Also, after this weekend I'll have the Drum Rocker, which is made for Rock Band but also can be finagled into being a decent set of MIDI drums, so I'll be able to play live drums as well as loops.

So, NinJAM itself is free, but is extremely limited when it comes to sofware inputs, largely because it does not allow input from one source (a virtual audio cable) and output to a different destination (e.g. your sound card). This makes it impossible to hear your own music while sending to the server.

There's probably a few ways around this. Ideally, there would be an all-free (at least as in beer, preferable as in speech) way to do this. There's a few near solutions, but none quite add up:
- Most audio apps allow you to output an extra signal as a monitor or cue channel. The problem is that most have the same limitation where they can only output to one type of destination (the virtual audio cable) and can't simultaneously output to NinJAM and the sound card.
- There isn't, to my knowledge, a free virtual audio cable. JACK Audio Connection Kit is free as in speech. I'll experiment.
- There may be a way to have NinJAM or the synth program output on a virtual cable (like JACK) and have yet another program send that to the sound card.

NinJAM is a collaborative music tool that allows you to play music with other people over the internet. The basic client works great with sound card inputs, but a lot of people use software synths, beatboxes, or effects, which requires an extra computer, some tricky sound card wiring, or a virtual audio cable to wire up sound between applications. This guide is for the latter option.

1. Install ASIO4ALL, which provides decent ASIO drivers for sound cards that don't have ASIO. You probably need these unless you have a high-end sound card aimed at music production.


2. Install Reaper (not free). Be sure to install ReaNINJAM (under Audio Processing Plugins) and ReaRoute ASIO Driver (under Additional Functionality.) This last one is a virtual audio cable, allowing you to output sound from other programs (Ableton Live, for example) to a virtual sound card and input them into Reaper (or anywhere else, for that matter). You can also use ReWire or Virtual Audio Cable, or several others, so long as they support ASIO.


3. Start up Ableton Live (or whatever). Configure audio output. Use Driver Type: ASIO and Audio Device: ReaRoute ASIO (or whatever virtual sound cable you chose).


4. Start up Reaper. Under Options/Preferences/Audio/Devices, configure Reaper to use an ASIO driver (e.g. ASIO4ALL from above).


5. Add a new track in Reaper. Drag the bottom of the new track down so you can see the green->red sound meter. Click the 'In' button, go to 'Stereo Input' and select 'ReaRoute1/ReaRoute2'. This tells reaper to listen to input from the virtual cable. Hit the R button on the left side of the track to start recording input from the cable. (I don't think I had to push the R button the first time I did this, but now I seem to.)


6. Start up ReaNINJAM, which is done by hitting the FX button on the master track. In the FX chain window, click 'Add', go to 'Cockos', select 'ReaNINJAM'. Once you've added this, back in the FX chain window you'll see VST: ReaNINJAM. Highlight that if it isn't already, and click the "Show ReaNINJAM Console" button.


7. In the ReaNINJAM window that comes up, go to File/Connect, enter the server settings, and click 'Connect.'

You should now be able to make noise in Ableton Live (or whatever) and both hear it and send it to the NinJAM server for other people to hear. You should probably save the Reaper project to avoid having to re-create the input track.

There's a really good chance that if you are playing loops, they're out of sync with the NinJAM metronome. This will be really irritating to other people on the server. To fix this:

1. In ReaNINJAM, uncheck the 'Xmit' box to stop transmitting to the server while syncing up.


2. In Ableton Live set the BPM to the same as the NinJAM server. Listen for the start of a two-measure phrase (the tick of the metronome is a different tone on the very first beat) and press play at the same time. It's good to start with a beat that's simple, so you can hear if the timing matches.


3. If the timing doesn't match (and there's a good chance it won't), hit stop, wait for another measure, and hit play again. You can also use the tap key to sync up, but you probably want to do that just to get the offset correct and then set the BPM again or you'll drift.


4. Once you're synced up, you can hit the 'Xmit' button again to start sending to the server.

Going back to possible hybrid architectures, there may be a way to use 'Log and re-run' without implementing the same game engine in two languages (python and ActionScript).

( Solving the client/server issue )

I've been working with jearl on a game project for about a year and a half now. It's a flashed-based multiplayer roguelike, featuring dungeon crawls finishable in about a half hour. Hence the working title, 'Lunch Crawl'.

Anyway, about three months ago, after a longer period of stagnation, I was inspired to rebuild it using Google App Engine, to solve a conceptual problem I was having regarding cheating. It all boils down to running the game logic on the client (the browser) or the server.

Client-side architecture:
In this design, the game runs entirely in the browser, uploading results (like a score) to the server, either periodically or at the end of the run.
Flash-based games in which you keep score have a long history of cheaters. In the simplest of abuses, you notice that a flash game submits a score to a website with the score in the url, or in some data posted 'invisibly' in a form. There are some minor things that can be done to make it harder to cheat, but that tends to just motivate people to try harder. You can decompile flash and replace the "attack-the-monster" function with one that always hits for 1000 damage. Or you could replace the "player-take-damage" function to make it so the player always heals immediately afterward. This isn't strictly a flash problem, either; First Person Shooters like Half Life, Quake and even Halo have cheating problems. An early example is to replace all the wall textures with transparent textures...

Server-side architecture:
This gets rid of most of the cheating problem; the server rolls all of the dice, keeps track of all of the stats, and even limits what the client knows about the player's surroundings. In Half-Life terms, it does no good to replace the walls with transparent textures if the server doesn't tell the client the locations of anybody you can't see. Games like World of Warcraft use this approach.

The downside is that it's very expensive to run everybody's actions through a server. It's the main reason WoW needs several servers per realm, despite not having particularly complex AI outside of instances. It's also the main reason that multiplayer FPSs still struggle with players-per-server limits.

This downside is especially painful for development of Lunch Crawl, as the plan is to use a web-based service to process turns. In short, every time you do something in the game, it sends the command to the server as if you were going to a web page, gets the results, and renders them in flash. So in addition to the time it takes to process the action and send the data, there's a per-turn price that comes with using HTTP. I'm optimizing as much as I can, but until HTTP requests magically become faster, I doubt I can make a turn take less than 250ms. Seems like not very long, but trust me, you'll notice it when you're trying to walk across a big room.

Hybrid architecture:
There's a few ideas out there for doing a little bit of both. The biggest problems with all of these is that they're complicated to write and they don't completely prevent cheating.
- Guess-and-check. FPSs do this, to some degree. You allow the client to predict the outcome of an action while waiting for the response from the server. This still requires the client to wait whenever things get complicated (like, you shoot somebody) or when network congestion makes responses come back out of order. Usually, when things get messed up, the whole session needs to be rebuilt using what the server knows about the situation, and it tends to be jarring. In a roguelike, this could mean that you run down a corridor only to get yanked back a dozen squares because the server just realized you noticed a secret door.
- Log and re-run. In this approach, the client is allowed to go hog-wild. The dice are rolled on the client, the character's kept track of on the client, etc. Everything that happens is logged, though, and periodically the log is sent to the server. The server then re-creates the scene and runs through the player's actions again, using magic dice that roll the way the client was expected to roll. If the end results don't match up, the client must be lying. There's a couple problems with this approach, however. First, it means coding the same application on both the client and server, which often means two different languages. This could be extremely complicated - one small bug in a routine to determine if you could see a monster, for example, could result in clients being labeled cheaters when they aren't. The other major problem is that it won't stop all cheating. If the client knows where all the monsters are, there's nothing stopping a cheater from displaying them all on the map. Same thing with invisible monsters, secret doors, and fog of war.
- Log and detect cheating later. The weakest of the anti-cheating methods. The idea is to know ahead of time what the highest possible score for a game is and predict, roughly, what score the player should realistically get. If you keep enough of this data, it should be possible to detect the worst of the cheaters. For example, if a dungeon is created that is inhabited by 30 goblins that are each worth 2xp, then it's pretty obvious that a player reporting 75 xp is cheating. Likewise, if a player always reports always killing every monster, thus getting maximum xp, but never reports that they used a healing potion, they could be a cheater. Statistical analysis could be used to spot the likely cheaters. Traps could also be laid for players suspected of cheating, such as placing some monsters in unreachable rooms or swapping out the client for a different one (see the next option).
- Encryption, obfuscation. This isn't really a method to prevent cheating so much as a way to make it harder to cheat. If an encryption key is hard-coded into the flash client and changed periodically, then any cheater that is using an altered version of the flash client (with, say, god mode) will suddenly be unable to communicate with the server. This does mean forcing players to download new clients periodically, as well as requiring some scheme to prevent cached flash clients from being immediately invalidated.


Anyway, I'm stumped. I was toying with a combination of the hybrid architectures when I last got stuck due to overcomplexity.

This is my third year playing and judging IFs for IF Comp, a competition of interactive fiction (think Zork or Adventure). This year, I'm taking a cue from fellow If Mudders and reviewing entries instead of only assigning a number.

There's 35 games. I'll post with cuts, but I'm not filtering (so that non-friends can read), so it may get a little noisy.

( Judging 'criteria' )
( The Games )

The most common retort against privacy advocates -- by those in favor of ID
checks, cameras, databases, data mining and other wholesale surveillance
measures -- is this line: "If you aren't doing anything wrong, what do you have
to hide?"

Some clever answers: "If I'm not doing anything wrong, then you have no cause
to watch me." "Because the government gets to define what's wrong, and they
keep changing the definition." "Because you might do something wrong with my
information." My problem with quips like these -- as right as they are -- is
that they accept the premise that privacy is about hiding a wrong. It's not.
Privacy is an inherent human right, and a requirement for maintaining the human
condition with dignity and respect.

Two proverbs say it best: Quis custodiet custodes ipsos? ("Who watches the
watchers?") and "Absolute power corrupts absolutely."