asgaard

Tribes Ascend Survey

Tue, 27 Nov 12 22:37:23 +0000

Hi-Rez sent out a survey for all 'VIP' players in Tribes Ascend today.

They've probably realised their player numbers are tanking (hastened, I imagine, by PlanetSide 2's release last week). It's a sensible change of tactic; previously they've only canvassed feedback from serious competitive players, whereas they've sent this one out to everyone who has a VIP (paid) account.

The survey is a bit strange and apparently proposes very big rule changes without explanation of how they'd be incorporated alongside existing rules. One of Ascend's big problems is that it incorporates concepts from previous Tribes titles with no apparent understanding of how Ascend is different and why you can't just bring things in from different games and expect them to work.

The problem is that Ascend basically needs its rules rewriting from the ground up. Fortunately, someone's already written those rules. They're called Tribes 2.

I kind of hope they do implement massive changes just to make Ascend's death a bit less depressing than a gradual fade out over the next 6 months. Despite being better than Vengeance, it seems set for an even shorter lifespan.

PlanetSide 2 Base Capturing Explanation

Sun, 25 Nov 12 22:18:22 +0000

I found the base capture mechanics completely impenetrable when I started. Here's an explanation of how to capture facilities.

Facilities (biolabs, tech plants, etc) are controlled by one of the three factions. If the facility's territory borders some of your territory, you can capture it!

You're automatically a part of a base-battle if you're fairly close to it. At this point, the capture HUD pops up on the left of your screen to tell you a few interesting things.

The name of the place you're vying for (Crossroads Watchtower) and the faction who currently owns it (NC).
The capture points and who owns them currently. A, B, and C. A is green (friendly - NC for me), B is purple (Vanu) and C is red (Terran Republic).
The presence at the capture points - 2/2, 1/2, 2/2. Basically, having a presence at a capture point will contribute to the capture process, or block other factions' attempts at capturing your facilities. If you've got a presence at all points, it's simple, the capture goes in your favour. If other teams have presences, it gets more complicated. It may be that the defending team successfully holds the facility, or it may be that a third faction manages to overwhelm both of you.
A pie chart showing influence - this is the surrounding territory owned. The higher a faction's influence, the more capping power they have. TR have a massive advantage here.
The progress meter: It's purple at the moment which means Vanu have the most control at the moment, but according to point #3, they're losing it. When the bar reaches zero, another faction's will start to fill up. When it reaches full, that's when the facility is won or lost (and you get lots of points).

You'll often see during longer fights a lot of your team camped outside the enemy's spawn, and all of a sudden, the spawn will switch sides. The switch is determined by the capture points mentioned above. Capturing is a slow process and battles can last for hours if one team doesn't have the advantage. You should be aiming to be at or around one of the capture points, if you want to be of most use to your team, although obviously, disrupting a Sunderer or bringing in armour/air support can be more effective than digging in at a capture point. The capture points are in the form of node structures ad you have to be relatively close to have an effect on it. A display on your HUD will pop up at the top of the screen when you're close enough.

The decision to have three factions in the game has some amusing consequences. We were struggling to hold the Crossroads Watchtower against the Vanu and their bar was starting to fill up worryingly fast. Then the Terrans turned up and gave us some free time while they reverted the Vanu cap (we still lost though).

Undo/Redo in JavaScript

Wed, 21 Nov 12 21:24:01 +0000

Undo and redo is a fairly interesting and surprisingly difficult concept to implement in an application. Thanks to the way JavaScript treats functions, it isn't too hard to come up with a general and powerful approach for JavaScript apps.

We really just need a command stack and a set of actions we want to perform. We can wrap up the details in closures.

Here's a bare bones command stack implementation:

/**
  * @constructor
  */
Commands = function() {
    this.undoStack = [];
    this.redoStack = [];
};

/**
  * Executes an action and adds it to the undo stack.
  * @param {function()} action Action function.
  * @param {function()} reverse Reverse function.
  * @param {Object=} ctx The 'this' argument for the action/reverse functions.
  */
Commands.prototype.execute = function(action, reverse, ctx) {
    this.undoStack.push( {action: action, reverse: reverse, ctx: ctx} );
    action.call(ctx);
    this.redoStack.length = 0;
};

Commands.prototype.undo = function() {
    var c = this.undoStack.pop();
    if (c) {
        c.reverse.call(c.ctx);
        this.redoStack.push(c);
    }
};

Commands.prototype.redo = function() {
    var c = this.redoStack.pop();
    if (c) {
        c.action.call(c.ctx);
        this.undoStack.push(c);
    }
};

The third argument of execute(), ctx, is just a context for the functions to execute with. If you're not familiar, that means you can pass in a reference to a class method and then set ctx to an instance of the class, and it will all work fine. It sets the 'this' value for the function. It's optional.

Now we need some actual action functions. Your functions might be to add an item to a basket, to draw a line between two points, or anything. For ease of demonstration, I've chosen add and subtract. The important thing is that for every action you want to perform, you should have an opposite action defined. Add item to basket/remove item from basket. Or at least, a way to roll back the state (perhaps not a direct reverse operation, perhaps clone the state then reapply it).

function add(n1, n2) {
    return n1 + n2;
}

function subtract(n1, n2) {
    return n1 - n2;
}

And here's how we start executing, undoing, and redoing actions:

var number = 1,
    commands = new Commands();

commands.execute( 
  function() { number = add(number, 5) },
  function() { number = subtract(number, 5) }
);
console.log(number); // 6

commands.execute( 
  function() { number = subtract(number, 2) },
  function() { number = add(number, 2) }
);
console.log(number); // 4

commands.undo();
console.log(number); // 6

commands.undo();
console.log(number); // 1


commands.redo();
console.log(number); // 6

commands.redo();
console.log(number); // 4

That's pretty simple. As a thought exercise, what if addItemToBasket(item) returned an ID which had to be passed to removeItemFromBasketById(id)? Is that hard to handle? Turns out it's not:

var item,
    basketItemId;

commands.execute(
  function() { basketItemId = addItemToBasket(item); },
  function() { removeItemFromBasketById(basketItemId); basketItemId = null; }
);

// NOTE: at this point, basketItemId will have a proper value, supplied 
// by addItemToBasket(). Therefore, you can:

return basketItemId;

The closure scope handles it for us, and the stack structure ensures that basketItemId will always be current. So we can use this approach to have our program functions arranged mostly naturally, and then take advantage of closures and scoping to wrap them up for the undo/redo stack.

JavaScript prototype gotchas

Tue, 13 Nov 12 22:48:13 +0000

JavaScript is a strange language:

MyClass = function() {};

MyClass.prototype.myProperty = [];

var a = new MyClass();
var b = new MyClass();

a.myProperty[0] = 'xyz';

console.log(b.myProperty[0]); // prints 'xyz'

The issue here is one of prototypes and references. In most OO languages, defining a property on a class is done so in such a way that each instance of that class has a property completely independent of all other instances. Even if it's a reference, it still gets a reference to its *own* value.

In JavaScript, things are a bit wacky. Mostly they work as normal (with strange syntax), but occasionally things like this trip you up. It happens because, as you can see, both a and b have a reference to the same object; the one assigned to the prototype. This is only a problem with 'rich' types like arrays and objects, immutable types like string and primitives sort themselves out.

The way to avoid it is to assign 'myProperty' within the constructor:

MyClass = function() {
    this.myProperty = [];
};

It's strange that JSLint doesn't pick up on this.

For what it's worth, gjslint (Google's closure linter) will emit:

Line 3, E:0100: Member MyClass.prototype.myProperty cannot have a non-primitive value

but I find Google's linter is generally not as universally useful as JSLint (it's better for Closure development though, i.e. if you're planning to feed your source through the Closure Compiler).

PlanetSide 2 beta keys

Mon, 12 Nov 12 08:30:38 +0000

The next four people to leave their email address in the comments below (I won't publish the comments, don't worry, your address is safe) will get a PlanetSide 2 beta key/code.

You can read some more detailed PlanetSide 2 beta impressions.[...]

Planetside 2 Beta Impressions

Fri, 09 Nov 12 20:42:17 +0000

PlanetSide 2 is an MMOFPS (you know what that means). I'm lucky enough to be part of the beta test. It releases in a few weeks (20th November).

I've been playing PlanetSide 2 for the last week or so. I've never actually played an MMOFPS, and off the top of my head, I couldn't name any others except for PlanetSide 1.

I vaguely remember a fair bit of buzz within the Tribes 2 community for PS1 (because they were both relatively unique in that they implemented big, open rolling terrain in an FPS, but I suspect that's where the parallels ended), but I never got around to checking it out. History repeats, and there's a fair bit of buzz in the Tribes: Ascend community for PS2. But my impressions are that it has little in common with Tribes, and is in fact strikingly similar to Operation Flashpoint (the original 2001 version, not the rubbish sequel). Because:

It's vast.
Most of the time you get shot at by people you can't see.
Your bullets go just about everywhere except where you're aiming.
It's really confusing and often frustrating.
Movement is clunky.
Vehicles are terrifyingly powerful and indestructable when you're not in them.
Vehicles are fragile and giant "I AM HERE" markers when you are in them.

That about sums up both OpFlash and PlanetSide 2. Sort of. They are both "combat on an epic scale", and it turns out that 'combat' in such context is synonymous with 'confusion'. PS2 is bewildering. That's really the biggest thing you'll take away from it. OpFlash was bewildering too; it was a realistic portrayal of warfare inasmuch as any game can be a realistic portrayal of warfare. You could easily spend 20 minutes lying on a hill because you were too terrified to do anything else (missions were long and the save feature was stringent). Combat was often a few loud bangs followed by a black screen, as indeed is PS2's.

But Flashpoint was single player and back in 2001, the technology for such an environment wasn't exactly widespread. To a certain extent, the helplessness and the irrelevance it thrust upon you was a very convincing illusion, and given time, you could learn to game its limitations. PS2 appears to be much more genuine in this respect, and I'm not sure that's necessarily a good thing, as it is basically an exercise in frustration.

But it is fun. There are genuinely engaging experiences when at night you're cowering behind a rock shooting hapless foes who appear on the horizon. You're positioned well, you're part of a squad, and your superior tactics win the day. The scale makes it even more exciting. But then there's the flipside: You travel across the world uneventfully until suddenly you find you're getting shot by people who are well positioned and well organised, and you're lucky if you even figure out where they are before you get killed. Then you respawn back somewhere totally undesirable and have to make the trek again before you can again be part of the action. This is a frustrating experience.

PlanetSide isn't helped by its complete lack of explanation of virtually everything. You spend a lot of time running around following other people, hoping they have more idea than you. You will likely spend your first half hour in the game wandering around empty terrain wondering where all the excitement is. Perhaps PS1 veterans will understand what's going on, but most people certainly won't. Even the mechanics of capturing locations are unclear.

Luckily, there is in-game voice chat and most of the players using it are quite helpful in directing their team-mates. Players are organised into squads (12 people) and platoons (4 squads) with a leadership hierarchy. If you end up in an organised squad/platoon, the game is *incredible*. You bounce along the terrain in an APC with your squad-mates while your squad leader tells everyone to hold fire to keep your position unknown to the enemy, then a hostile aircraft notices you and your squad leader announces "weapons free", as just a moment later both turrets spring into anti-aircraft action. It really is an amazing form of gaming. And the beauty is that the voice chat is in-game, so regardless of whether you actively partake (or even have a microphone), you can still benefit from it passively without having to worry about third party software and servers.

It is beautiful, too. The lighting effects are lovely. The night time battles are wonderful; the skies and hills light up vibrantly with fleeting flashes, explosions, rocket trails, AA fire and muzzle flashes in a way screenshots just can't capture. This isn't just cosmetic: muzzle flashes have tactical implications. Some weapons even have add-ons that promise to reduce muzzle flash. And night is very dark. You have to stay alert. You can easily end up a few feet away from an enemy before either of you realises the other is there.

PlanetSide 2 will be released officially in less than two weeks (20th November), but unfortunately, in its current state, while the game is interesting and has huge potential, the experience is just too hostile to new players. For a game designed for huge numbers of people, launching in such a state as to give a poor first impression will irreversibly harm its lifespan. It *needs* players more than any other FPS. Tens of empty square miles of terrain is of no interest to anyone; you need hundreds of players online at once to populate it. Hi-Rez Studios made the same mistake with Tribes: Ascend in that they pushed it out to the masses before they'd polished the new user experience sufficiently (and then repeated the mistake with the Steam release), and now, 9 months after release, the playerbase is worryingly low because it failed to hold on to players. T:A can just about sustain itself for now, but PS2 is fundamentally reliant on a large, active playerbase.

The game is also a bit buggy and not entirely professionally run (it is Sony after all). This close to release, bugs like your tank falling through the terrain shouldn't really happen. And maintenance shouldn't take down the game at European prime time two consecutive nights. There seem to have been a lot of failed patches and rollbacks recently, too, which doesn't inspire confidence.

It's ironic that many Tribes: Ascend players were pining for PS2 as a saviour, but PS2 seems to be falling into exactly the same trap of releasing far too early and making the wrong impression.

Update

Here's a concrete example about how badly Sony is handing the software and support side of the game. This morning (11/11, a Sunday) Sony pushed out a 3GB patch, and took the servers down for a few hours. When the servers came back up, players found that they had lost certs (certifications - things you earn and can spend on upgrades of various kinds). So this afternoon, they took the server down again for almost three hours to fix the missing certs. And they're talking about patches to fix crashes (which I think were introduced this morning...), which will no doubt mean more downtime. I know it's a beta, but it's now very close to release and we are seeing large amounts of downtime because they don't test their patches before deploying them.

This is the third day in the last 7 that I've attempted to play at a perfectly reasonable time and found the servers have been down.

Sony Online Entertainment is not a company that is good at this stuff.

Planetside 2 Beta Issues (Servers full)

Sun, 04 Nov 12 22:58:11 +0000

I received my PlanetSide 2 beta invitation today. The EU server seems to be having capacity problems; it often says "Server full" and then seems to suggest it's put you in a queue. These are lies: the best way to get into the game is to press cancel and then try again. A few times. I hope they sort this out soon.

I'll write more about it soon, but so far I am bemused, baffled and indeed bamboozled by the whole game. I wouldn't know where to start. Initial impressions are mixed; it seems incredible, but it would benefit greatly from having some kind of guide or explanation because I generally have no idea what I'm supposed to be doing.

I intend to check it out more, anyway.

Edit (5/11): I'd love to try playing this game but it's always down. 8PM European time, yep, that's a great time to do maintenance! Unimpressed.[...]

Counter-Strike: Global Offensive Impressons/Review

Sat, 03 Nov 12 23:04:47 +0000

I hesitate to call this a proper review as I'm only playing the free weekend on Steam, but nonetheless, here are my impressions.

I used to play Counter Strike back when it was beta and it took two hours to download a 16MB patch. From an .exe, because Steam didn't exist. I played it to death. I was pretty much bored of it by the time CS:Source came out; I had long since moved on to Tribes 2 and Day Of Defeat (not the heretical Source incarnation, which is rubbish). I used to play on the server of a clan called -=P@ntz=-. Their website's member list was called "Who's in my P@ntz?". There was a regular clan there called I Can't Believe We're Not Better. I remember (by the time we were getting a bit bored of CS) we once did a DoD 4v4 mini-tournament with all the server regulars. It was glorious. I have fond memories of CS, and many hundreds (thousands, most likely) of play hours.

Although it seems absurd in the age of endless Call of Battlefield sequels, CS was a innovative game that stood alone in its own genre at its release. It has since been copied and cloned into oblivion, but the notion of one-shot headshots, recoil and permanent (per round) death made CS considered unique, realistic and tactical at a time when online first person shooters meant deathmatches with railguns, rocket launchers and bunny hopping.

Anyway, cut to the present day: Counter-Strike: Global Offensive is a modern reboot of Counter-Strike and was released back in August. I'm not convinced it's strictly necessary considering that the original CS is still one of the most popular games on Steam and the whole franchise combined is far more popular than TF2. But we have it, so let's work with that. It is a modernisation - it comes with all the simplification you'd expect and dread. There's a quickplay option and the weapons buy-menu has been replaced by a bizarre wheel thing which I can only assume was dreamt up by somebody who has never played a game, or in fact used any form of computer, in their life. You can press F1 to auto-buy, which seems to give you either an AK47/M4 or an SMG if you can't afford an assault rifle. And I think kevlar and helmet has been removed in casual play. It all seems to work, but it's not an improvement.

There are lots of new weapons, which are confusing, while some of the classic ones make a reappearance. One of the big things that jumped out at me was how underwhelming the weapon models are; the iconic M4 is now pretty ugly, and a lot of the other models are visually indistinct. There's also music. Which is nice, except when you consider that CS rounds often end with a sneak-em-up where your ability to hear footsteps correlates positively with your life expectancy.

At its core, CS:GO is pretty much a carbon copy of previous games but with a slightly more modern slant. There are new game modes, which seem to be uninteresting additions, and the best part is the 'classic' mode.

It's business as usual, really; the core mechanics are all still there. It's still a game of whichever team successfully rushes to the best camping/ambush location will win by surprising and slaughtering the other team trying to do exactly the same thing. The gun mechanics still work such that crouching and burst fire discipline will have you fare far better than someone running around on full auto. The AWP is still game breaking, sadly.

Some of the classic maps we all loved are still here, including Aztec, Dust/Dust2, Inferno and Italy. There's more focus on the bomb maps than the hostage rescue, but that kind of makes sense given how many rounds used to devolve into watching counter terrorists wrestle with the imbeciles getting stuck in doors. Surprise map inclusions are de_nuke and de_vertigo. I don't think they made it into Source (officially).

But despite the negatives and the fact it's not very original, CS:GO's underlying game is pretty fun (we knew that already), and at £11.99 full price on Steam (currently £7.99, on sale this weekend), it's not really claiming to be an AAA title.

Verdict: Fun but nothing new. The fun, however, is honest, open and transparent. There are no Skinner boxes here. What you see is what you get.