Tuesday 31 May 2011

More Update

Long break due to Bethan getting ill, meaning I've been home pretty much continually. May continue for a little while, as it's half term. Progress is still happening in fits and starts, however. Axon leg joints now work perfectly, including torque. Am now implementing the "spasming flex" movement, and the fusing of legs that contact with each other.

Wednesday 25 May 2011

Gestalt Update

Got the physics working absolutely perfectly. Then realised I needed to implement rotating joint motors, so added joints in. Broke the physics again.

Monday 23 May 2011

Writing Games Is Easy

Oh dear, here we go again. Just saw another string of comments on a RPS post about how "writing games is easier than ever."

Um, no. Writing games is every bit as difficult as its ever been.

Someone else made it worse by saying "No, it's just coding games is easier than ever."

Um, no again. Coding games is every bit as difficult as its ever been. Its just that a lot of very clever people have very nicely done a whole lot of the work for you. This is a good thing - it democratises the process and means that we have more bodies to throw at the "writing games is hard" problem we mentioned earlier.

Unfortunately, if you want to step outside the boundaries of those elements they've been nice enough to provide, you're still going to discover to your horror that coding games is as difficult now - if not more so - than it ever was before.

Gestalt Update

Nearly there. Legs are in place, IT ALL WORKS OMFG. Moving towards having an actual game at quite some pace.

Thursday 19 May 2011

Rare Stuff and the Retro Sideline

We've been doing this vintage/retro computer stuff for a few weeks now (well, Ian has, I've been coding and parenting) and we keep stumbling into rarities. We've got an Issue 1 Spectrum hanging around, an Atari 2800 (yes, that's correct, a 2800 - the ultra-rate Japanese version of the 2600) and Ian's just shown me a copy of "Grid Master".

Now, setting aside the fact that "Grid Master" sounds like the most awesome name for a CBBC cartoon ever (just try saying it in an evil villain voice), it's apparently even listed as MIA on "World of Spectrum", the definitive online catalogue. So just a teensy bit rare, then.

And this is after 3 weeks.

The sideline's not making any real amount of cash (plenty of turnover, but profit is held back by the need to constantly bring in more stock along with the ungodly postage costs) but it's certainly interesting as all hell.



If you're wondering, the general retro shop is here and you can find the rarest pieces generally on our eBay shop.

Wednesday 18 May 2011

Gestalt Update

Node ownership done. Axon spawning (and ratios) done but not fully tested yet. Custard level checks done. Update cycle implemented. Axon physics implemented but completely and utterly broken.

Once Axon physics is working, I'll prettify what I have so far a bit and release some super-early screenies and vids.

After that, here's the tasklist to first demo release:

- Axon joins - creation of Yellow chains, Blue definitions, Red assemblies
- Mouse picking and live control of Axons
- Red auto-assemble

- Red assembly templates
- Clickables on the node screen

- Paradigm Screen
- Feed in paradigm output already calculated
- Unlocking mechanism
- Create 3 or 4 demo paradigms

- Axon structures on the Node Map Screen
- Animation of defensive structures
- Travel along node lines

- Demo enemy Gestalt
- Node possession
- Yellow dump
- Red defences
- It won't expand or produce Paradigm for upgrades

- Frontend improvements
- Updated intro screen
- Music
- SFX
- Higher quality UI elements

Optional demo tasks:

- Throbbing/Electrical arcing effect on connected Axons
- Invasive enemy AI
- 3-Axon

Tuesday 17 May 2011

On Minecraft, Zeitgeist, And Indie Megahits

Gamesindustry.biz has an interview up with Daniel Kaplan (reg rqrd), the business development officer at Mojang Developments, the firm created by Notch from the success of Minecraft.

Much of it involves discussion of the corporate "noise level" they're under, but a big part of it leapt out at me.

Kaplan: "It's really hard to tell. We still don't really know how big Minecraft is - we have to update ourselves! Last year the huge blogs like Kotaku and Penny Arcade wrote about it - that's when we realised, wow. It was already going well, but when those guys wrote about it it started to go great. They made a huge impact on our sales and brand. We are very grateful."

He's keeping it somewhat under his hat, but essentially this has been the way Notch and Mojang as a whole have been all along on the subject of the "secret of their success." This has been in stark contrast with most of the journalists I've read on the subject, who seem to be determined to promote the idea that Minecraft's success was some kind of magical feat, some organic thing that simply leapt fully-formed from the quality of the product.

This is, of course, purest crap. Minecraft is an excellent game with one absolutely masterful mechanic (the spawning of monsters at night-time, which forces the player to engage the world around them with a strict time limit from the start). But there are thousands of excellent indie games out there, and hundreds using the same paid-alpha approach as Minecraft. Why haven't they all made millions?

Because Minecraft, good a game as it is, got its success not from any inherent quality of the game, but from a thousand blogs, forums, and online magazines carrying the news on Mojang's behalf. There's a name for this kind of self-perpetuating narrative. It's called a "zeitgeist". And this zeitgest, this sheer blind luck of the draw, is the secret to Minecraft's success.

Contrary to the magical thinking of journalists, Notch/Mojang simply got lucky. That's all. Nothing more, nothing less.

But here's the awesome part. This industry isn't filled with humble people, not even amongst the indie community (hello there!). And yet... Mojang pretty much hand credit for their success to the blogs and comics. They're talking about funding other indie titles they like the look of. They even recognise the nascent block-em-up genre they've popularised (not invented, incidentally, the concept is a decade old) and seem happy to watch contenders arise. They're pushing to make their users happy, not to broaden their fan base or increase their DLC sales, but pretty much just because they want to. They've made millions, and yet they're displaying humility.

I don't say this often... but in this case it's deserved. Mojang Developments are The Good Guys.

Progress, of a sort

Well, I got Box2D integrated. Which was amusing, as it shares some typedefs with ProtonSDK. It's kind of amusing that the B2D guys went to the trouble of naming everything b2 and then put a base header with some highly generic defs like int32 in there.

It's not quite working at the minute, the renderer's getting a weird-ass fixture list. I think I'm missing something. Will be at it today. Once it's working, axons should start dropping into nodes.

Saturday 14 May 2011

Blogger Fail

So. I finally write something people pay attention to, start getting decent page views and get a bunch of comments... and then Blogger dies on its backside. Taking my pageviews and comments with it.

Nice going there Google!

Regardless, there's still one thing I can say about it: At least it's better than LiveJournal.

Thursday 12 May 2011

Android Grumbles

We've pretty much ceased to bother with Android, apart from occasional updates to the 3 titles currently on there. The reasons for this:

1) The service is absolutely swarming with crap. There are certain developers that specialise in cyclically releasing virtually invisible updates to their collection of match-3 games every 2 weeks just to stay atop the "just in" list as much as possible. There are hundreds of apps using the name of well-known film or TV shows, musicians, whatever it takes to get search results. There are apps which are basically the name of well-known iPhone games, followed by a lot of spaces, followed by "... Wallpaper". There's a dog fighting game on there. It's not a healthy and competitive market, it's a morass.

2) As a result, most Android users just look at the "Top" section, but that just means that those titles which already sold ungodly numbers sell more. Everyone else is ignored.

3) There's no practical way to market a game to Android users. There isn't even an equivalent of "freeappaday".

4) Android is supposed to be a blanket solution across multiple handsets, but apps have a horrible habit of randomly not working on some handsets when they work fine on others. This gets much worse if you use anything remotely complex such as wifi, bluetooth, or even the Contacts API.

5) Android users don't email to tell you it doesn't work. They give you a 1 star rating and post a comment to that effect. Neither can ever be removed, even if you fix the problem. This routinely happens due to point 4), above.

6) Android users consider £1.50 too much to pay for a game that took multiple people over a month to create, and get angry if the game they bought for 80p doesn't get updated every single month with new content.

So, as you can probably tell, we're not going back there unless someone pays us to, except to bugsplat our existing apps. It just isn't worth the time and effort.

Tuesday 10 May 2011

Gestalt - Concept Part II

Paradigm and Evolution

The Gestalt is capable of improving itself via its Blue Axons. As mentioned in Part I, Blue Definitions (geometric structures of Blue Axons) produce a resource known as Paradigm. This represents the Gestalt's slow modification of its own nature.

These modifications are handled via the Paradigm screen. On this screen is a list (with visual cues to the right, just before the Custard Bar) of different possible developments, each marked as an Evolution or a Paradigm (and highlighted in green-blue and darker blue respectively). One can be selected at any time and this will fill at a rate related to the amount of Paradigm resource produced, however all of the entries on this screen will very slowly fill over time regardless of being selected or not.

Once a requisite amount of Paradigm resource has been gathered, the result differs.

Evolutions unlock instantly, allowing some new possibility for the player - examples include 3-legged Axons (called 3-Axons), new Assemblies for Reds, and so forth.

Paradigms, on the other hand, remain on the Paradigm screen. They must be selected again to become "Active", in which case they begin to have an effect. Once this occurs, they turn steel-grey and can no longer be deselected. However, once this occurs, a new Paradigm will appear that has the capacity to reverse the effects of the previous one.

Paradigms act as selectable "powers" for the player. The player may decide to leave a number of Paradigms charged and ready to be activated if needed, and several Paradigms will be designed around that role (Survival Mentality Paradigm is a classic example).

While most Evolutions provide only simple incremental improvements in the player's capacity (in a manner similar to Tech Trees in other games), Paradigms serve to allow us to almost entirely change the way the game is played based upon the player's decisions.

Enemy AIs

Enemy AIs in Gestalt currently come in three forms:

1) Gestalt
Other Gestalt can come into existence, and these will generally be the most common enemies. They operate in the same way as the player's Gestalt, including Axons, Paradigm and so on. Much as with the player, they can only be defeated by destroying their Yellow Axons.

2) Virii
Rapidly moving enemies that piggy-back the natural data packet traffic on the network (represented by occasional pulses of brightness along the link lines). They slip by most defences, steal resources (which can cause Axons to die), and then flee before they can be attacked.

3) Deus Ex
Monolithic intelligences can arise that are based in a single conscious process on a single Node. These opponents are extremely difficult to directly take on, but have limited ability to affect the network beyond their home Node. They can take over the processing of other nodes and weaken them, and can with great effort move their entire selves to another Node to gain better strategic position.

Suggestions for other kinds of opponent AI are gratefully received. Variation in this will be key.

Play Modes

The simplest play mode is called Conquest Mode. The player is let loose on a network and wins by dominating every Node. There will typically be a number of enemy AIs of the three kinds mentioned above.

Slightly more complex is Singularity Mode. This presents the player with a couple of wrinkles on the basic structure laid out in the Conquest Mode. Firstly, many of the Nodes in the network will have some oddity or special property that might help or hinder the player, and secondly, the game will be accompanied by a relentless red ticker counting up as the game progresses. After a pre-defined period, counter-actions will be taken by the humans running the network; deadly viral counter-agents and worse will be released into the network, and Nodes will begin to disappear. The player can prevent this by controlling certain Nodes that will allow the player to keep the human watchdogs away from their own network.

Multiplayer is intended to be effortless and seamless. If it is enabled in the options (which it will be by default), the game will permit other instances on the same local network to connect to it silently and without mentioning the fact to the player. The game will only stay actively able to host for a few minutes after a game is started, however all games will perform this check on game start.
The maps of players in Multiplayer are connected at their edges, so when the player explores outside of their starting area, they will eventually discover their Multiplayer opponent.

It will of course be possible to whitelist, password control, or turn off this feature entirely.

Monday 9 May 2011

Develop 2011

I just took a look at the list of speakers and topics for Develop this year:


Is it just me or is the entire line-up (Art keynote notwithstanding) just completely nauseating? It's just endless rehashes of the same "EVERYTHING MUST BE CASUAL! EVERYONE MUST USE AGILE! EVERYTHING MUST BE ON A PHONE AND COST 20p! EVERYTHING MUST HAVE SOCIAL FUNCTIONS!" crap that infests Gamasutra these days.

"Can you whether the challenges of the New Games Market?"
- Er, no. Because I can't live off games that sell for £1. I'm neither a student, nor an Indian sweatshop labourer. I have a family to support. Go away!

I suspect the whole thing is going to be more of the same - fashionable braincheese for people with no idea of how to make games and every idea of how to spend the money they make from people who do know how to make them. Coke snorting idiot central, in other words.

That said, I may consider the Indie Dev Day section. £50 ain't going to break the bank and I can go and be patronised and patted on the head along with the rest of my fellows.

Ouch

Went for a very interesting talk with a rep from GO Wales this morning about possibly taking on short-placement interns and maybe even getting funding for conferences. Also, free coffee, which is always good.

I then managed to twist my ankle and as a result I'm not in the office today. I'll try to get some work done, but I doubt there'll be much progress to report.

Friday 6 May 2011

Progress Again

So, interim map generating and rendering fine. Scrolls around OK, have set up correction for position in the click detection. Floating text and other elements work fine, both under and over, which means the Custard Bar will be a doddle. Something's wrong with the node selecting though, I'll fix that this weekend, then we can move into the Node Screen. Not gonna do zooming until later.

Definitely warming to Proton SDK. Takes a while to get used to (lots of boost, ClanLib and so on) but it's good stuff.

edit: Fixed the node selection in the middle of the night while panicking about family stuff. Suffering for it the next morning, though.

Monday 2 May 2011

Delays

Not been at work due to illness (Cerys', not mine). So no new progress.