Boring? No. Bloody brilliant. Don't go mistaking words like "quieter", "thoughtful" and "mature" for meaning that Glum Buster is some overly-pretentious artgame. The gameplay is spot on - it was clearly put together by someone who knows how to present game mechanics and puzzles. The whole thing is constructed from these amazing little "Aha!" moments, where you get a bit stuck, hammer some mouse buttons at random, and suddenly your brain echoes with the sounds of pennies dropping, whilst the game wryly smiles and says "I knew you'd work it out". It's slick as hell, and if you give it just a couple of minutes of open-minded willingness to get sucked into its world, it'll get under your skin and reward you a hundredfold. It's a seriously under-rated thing, is Glum Buster. Just talking about it makes me want to play it again.

Got this bit of code... It's one of those mind-meltingly clever bits of C++ template stuff which is actually really useful (rather than just being template wankery for the sake of it) but reads like an absolute nightmare as a result. I spent 6 months writing it in 2004, and have used it in basically everything since then, without ever wanting or needing to look at or modify the scary template code.

Now I've looked at it. And modified it. And made it about twice as good. I am super pleased with myself.

create paradox
become paradox
enter paradox

Doesn't work. I am disappoint.

Sold. 

Your idea of fun seems very similar to my idea of fun, and I thoroughly approve. Mostly I just wish that I'd made this myself, or that I knew you well enough to have helped you make it. I do code rather than pixels though, so I doubt I can be of much use

Got a 3DS today. Still trying to adjust to the 3D screen. In some things I'm hard pressed to see the 3D effect at all, in others it's just eye-hurty, and then from time to time it'll do something that just looks incredible. The first time it looked like that for me was during the AR stuff. I'm really impressed by that. Actually, I'm kinda impressed by all of the features. Confession: most of the time when Nintendo release a new thing it just seems gimmicky to me - I didn't get the appeal of the DS at all to begin with (although it grew on me), and the Wii still seems bizarre to me despite its massive success. With the 3DS, for the first time with new Nintendo hardware, I get it. The 3D is cool (although it'll take a while for the software that uses it to settle down), the AR is super-cool, and I reckon the stuff like StreetPass is going to turn out to be a really big deal.

Hehe, thanks I think in this age of polarised light, lenticular lens displays, head-tracking and steam cloud holograms, it's important that peopl don't forget the Old Ways of doing things. There is much wisdom in those little translucent red and cyan lenses...

@Cellulose: I'm well aware that the linear interpolation thing I described is not perfect, and is not without its problems (Nyquist being a constantly-looming evil in pretty much any DSP-related thing). My main reason for describing it is that it's likely to give better results than simply removing every Nth sample, it's an easy algorithm to understand and implement, and it can be done quickly enough to do in realtime, even multiple times if you want to layer pitched sounds together to make a sort of cheap & cheerful tracker player thing. As has been pointed out though, it really depends on what you're trying to achieve. Audio programming can get to be quite a deep rabbit hole once you get into the really technical stuff, but unless you really need to get into hardcore Fourier transforms and whatnot, there are still plenty of cheap and cheerful effects you can do that sound pretty good and are easy to understand and fun to code.

Just use the standard C way to open and read the file. Look up fopen, fread, and fclose... I can try to write you some more specific example code if you need it, but those functions shouldn't be hard to use.

As for pitch-shifting, since mcc asked, and this thread seems as good a place as any to talk about it, here are some thoughts. First up, for DSP code in general, this place is your friend: http://www.musicdsp.org/showmany.php

I went there looking for a nice pitch-shifting filter (seems like everything that's done "properly" in DSP is done with filters) but I didn't see one. So I'll try to explain the problem in horribly misguided layman's terms. Imagine a nice simple waveform - a sine wav, in a monophonic PCM WAV file. This randomly-googled image demonstrates quite well what happens when a "pure" sine wave is approximated by a bunch of samples in an audio file:



You can see that it's an approximation, but at sufficiently high resolution it's fine for human ears. If you remove every 10th sample, the overall waveform gets shorter (so, higher pitched), but it doesn't do it uniformly. You introduce (relatively) big jumps in amplitude in an otherwise smooth waveform, and those jumps in amplitude get heard as a higher frequency that gets introduced to the sound. The frequency of a sound is defined by how quickly it changes over time, so by introducing a bigger change every 10th frame you introduce a higher frequency.

I imagine there are a few ways around this problem. Just adding a suitably-configured lowpass filter might do a decent enough job of removing the higher frequencies before you write the new waveform to a file, for instance. The way that I've used, that I know works, is as follows. It's not perfect, but it's less horrible than just removing samples. So, you've got a buffer, which is basically an array of amplitude values, like the values marked as * in that diagram. Imagine a playhead moving through that array, reading off a value, and writing that value into a new buffer. The playhead speed determines how we move through the array, to read values. A playhead speed of 1 is pretty straightforward - we read off every single value in the source buffer and copy it to the output buffer, so we get an exact copy of the the original fine. A playhead speed of 2 is easy enough, too: you read every second value from the source array and write it out, which gives you a sound with double the frequency of the original (i.e. an octave higher).

Where it gets interesting is with the non-integral values. If you want the output 10% faster/higher than the input, you want a playhead speed of 1.1. If you make the playback head move at 1.1, then you end up needing to know the amplitude values between the points in the source array. So, the first value you write is a copy of sourceWave[0], but the second value is something 10% of the way between sourceWave[1] and sourceWave[2]. There are nicer, more audio-friendly things to do at this point than what I'm about to suggest, but the easiest thing to write is a linear interpolation of the values between sourceWave[1] and sourceWave[2], and write that value to the output. You're now writing values that might not even have appeared in the original source file, but they do (roughly) correspond to the waveform the file was trying to describe, so you should be all good.

Does that make sense?

Hahaha, the first post makes it look like you fed an image of Lena into the image compression algorithm, and got the image of the baboon as an output 

Cool to see people figuring out what the Gameduino is capable of. I'm looking forward to playing with one myself.

I don't know about Babel or about the exact order of things, but I was playing Crush on the PSP before I first heard about Fez. I always presumed that Fez was inspired by Crush, although I guess they could have had the ideas independently of each other - happens a lot.

Phil, is there anything resembling a launch date yet? I'm keen to play it, and the recent videos seem to show something that's at a pretty near-complete level of polish.

Nope, no playback is needed. This is a batch process that I am trying to do. I actually want to create a function that removes every 10th sample (this will probably speed up the song by 1/10th) and another function to change the value of the samples. Once I know how to do this I can easily create a text user interface so that the user can enter his/her own values. Say if they wanted to remove every 20th sample instead or randomise the samples by ±50%.

Am I right in thinking that the main purpose of this project is to learn more about C++? Just double-checking, because if the aim is genuinely that you've got a bunch of samples you need 10% faster, and everything else is a means to an ends, then writing your own code might not be the best way to go about it.

Incidentally, removing every 10th sample will (I think) sort of do what you expect and give you a rough and ready kind of pitch-shift - but it will very much be rough and ready. There will be side-effects, new frequencies introduced to the sounds. Audiophiles who listen to the resulting files will cry. If you want a better algorithm for pitch-shifting sounds up by 10% then let me know and I'll try to explain the little bit that I know. If you're fine with the junkiness and side-effects, I'll keep quiet 

You can hire people without putting them in a team, freelancers.

Yes:

outsourcing

Because clearly Derek or Cactus et al are going to want to live 'comfortably' (which would take in excess of $3 mil, fyi) while generating several times that value for someone else :/

My bad, I suck at currency conversion.

Hey hey,

For my sins I have done a fair bit of audio programming in C++, including editing WAV files amongst other things. The IDE shouldn't make a difference - you can write code to manipulate WAV files in C++, and you can write C++ in any IDE which allows it, including XCode. You don't need any special libraries whatsoever (well, except the standard file library) to load a WAV, make an edited copy, and save it. You'll need other libraries if you want your program to be interactive - so, a graphics library if you want to display waveforms onscreen or edit them visually, and a sound library of some sort if you want to play the sounds from within your program rather than just save them out and play them in your favourite media player.

So... What do you want to know, exactly? What exactly is it you're trying to create?

you're wrong, you're wrong, you're wronnnngAJUOHFijhbedgkhd

I stand corrected.

Alright then: Passage. 

I would like to see what would happen if we took 1-2 millions and give them to a prolific indie, someone who has a good record of releasing good games all by himself (someone like derek or cactus or whoever), and make sure that the money is used solely on game production assets (not cocaine or stuff).
The result would probably be awesome.

Hmmm. The only way for a one-person indie team to usefully spend that money would be to get more people involved, by hiring, subcontracting, or outsourcing extra work. At that point, the whole thing starts to look fairly indistinguishable from a team which is about a quarter of the size of Double Fine.

Or were you thinking of seeing how many one-man games you could get for $1m? Just pay for Derek or Cactus to live comfortably for the rest of their natural lives, and get 40 or so games of out Derek, or a couple of million from Cactus? That'd be pretty cool.

GTA IV was rumoured to cost $100 million. For that money, I reckon you could cobble together a rudimentary cloning facility for indie developers, and have them make more games than the whole of humanity could ever play.

MindEraser, I had exactly the same problems at university - overcoming my hatred/failing at maths, and still dealing with group projects (in fact, ones in which you're marked on the teamwork itself rather than the end product - which is doubly weird). It sucks, but it's useful, and it's... What's the eupemism? Oh yes - "character building".

Hugs for you.

$1 - $2 million is the square root of fuck all as modern non-indie game budgets go. A lot of AAA studios can burn through $1 million in a month, on staff wage bills alone.

I'm loving what Double Fine are doing. It seems like Schafer has always been about doing stuff with interesting and risky new IP, and I guess he's found that that's been less viable as budgets have risen. To solve the problem by reducing the budgets, whilst simultaneously mitigating risk by working on so many games simultaneously (i.e. "If Costume Quest doesn't go so well, it's okay because Stacking is just around the corner") is a smart thing that other studios should be looking to emulate. The AAA business model as it stands is unsustainable - if the PS4 or Xbox 4pi came out right now, the budgets required to make games for it would likely kill most of the major games studios. So, working out how to do more things, where each thing is cheaper, makes a lot more sense than trying to do one ever bigger, ever more expensive thing.

Not mentioned so far: Limbo.

Don't know if they count, but a fair amount of arcade games were pretty short too, even once you got good enough to complete them. I could blitz through Chase HQ in about half an hour, as I recall.

If you joined a big company, then I believe that you're aware that everything you create there won't be yours.. and I don't see what's the problem with that, they gave you all the resources and the people to do the game, and I the end of the project the game will certainly have changed a lot with input from the company. If you're good enough to create a successful game then you should be able to make another when you leave it.

If you want to start an IP with your name and get all the fame for it, then start your own company

I don't think anyone is disputing the fact that the company should own everything you do there as part of your job. Where it gets iffy is the contract states that they own EVERYTHING you do from when you start working for them until you leave. So, at the companies that interpret that clause particularly harshly, if you work for some games company, you can't write indie games in your spare time and release them, because they own every line of code and every pixel. You can't contribute to open-source software or hardware. If you're in a band, the company owns all the music you make.
I knew of someone who worked for a games company and in their spare time invented a new type of hair grip - they had to talk to the legal department to see if they were allowed to see if they could develop and sell the product, or if the games company would claim ownership or demand a share of the profits.

Perhaps you don't see a problem with that, but I kinda do. My employer owns what I do during office hours, but what I do in my spare time should be mine.

eXistenZ came out at about the same time as The Matrix. Being a poor student at the time, I could only afford to see one of those films. I saw The Matrix a couple of years after that though, so it's all good.