pouët.net

I've written a rather extensive piece on the state of the demoscene, by looking back 20 years (and even further, in a few cases) and would very much like to get (constructive) feedback on it: http://blog.subsquare.com/state-of-the-demoscene-in-numbers

Proofread!

Also, you use a Mac, lol.

Fixed the typo, can't fix the Mac-usage, unfortunately :)

Honestly, I don't see any leveling out in that graph... at most an exponential decay. (Which of course I'll take over a linear decay any day). Good job!

Thanks for the insightful blog post. I added it to the Demoscene research site, too. I've done very similar visualizations myself and you can find such images on Daniel Botz's book, too, and they all appear to be in line. Some thoughts based on the feedback I got and other ideas:

- The database dump from CSDb brought a huge amount of C64 productions to Pouet, which partially explains the perceivably high activity
- It's only during the last 10 years or so that "all" demos have appeared on Pouet, which leaves the nineties underrepresented
- Back in the day, especially when considering diskmags, productions were rather small, so the numbers don't tell the whole truth: the amount of time and effort invested in them hasn't fallen as rapidly
- Something that is not visible in the number is that in the 1980s and 1990s there was a lot more circulation in the scene: old beards quitting and newbies stepping in. Now it's largely the same 90s posse that just keeps going.

My own take on the same subject can be found here: http://www.kameli.net/marq/?p=595

Marq: the data is based on the release dates, not the entry submission dates.

And I thought I checked the links carefully. Demoscene Research.

Gargaj: Of course, but there's a lot of stuff missing from the 1990s, not so much from the last ten years.

I guess "leveling-effect" is the wrong wording. What would you recommend? What I wanted to point out was there is a little evening out of the dropoff, especially when taking the complete database into account.

Thanks for the excellent insight, I'll add that fact to the post. However, it's just a matter of coordinating data that was already there, so while some overrepresentation is probably the outcome, it still makes sense to count the data.

What Gargaj said above.
[quote]- Back in the day, especially when considering diskmags, productions were rather small, so the numbers don't tell the whole truth: the amount of time and effort invested in them hasn't fallen as rapidly[/quote
I agree, but the best way to count the productivity of any given culture, company or group, is to tally the tangible outcome of their efforts. In this case that's actual productions.

It's fascinating, and maybe a bit depressing too. And it's making me think that demos just aren't relevant anymore.

Back in the amiga days (when I first discovered the scene), I'd watch demos with various friends. We'd watch them because they'd do amazing stuff we didn't know was even possible, and they'd do it with cool music and cool attitude. Besides, just seeing some amazing code was new and interesting back then. Now the coolness is still there, but people get their "whoa, that's possible?" moment from the big games. A demo doesn't have the same appeal as it did. An amazing 4k sometimes does, but they're rare and appeal mostly to people who understand what they're seeing.

There's so much competition for the scene now too - webgl, processing, the hackerspace type stuff. For somebody into this kind of thing to join the scene, it has to appeal more than them. Does it? I think the answer is no in most cases. Some of the stuff I think is wrong:

- Too many old rules and traditions. File size limits on demos, say. I do tons of stuff with video, and I'd love to do a kick-ass demo using HD video. File size limits mean it's not possible (plus using video is often frowned upon - even if you do something seriously cool with it).

- Limited range of platforms. You do something cool on a new platform, and you have to be careful about where you release it - otherwise it ends up in the wild compo, where it might be competing against crappy videos.

- No interaction/too little flexibility. A compo where we can run through the crowd with a wireless video camera connected into the demo or control the lights in the hall would kick all kinds of ass (and tons of people would love to do shit like this).

There's plenty more. In some of the other scenes you can just turn up at an event with your equipment and your code and show your shit. Amazing stuff is being done out there, but it's not going to happen in the demoscene with the current definition of what a demo is.

To me all that really matters about a demo is:

a. it's coded
b. it's realtime
c. it's cool and entertaining

Maybe that's all the rules we need? If it needs exotic hardware, bring that to the party. If it's interactive, you have to demonstrate that live in the compo (or show a video but have it running in the hall for people to try). No platform restrictions, no size restrictions (outside the 4k etc. compo). Let people submit processing or webgl stuff to the main demo compo.

This is more of an advantage than a hinderance I think -- or at least it can be. The talent in the demoscene is ideally placed to create kick-ass stuff in WebGL which completely blows away the stuff from traditional media companies and other punters.. it's just a matter of seizing the opportunity.

nice initiative. next time include the graph of intros below 4k? or was that included with 4k. anyway nice initiative again and an informative read.

Great article/post.. good to see how platforms trended in both the 90s and in the past few years.

I would like to see cracktros vs demos (& intros). Would be interesting to see how much of a role the cracking/warez groups played in the C64/Amiga/Atari demoscenes.

I would also like to see Pouet's perception of quality over the years.. some kind of thumb-up ratio graph with years and maybe platforms. Or should it be more about how many people thumbed up prods by each year (maybe not.. would be similar to Pouet usage trends).

Glad you pointed out the bump in recent DOS releases was due to 32-256b intros.. couldn't explain why it didn't drop to zero otherwise. :)

Also, a quick grammar suggestion - in Groups, "amount of" should be changed to "number of", when discussing things that can be counted.

that sounds more like a concert-show to me. or wouldnt it just be better that all attendants had a VR-helmet on and just watched the demo like that. interactive segmented reality demos. that sounds weird!

typo.. interactive=non-interactive perhaps

in a way i agree with you, but the demoscene has to think where it can stand out imo. remember that most webgl/processing stuff out there now is totally shallow, simple, and pretty basic technically too.
we've had way more success outside the scene as well as inside when we started doing demos like agenda - which were interesting and relevant to some non-sceners because they didnt look like regular stuff they had seen in a game or otherwise, and in addition they were something that you couldnt go and do in webgl/processing because you really needed to get a lot of the power of the machine to do it at all.

there are ways where the scene can still win out imo. we do have the technical expertise that these other scenes on the whole just dont, if we use it right, and we have 20 years+ experience with the medium.

if all we're doing is something that's basic enough that it could be done in webgl or processing, then yes - no wonder competition will be a problem.



recent major demoparties have more or less eliminated that rule.



ok, im almost with you there, but not in terms of a "demo" compo. as another compo i think it'd be pretty cool though actually. we were talking with a certain organiser about doing an "interactive installation exhiibition/compo" - preferably with sponsor support, because setting those up with good projection hw etc isnt free.

I did notice the push for WebGL demos in 2011.. due mainly to the efforts of Mozilla Labs. Whether or not there is demand for web game/graphics programming out there, I'm not sure the scene is ready to embrace it just yet.

anyway! there's something called a gamedev-compo?

Great article, interesting cycle of platforms. Maybe the Raspberry Pi might appear in the next few years ;)

Speaking of the CSDb dump, is Bitworld and Pouet linked in anyway? I'm sure I've found stuff on there that isn't on Pouet?

And yes I am just looking for the Amiga graph to go higher!!

My idea: the technical and artistical level of demos, intros and 4k has risen really high, from 1991.

So it's harder to be a "scener", so it's harder to produce something worth releasing and watching in 2011.

So I agree with the author: the scene is not dying, the scene is simply changing.

Excellent article!

A bold conclusion? dekadence.

friol: depends on how you look at it. you didnt have glbegin, glVertex3f, glEnd (okay: BAD example) in 1992. you had to code your own polyfiller!. but yea, its changing. but still people are coding c64 demos and people are appreciating it. also because new unimaginable things come out of that machine every once in a couple of years

Great article (=