The state of the demoscene: 1991 - 2011
category: general [glöplog]
rudi: he's right though ;)
I wanna see that "number of demos" graph matched up with a "number of people at demoparties" graph. Hell, even "number of demoparties per year" could be interesting.
Besides the point Keops made about Atari popularity, there's another factual inconsistency that caught my attention:
While the ST was clearly less powerful than the Amiga OCS/ECS machines, the Falcon is actually significantly more powerfull than the A1200, even if you include accelerated machines in the comparison (the only exception could be that AGA accelerators came about a bit earlier).
Other than that, great read, thanks a lot! :)
Quote:
The Atari hardware could also not quite stack up to that of the Amiga, especially in the AGA-age.
While the ST was clearly less powerful than the Amiga OCS/ECS machines, the Falcon is actually significantly more powerfull than the A1200, even if you include accelerated machines in the comparison (the only exception could be that AGA accelerators came about a bit earlier).
Other than that, great read, thanks a lot! :)
Really interesting, thanks for writing this!
What about a scene award for the best graphs of 2012?
Quote:
The reason why Amiga was more popular was because it was a better machine overall for games and demos thanks to its additional chips, allowing better graphics and audio, thus luring many more artists and leaving more freedom and place for design and ideas, whereas Atari demos where mainly about code for a long time, a few exceptions put aside.
Well put right there.
D.Fox: yeah, for the masses :)
Another typo / confusing sentence:
[qoute]However, the most alarming part of this chart is of course that there has been almost no positive growth in none of the most popular categories since 2006[/quote]
The double negative actually implies that there's been growth in all categories, which is not the case at all, unfortunately ;-)
Great article none the less!
[qoute]However, the most alarming part of this chart is of course that there has been almost no positive growth in none of the most popular categories since 2006[/quote]
The double negative actually implies that there's been growth in all categories, which is not the case at all, unfortunately ;-)
Great article none the less!
Arrrh dammit. You make out the quote part yourself ;)
Youtube killed DemoStars!
Quote:
I thought that as well, which is why I specifically checked towards cracktros. The numbers didn't change significantly so as far as the Pouet.net/CSDb-data is concerned, no - the cracktros were not a significant source for the C64-spike.I think the peak of C-64-stuff around 1988 could also be explained by the constant productivity of the cracker age.
Rowley: oops! That's actually entirely my fault, it was supposed to read OCS-age. Fixed now, and added ECS for good measure.
Punqtured: corrected "none" to "any".
Ferris - Slowly Killing the Scene Since 1991
:D
Without having read the article, isn't it a plain logical thing that an increase in quality means a decrease in quantity, given a steady population of demosceners?
It's what Plek said:
Just compare the amount of graphical work going into say, Second Reality and Finally Inside.
Quote:
There simply is no denying that in order to create a top contender that's not a conceptual lucky shot (i.e. highly creative idea), the absolute amount of work has gone way up. So has automation (e.g. better tools, (usually) not programming the system directly in assembler), but that still doesn't make up for it, certainly not entirely :)
Just compare the amount of graphical work going into say, Second Reality and Finally Inside.
Or if that example sounds unbalanced, consider the graphical work going into an Amiga AGA demo now (Starstruck or Human Traffic) and 15 years ago (Nexus-7, Tint).
ferris: haha, props :)
I don't really support the "More polish = less prods"-theory, seeing as "polish" in 1991 was "code". It's just a change of focus, but the amount of work required should be more or less constant.
whereas nowadays "polish" means "blacklisted at german parties".
*drumroll.wav*
*drumroll.wav*
: D ferris
Gloom: firstly, what Gargaj had to say.
This is pretty much different for anyone, each platform, each genre, you name it.
But for the sake of discussion let's limit ourselves to PC demo.
There is no denying the amount of assets has skyrocketed. And these assets are quite a different discipline than busting your balls over 1200 lines of assembly that draw a wobbling torus at top speed, as per your comparison.
The scope of problems a programmer needs to solve to get something of respectable proportions up and running has also greatly expanded. This makes it harder to balance and also tends to leave less time to dabble with single routines to get that cool effect running fast, dandy & goodlooking. Shaders and GPU power have long opened quite a few (back)doors in the visual department, but really impressive stuff will generally still eat heaps of time out of the budget. And this type of programming really is no different than the aforementioned ballbusting on 1200 lines of assembly for 1 screen.
Many groups now have spent enough time on their knowhow, codebase, workflow and assets to quickly or less quickly (if they go for something special) release good demos. But don't forget to accumulate all the hours that have gone into the whole thing instead of the week or 2 that's spent on the release itself. If it ain't more.
Guess why size-constrained productions are so popular? Perhaps because it's relatively sane to set achievable goals without going south of the quality equator. That, I strongly believe, is a major factor in the explosion of 4Ks and whatnot we've seen as of "late". Coders doing what's fun without bothering with huge expectations, 3DSMAX exporters and other bundles of tepid busywork that's hardly even bearable if you're getting paid to do it :)
I'm also willing to go so far as to state that oldschool platforms can be a more relaxed target to work for simply because there's a bit less going on.
And those demos have gotten more complex over the years as well.
What these changes do bring to light though is how enormously important it is to be efficient, both on your own and as a team.
This is pretty much different for anyone, each platform, each genre, you name it.
But for the sake of discussion let's limit ourselves to PC demo.
There is no denying the amount of assets has skyrocketed. And these assets are quite a different discipline than busting your balls over 1200 lines of assembly that draw a wobbling torus at top speed, as per your comparison.
The scope of problems a programmer needs to solve to get something of respectable proportions up and running has also greatly expanded. This makes it harder to balance and also tends to leave less time to dabble with single routines to get that cool effect running fast, dandy & goodlooking. Shaders and GPU power have long opened quite a few (back)doors in the visual department, but really impressive stuff will generally still eat heaps of time out of the budget. And this type of programming really is no different than the aforementioned ballbusting on 1200 lines of assembly for 1 screen.
Many groups now have spent enough time on their knowhow, codebase, workflow and assets to quickly or less quickly (if they go for something special) release good demos. But don't forget to accumulate all the hours that have gone into the whole thing instead of the week or 2 that's spent on the release itself. If it ain't more.
Guess why size-constrained productions are so popular? Perhaps because it's relatively sane to set achievable goals without going south of the quality equator. That, I strongly believe, is a major factor in the explosion of 4Ks and whatnot we've seen as of "late". Coders doing what's fun without bothering with huge expectations, 3DSMAX exporters and other bundles of tepid busywork that's hardly even bearable if you're getting paid to do it :)
I'm also willing to go so far as to state that oldschool platforms can be a more relaxed target to work for simply because there's a bit less going on.
And those demos have gotten more complex over the years as well.
What these changes do bring to light though is how enormously important it is to be efficient, both on your own and as a team.
(and this from the man who hasn't made jack shit in 7 years, yes, I see the irony)
Actually I found those charts nice to look at - but I'm not scared in any way by those nice colored lines on a white background with lines.
That's a core point here. But getting to that point - is really tough work - especially for new groups/sceners it's really tough to beat anything technically cool that's currently out there.
And now please - stop wasting time killing the scene. It wont work anyways.
Oh and another thing:
Perfection is stagnation.
Quote:
What these changes do bring to light though is how enormously important it is to be efficient, both on your own and as a team.
That's a core point here. But getting to that point - is really tough work - especially for new groups/sceners it's really tough to beat anything technically cool that's currently out there.
And now please - stop wasting time killing the scene. It wont work anyways.
Oh and another thing:
Perfection is stagnation.
Plek is right on the money. At a certain point, democoding becomes similar to playing GTA 4. You just died, come out of the hospital, need to spend another 15 minutes driving back to the mission startpoint just to get *a chance* to compete.
After debugging another 3ds max exporter bug, I just went 'fuck it' and found another hobby :)
After debugging another 3ds max exporter bug, I just went 'fuck it' and found another hobby :)
I wholeheartedly agree with plek's last post... including the mini-addendum next to it :-)
