pouët.net

Assembly Summer 2017 oldskool platform lottery

category: parties [glöplog]
or let's just reintroduce an accelerated demo (pc, xbox, android, raspberry) and non-accelerated demo compo (amiga, atari, pokemon mini, mostly the old stuff).. at least technically AND artistically speaking the competition would be relatively fair ;)
added on the 2016-08-16 12:49:04 by maali maali
Maali: you mean turning the oldskool democompo to an unaccelerated "anything goes"-compo?
added on the 2016-08-16 12:51:01 by britelite britelite
Quote:
Coming back to the Assembly oldskool rules and specs, one way to convince us to keep the current specs is to release a few awesome DOS-demos aimed at Pentium-class hardware :)

Working on it as promised at Boozembly. Januarish release date sort of promised. Polishing stuff, adding better scenes.
added on the 2016-08-16 14:04:19 by MuffinHop MuffinHop
This is the story of asm oldskool. Don't touch the rules, no-one makes productions. Touch the rules to get more productions, everyone complains.

Three consecutive years of asking around for compo filler almost made me quit organizing the compo some years ago. That's when Brite joined the asm compo crew to potentially take the oldskool mantle if I ended up quitting. After a bit of fine tuning and lots of advertising last year, we've had two years with a healthy amount of entries now, I'm still in the org and we're a good team with Brite.

This year's rules were a direct result of me getting pissed off at the constant moaning about the rules. Let the people vote for the rules. They did, we got entries, the compo was entertaining, hooray.

We don't have a set plan about next year. C-64 only is pointless at Asm as the C-64 scene will not participate at a "lamer party". Going back to 1992 or older or any 8-bitter will probably cause a lack of entries again, leading to a very unentertaining compo and lots of existential crisis for the compo organizers. Repeating this year's rules in 2017 sounded like a good idea when you mentioned this face to face, and it still sounds like a good idea.

Let it be known, that I don't appreciate the tone of writing you used to start this thread.
added on the 2016-08-16 14:26:57 by Jope Jope
I'm sorry for the way I started this. I could see it after Brite's pissed-off message. The facilities for editing posts are slightly limited around here.

I guess I tried to get different people to throw their ideas out in the open. Where do these ideas about "Pentium is too fast" and stuff like that come from, and what do I actually think about it.

One of the things I cannot sympathize with very much is PC guys speculating about how fast Amiga is and trying to level out the processing power, to make it feel "fair". I can't see actual Amiga folks complaining, and that's how it was 20 years ago. What's wrong with accepting that each platform is inherently different in very many ways, not just on a one-dimensional scale. If the platform is just "chunky pixel buffer + PCM audio buffer + C compiler + some amount of CPU operations per second" then it's not PC, it's not Amiga, it's just boring. Colorless, odorless, tasteless, unattached, uncommitted, generic. Why not get your hands dirty and go low-level deep down into the actual platform?

Another thing that annoys me is "I possess the Universal Truth of what oldskool means, and blah blah feature or tech or something 'is' not oldskool".

About the oldskool PC platform, say, Pentium1+GUS+DOS, being dead. What does it mean that a platform is alive? I say, new prods are being made for it, and compos have it as a platform regularly. This is true for C64, Amiga etc. But Pentium1+GUS+DOS... I think it needs more than one compo to really rise from the dead. Asm 2016 oldskool compo was a good start, but a bit bland. It was possible to "get a way with" just generic chunky effects... this time!
added on the 2016-08-16 17:54:58 by yzi yzi
I don't think there is a problem with Pentium+SVGA being unfairly faster than the Amiga, it's near enough and the audience is supposed to know what they're voting for. The soundtrack and sync is more important for making a demo impressive anyway, both platforms are roughly equally demanding for those. Now we just need more consistency, don't change the rules too much. Tho, as a Pi coder, I will second Maali's suggestion of dividing the compos to accelerated and unaccelerated. The RasPi, gaming consoles with GPU, Android etc should compete against the PC and each other so we wouldn't need to intrude into the wild compo with "milder" stuff.
added on the 2016-08-16 18:23:26 by luutifa luutifa
Quote:
What does it mean that a platform is alive?


What I meant with it is that there are groups still actively developing for it, having hands-on knowledge, and a working codebase. This is the situation with Amiga and C64, where a number of cutting-edge demos are released every year, pushing boundaries further.

Granted, for a large part it's still the same people who did Amiga/C64 demos 20 years ago... But there were PC sceners back then as well. I think the best way to try and resurrect an oldskool PC platform is to have some old PC groups dust off their PC code and build something new.

Perhaps a good idea is to have a standardized 'oldskool PC' platform across all demoparties.
It's not as interesting to develop a P133-demo if there is only once chance to compete with it during the year (and no guarantee that the compo will still be there next year if you miss it this time).
added on the 2016-08-16 19:12:33 by Scali Scali
If there's more pentiums demos released I think that will also increase the likelyhood of there being similar compos at other parties.
added on the 2016-08-16 20:17:46 by MuffinHop MuffinHop
luutifa: when you talk about Raspi, I guess you mean "non-mainstream PC" or wild compos, not oldskool specifically?

Scali: the standardization is pretty much my point exactly.

But if someone doesn't like my opinions, I have others. A technically oriented argument against Pentium and 68060 class power is: as CPU horsepower increases, it gets easier and easier to take the easy way and look at the platform as "C compiler + audio/video buffers". We've seen these kinds of generic demos, where the rest of the components don't do anything significant and meaningful compared to the raw CPU power. The easier it is to port a prod to a different platform, the less "platformy" and the more generic the prod is.
Sometimes the coder actually has something to say visually even without using any of the platform-specific features, and without using the platform's unique characteristics as a source of inspiration. And that's great. But for me it doesn't feel like a very good starting point. Even music is better if each player and instrument is allowed to contribute to the whole in their natural ways, instead of a composer trying to express some sort of purely abstract musical ideas, as if the so-called real world was a mere nuisance.

However, my liking the 1997 spec doesn't come from technical thinking at all, it's more of a cultural point of view. 1995-1997 was something of a peak, the heyday, of DOS demos, at least for me. But did we actually see the finest stuff that could have been done with those very powerful PCs? I don't think we did. The mainstream PC platform kept on evolving rapidly, and nobody wanted to keep studying the 1997 platform indefinitely. Everybody was busy with new powerful devices, more and more each year. Not so with, say, Amiga and C64, which stopped evolving and were frozen as established "painting canvases". (PPC Amiga and MorphOS and stuff like that aside) The sceners themselves kept on learning and now some of them may think, "if I had had my current skills and knowledge in 1995, I could have done so much more..." Or maybe some younger sceners look at 1995-1997 demos and think that they could have done something better or fancier than those.

What characteristics worth using does P133+GUS/SB+DOS PC have then, besides its raw CPU power? Myself, I'm going to keep trying to do stuff in planar modes (instead of old boring chunky modes), make the screen 60 Hz if it's 320x200, or better, use 640x480 which is 60 Hz naturally. Use the timer interrupt for copper-style tricks to beef up things. It's plain DOS, so the machine is all yours. Maybe use OPL3 + PCM on SoundBlasters? Could be interesting. Try do something tasteful with these.
added on the 2016-08-16 20:19:50 by yzi yzi
Not that I've organized any proper parties to know well enough tho... But I think that's to be, expected? No?
added on the 2016-08-16 20:20:24 by MuffinHop MuffinHop
yzi: go make a demo about it
added on the 2016-08-16 20:49:57 by britelite britelite
Were the Pentium-level demos at Asm 2016 captured and/or presented from the real hardware setup, or from an emulator?
added on the 2016-08-17 07:44:38 by trixter trixter
Real hardware always. We don't do emu caps in platforms compos.
added on the 2016-08-17 08:18:47 by visy visy
If the pentium-crowd gets another 2-3 years of compos to sharpen their skills it might actually feel worthwhile to do an 060/AGA-prod again. :D

As for the (so-called) actual discussion here: I'm quite confident that BriteJope will come up with a sane set of rules / platforms - once again ensuring that the quality of the compo depends on the quality of the competitors (and at that point one can only pray to the demonic entity of choice....)
I think this years setup was alright, keep it as is!
About unfairness: People know what platform the demo is made on and the specs are wellknown to most of us aswell, they will acknowledge and vote accordingly.
The Extra-CPU-Power of Pentium is countered a bit by CoProcessors like the Blitter anyway.
Quote:
About unfairness: People know what platform the demo is made on and the specs are wellknown to most of us aswell, they will acknowledge and vote accordingly.


I think so too. If I look at the Revision oldskool compo, basically 'anything goes' as long as it's 16-bit or less.
I'm not entirely sure if this also works if you also include 32-bit platforms though.. but the Amiga compo had OCS/ECS competing with AGA, and that seemed to work as well.

But that gets me back to the process of standardization: at Revision you won't be able to release a 386/486/Pentium production in the oldskool compo.
The thing with C64 or Amiga prods is that there are compos all year round, so if you can't get your demo finished for one party, there's always the next one. I would like to see a similar environment created for some kind of oldskool PC platform. Not quite sure in which shape that should be. What would people think of allowing 386/486/Pentium prods in accelerated Amiga compo's?
added on the 2016-08-17 13:13:06 by Scali Scali
Is there an accelerated amiga compo somewhere?

I already started working on a DOS demo, btw. If there's nothing else to put it in, I'll just put it in a wild compo somewhere...
added on the 2016-08-17 13:21:32 by Preacher Preacher
Quote:
Is there an accelerated amiga compo somewhere?


At most parties? AGA with an 060 accelerator card is pretty much a 'standard' Amiga platform. AGA demos nearly always need an accelerator card, usually an 060.
I was thinking mainly of Revision, where the Amiga compo allows AGA+060, and since 386/486/Pentium prods won't go in the oldskool compo there, they could go in the Amiga compo. That's probably a better match than the regular PC demo compo or the wild compo.
added on the 2016-08-17 13:44:41 by Scali Scali
Quote:
a better match than the regular PC demo compo


Because they stop being PCs when they get old? Oh you and your ageism
Quote:
Because they stop being PCs when they get old? Oh you and your ageism


It's not about age, it's about capabilities.
Revision doesn't allow PPC-accelerated Amigas in the Amiga compo either. Apparently some people think that makes sense.

So, nice troll, but I'm actually seriously trying to make some oldskool PC competition possible.

And if you ask me, nothing other than the IBM 5150 is a PC. I would love to see a proper PC demo competition, but I know that is not realistic.
added on the 2016-08-17 14:18:57 by Scali Scali
I think entering an oldskool PC demo in a compo is pointless, if the machine spec isn't defined exactly. "386/486/Pentium" is crazy talk, because 386, 486 and Pentium are completely different things.
added on the 2016-08-17 14:22:30 by yzi yzi
Quote:
I think entering an oldskool PC demo in a compo is pointless, if the machine spec isn't defined exactly. "386/486/Pentium" is crazy talk, because 386, 486 and Pentium are completely different things.


I think you can't define exact specs for PCs, except for the original IBM PC (no clones).
There is so much variation in CPUs, chipsets, motherboards, video cards and even CMOS settings, that even something like "Pentium 133 with SVGA and GUS" is pretty damn vague, and you can have considerable differences in capabilities and performance between two machines that meet these 'specs'.

Aside from that aspect, I think that keeping the compo specs too limited will also limit the possible amount of entries.
As said earlier, the exact specs don't really matter. 8088 MPH competed in a compo where 'anything goes, as long as it's 16-bit or less', so it competed against GameBoy Color and Amiga 500, among others.
It could also have competed against a 386SX-40, which would have been the fastest PC to meet the specs.
It doesn't really matter, people understand the specs, and will vote accordingly.

So if you make the compo machine a P133, that implies that 386 and 486 prods would work on that machine as well. Nothing that keeps you from entering your 386 or 486 prod in that compo.
I mean, if we take one of the best demos for 486, such as Stars by NoooN... A demo like that would have stood up fine against the Pentium 133 demos from Assembly this year.
Specs aren't all that important, it's what you do with it.
The problem with choosing a Pentium 133 over a 486 is that you also raise expectations. If you choose the faster PC, and you only do things that others can do on a 486, it works against you, and will cost you votes.
added on the 2016-08-17 14:38:47 by Scali Scali
For people who are developing old skool demos on the PC side, are you using DOS box or something similar for your development or what are you using?
added on the 2016-08-17 15:10:50 by d3pth d3pth
Quote:
For people who are developing old skool demos on the PC side, are you using DOS box or something similar for your development or what are you using?


I use DOSBox in the early stages where possible (for hardware-specific hacking, you always have to test on real hardware, because PC emulators tend to be quite buggy), and when things are near completion, I move to real hardware, because at that point things often don't even work correctly on emulators anymore.

But depending on the complexity of the effect, I sometimes write a Win32-based prototype first, to at least get the algorithm worked out and verified, before trying to implement it on the real thing (eg I did a Win32-based delta polygon renderer first).
added on the 2016-08-17 15:19:37 by Scali Scali
Specs don't matter? You can't be serious. If the compo is run on a Pentium, then even mentioning 386 is misleading. And you have to know the specs in advance, with reasonable accuracy. Like compos were in 1997 and like they are now. There is a compo machine, and it is announced and people know its specs with reasonable accuracy. P133+16megs+GUS, it's not vague. Don't start splitting hairs about "exact". The stuff about chipsets and CMOS settings is bullshit.
added on the 2016-08-17 15:48:43 by yzi yzi

login