15 August 2019, 23:30 | #81 | ||||||
Registered User
Join Date: Jul 2015
Location: The Netherlands
Posts: 3,411
|
This is really off-topic. Perhaps we should consider discussing this elsewhere
Quote:
Granted, these Amiga games tended not to be action games, but that is besides the point here. This is not to say that action games are no fun. I can still remember playing Speedball II for hours. Great game, easily top tier NES (or even SNES) material. Oh and Dynablaster, which is literally a console game on the Amiga and a damn good one at that. On a personal note: what I always find striking in these discussions is how it just doesn't fit with my personal experience. Despite the so called quality difference in favour of consoles, all the 'console' kids I knew ended up playing C64/Amiga games for hours at my house, generally preferring it to playing the console games. Quote:
Quote:
Many NES platformers (including quite a few of the classics), for instance, commit the horrible, horrible design sin of making the player jump back on a hit. That is (IMHO) one of the worst things in gaming ever and the NES pretty much made it popular. Not a good thing. And yes, I'm dead serious on that point. It's totally arbitrary and generally leads to many unfair deaths. It can totally spoil a game play session or even game. It's one of the main reasons I thoroughly dislike Turrican III. I'd rank it as just as bad as leaps of faith or 'please redo the game again to get the ending', which is also really bad design. Quote:
Quote:
Quote:
|
||||||
15 August 2019, 23:50 | #82 | |
Warhasneverbeensomuchfun
Join Date: Jun 2001
Location: Rio de Janeiro / Brazil
Age: 41
Posts: 3,450
|
Quote:
Yeah, I am feeling guilty for derailing a topic for the 2nd time in less than 24 hours. Back to fast collision detection, hehe |
|
16 August 2019, 10:58 | #83 | ||
Registered User
Join Date: Jun 2016
Location: UK
Posts: 428
|
Quote:
Western developers did make good games, but they tended to be slower paced, more focused on puzzles, and worst of all designed to run on a wide range of 8 and 16 bit systems. Many Amiga games were hobbled by being co-developed for the Atari ST, for example. Quote:
|
||
16 August 2019, 11:02 | #84 | |
Registered User
Join Date: Jun 2016
Location: UK
Posts: 428
|
Quote:
Gamesack did a review of the CD32 once. They didn't like most of the games, and in some cases the criticism was fair because the CD32 wasn't all that competitive. Often it was just that the games were not console style and what they were used to though. Both are an acquired taste it seems. |
|
16 August 2019, 17:54 | #85 | |
Warhasneverbeensomuchfun
Join Date: Jun 2001
Location: Rio de Janeiro / Brazil
Age: 41
Posts: 3,450
|
AAAANNDD....... we're are moving to split another topic about console games...
Quote:
In the other hand japanese devs weren't trying to FIGHT the hardware, instead they tried to do good games that would work easily on the hardware. Look at Super Mario Bros... scrolls only one direction, isn't pushing the NES at all, friggin brilliant game. Another thing that I feel it was a big difference was the size of the staff. Yeah, japanese games also had those "Whole game coded by a single programmer who had just graduated from university working on a small office shared by other 5 coders", but the bigger releases had proper staff. Looking at Vampire Killer on MSX (Its Castlevania version), the credits roll at the ending show SIXTEEN people working on the game. What amiga game had 16 people working on it? And that was an 8 bits game! I agree there's a very obvious difference between American, European and Japanese devs, at least during the 80s and 90s. At that time you would never see a japanese dev doing a game like Cannon Fodder or Syndicate, both amazingly good games. But at that time what we were playing the most was pure arcade games, and on this area the Japanese guys were gods.... and I think that's the main reason: They were more focused in bringing a good game than pushing the hardware. I don't think it's a coincidence the games I find that push the Mega-Drive hardware to its limits are NOT Japanese, but western ones (Batman & Robin, Vectorman, Zero Tolerance, Red Zone...) Are they better games than the most impressive-coded japanese games on the system? (Alien Soldier, Contra:Hard Corps, Gunstar Heroes...) Absolutely not. |
|
17 August 2019, 00:27 | #86 | |
Moderator
Join Date: Nov 2004
Location: Eksjö / Sweden
Posts: 5,604
|
Quote:
I think most of us here have the experience of games from the 80s to today, especially 2D games such as arcade and the arcade ports we got on Amiga. I mention flickering because on arcade/port, this wouldn't be tolerated. I'm here to say that Bullet Hell is irrelevant for Amiga, and that competent collision detection is both possible and necessary to produce Amiga games that are better than the ones we got. |
|
17 August 2019, 02:09 | #87 | |
Registered User
Join Date: Feb 2007
Location: Melbourne, Australia
Age: 41
Posts: 3,773
|
Quote:
Name one Amiga game as good as Metroid? Or Contra? Or Final Fantasy? Or the Mega Man series? They simply don't exist. |
|
17 August 2019, 12:09 | #88 | |||
Registered User
Join Date: Jul 2015
Location: The Netherlands
Posts: 3,411
|
Quote:
The notion that European games that were good were only an 'accident' is really rather insulting to the people involved, don't you think? Quote:
Which might be why I keep being mystified with at all this 'consoles rock' stuff Quote:
For instance: I personally find the Mega Man series and Contra to be some of the most overrated games in history. I also think that the first Final Fantasy worth playing is FF-III/VI and the ones before that were average at best. Of the games you listed, the only one I personally feel is any good is Metroid. And I still think even that is overrated. I don't think it's bad, but I don't think it's amazing either. Over and over we get people claiming the Amiga is bad for games because they like some console action game or another more than what was on the Amiga. Well, I had more fun with Elite:Frontier 2 than all of those games you listed combined. I'd exchange all of them for the poorly converted OCS Amiga version of Civilization in a heartbeat. And that's not even getting into the wonders of those lovely SSI titles, which all stand head and shoulders above console action games in my mind. As for that one game. I'm sure it has to be an action game that could also be on a console for you to even consider recognizing it, so I'll offer you Dynablaster. This is Bomberman and it plays exactly the same as on consoles. Bomberman is widely recognized as a true classic console game and generally considered one of the best games ever made. It is available on the Amiga. Even runs in 60Hz. Now, if I understand the rules of the 'Amiga vs Consoles' game well enough, you're now going to claim that Dynablaster/Bomberman is actually not a good game because... Reasons. Or, you'll move the goal posts and admit it is good but then go and claim it's the only example and therfore it doesn't count. Last edited by roondar; 17 August 2019 at 14:26. |
|||
17 August 2019, 13:07 | #89 | |
Natteravn
Join Date: Nov 2009
Location: Herford / Germany
Posts: 2,500
|
(mainly played Arcade games)
Quote:
I guess the European games market was different to the American and Japanese, where consoles dominated. European gamers saw the numerous advantages in having a full computer. And many of them became creative with it. I loved the variety on the Amiga and most Arcade games quickly became boring to me, while I could play "Player Manager" or "Civilization" for months. Never felt the need for a games console in my whole life... |
|
17 August 2019, 15:48 | #90 | |
Warhasneverbeensomuchfun
Join Date: Jun 2001
Location: Rio de Janeiro / Brazil
Age: 41
Posts: 3,450
|
Quote:
Then comes Speedball 2. GREAT, fun game to play with lots of nuances on gameplay. Then comes The Chaos Engine, another very well designed game which is great to play. Then comes The Chaos Engine 2: Huge piece of crap. And later they try to remake Speedball 2 at least twice. Both times it's, again, awful. When I say "Got the game right by dumb luck" I am always thinking about Bitmap Brothers. But when I look at the dev team at Team 17, it's hard to not think the same thing. In the other hand... yeah, its unfair from me to put everyone on the same bag. I believe Factor 5 did know and care how to do good games, as obviously Sensible Software too (Even though I am really hugely dissapointed with Jon Hare and the Sociable Soccer debacle). But I loved that someone mentioned Metroid. I don't think there's anything like it on Amiga. I only got to play Metroid games recently (I never had any nintendo systems back at the day), and they are incredibly good for their time. Super Metroid was really lots of fun to play (even if I had some issues with its control scheme)... it's something the Amiga could handle very easily... I find it curious because there are actually lots of european games that are, like, open and about exploration and finding out things, yet I can't name one that's as well designed and polished as Super Metroid. |
|
17 August 2019, 16:42 | #91 | |||
Registered User
Join Date: Jul 2015
Location: The Netherlands
Posts: 3,411
|
Quote:
Quote:
Quote:
It is indeed curious no one else has tried for a 'Metroidvania' yet on the Amiga (past or present). Should fit the hardware quite well, actually - you could even do nifty things like having levels that don't use standard 16x16 tiles as the individual rooms/areas in Metroid tend to be quite small so it can probably all fit without the need for wrapping displays. The C64 did get a few over the last few years, but as yet, nothing on the Amiga. |
|||
17 August 2019, 16:55 | #92 |
Warhasneverbeensomuchfun
Join Date: Jun 2001
Location: Rio de Janeiro / Brazil
Age: 41
Posts: 3,450
|
It's a huge work to do something like Metroid for Amiga as a spare-time-hobby job. I've been thinking about something like that with flicker-screens on Amiga... but the work would be huge really.... and I still have lots of other stuff I would like to do on the machine
Nah, I agree with you. I was unfair. DMA Design made great games too. Those guys whoe made The Killing Game Show and Geometry Wars also did some great games (Forgot their name), and yeah, we could keep listing a few more. Yes, my statement before was kinda unfair indeed. |
19 August 2019, 10:37 | #93 |
Registered User
Join Date: Jun 2016
Location: UK
Posts: 428
|
|
19 August 2019, 11:07 | #94 |
Phone Homer
Join Date: Jun 2006
Location: 5150
Posts: 5,775
|
Bitmap Brothers games are shit I like brainless button bashing
|
20 August 2019, 22:57 | #95 | |
Moderator
Join Date: Nov 2004
Location: Eksjö / Sweden
Posts: 5,604
|
Quote:
But Amiga games are ofc still highly relevant if they follow the topic: Fast Collision Detection. And so are techniques from any platform that can be implemented on Amiga (while being fast, ofc). |
|
20 August 2019, 23:25 | #96 |
Registered User
Join Date: Jul 2015
Location: The Netherlands
Posts: 3,411
|
Well, I've not actually done any tests here but the one remaining 'out there' idea I have would be to process objects in groups that act as one bigger object - until the grouped entity hits anything and only then look into what entity in the group collided. (effectively this is bounding box collisions where some objects are grouped into a single, bigger box and only 'separated' when a collision with the big box is detected)
This would work best if the object tested is not a grouped one but rather a separate one and the objects tested against are grouped. In theory, this should allow the number of tests to go down because you're only testing against bigger groups. However, adding objects to the group naturally also takes time to do. And naturally, this is only useful if objects can be reliably grouped together over multiple frames. A game that has all of it's objects moving freely will probably not work for this. As such, I'd say such an approach might work best if you can 'pre-bake' the collision groups and use some sort of group-anchor as the point where the big box is 'drawn'. I'm not certain it'd actually save time though, it probably rather depends on object spread on screen. Other than that, your initial points are still the best overall idea: reduce the number of objects that need to be tested by using knowledge about the objects to exclude them from tests. |
24 May 2020, 17:03 | #97 |
Registered User
Join Date: Aug 2018
Location: Untergrund/Germany
Posts: 408
|
Another option is to use morton codes. This works very fast, and can be easily combined with position fuzzing to test also objects outside the current morton slot.
I used that approach in my game "Tiny Galaga". There i had to test 8 player bullets with 32 enemies and the player with 32 enemies and 8 enemy bullets (so over 300 comparisons in worst case). Axis sorting was not possible because enemies could be clustered on X and Y axis. |
24 May 2020, 18:25 | #98 | |
Zone Friend
Join Date: Jan 2005
Location: italy
Age: 46
Posts: 244
|
Quote:
very interesting tip the Morton Code ,thanks |
|
24 May 2020, 23:56 | #99 |
Registered User
Join Date: Oct 2015
Location: Landsberg / Germany
Posts: 526
|
Never heard about the "morton code" before, so had to google it. Pretty interesting concept. Not sure in how far it can be faster than simply comparing coords: If I understood it right you attach an quadrant attribute to each moving object, and simply compare if two objects move within the same quadrant, and just in case of a hit init more sophisticated methods of collision detection. But you still have to update this attribute each frame for each object, no?
|
25 May 2020, 22:16 | #100 | |
Registered User
Join Date: Aug 2018
Location: Untergrund/Germany
Posts: 408
|
Quote:
You could call it a 1-dimensional spatial hashing method. Anything that moves would need to update it's morton codes in the data structure. You can store the pointers to the objects sorted in a linked list or in fixed sized bins. For simplicity i used fixed sized bins with a counter. For 32 enemies it took about 8% CPU time (in my 'Tiny Galaga'). For potential collision testing it's then only a lookup into the correct bin. Testing was then much faster then building the bins. In general this method does only pay off if you need a large amount of tests. If you have only a few dotzend tests or widly different AABB sizes then brute force will win. |
|
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
Thread Tools | |
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Games with collision detection disabling cheat? | alkis21 | Retrogaming General Discussion | 9 | 03 January 2018 04:54 |
Fast(est) method for bounding box collisions? | MickGyver | Coders. Blitz Basic | 2 | 02 October 2017 21:17 |
Possible problem with Collision detection ! | amilo3438 | support.WinUAE | 5 | 11 January 2017 18:35 |
Collision Detection | sandruzzo | Coders. General | 5 | 10 June 2016 12:50 |
Collision detection: 2 doubts | nandius_c | Coders. Asm / Hardware | 6 | 30 November 2014 00:53 |
|
|