This page is a mirror of Tepples' nesdev forum mirror (URL TBD).
Last updated on Oct-18-2019 Download

The essence of great games on the NES

The essence of great games on the NES
by on (#44710)
In this post, bunnyboy wrote:
SMB shows a good game can be basic hardware, but SMB3 shows a good game can be more advanced hardware too.


I think both SMB and SMB3 are not only good, but revolutionary games.

The difference is, in order for SMB3 to exist, there had to be SMB first, adressing exactly my point: core idea and playability are the primary factors which make a game great. Especially on a retro system like the NES, where you simply can not fool your audience with graphics and sound.

If someone feels a good game NEEDS a complex mapper, there are lots of examples which clearly show that this is not correct. I personally think that if you can't design a fun game in 8K VROM and 32K PRG ROM, you most probably won't succeed with the best mapper in the world.

I mean, if people don't feel any fun in trying to achieve maximum effect with minimal hardware, why are they coding for the NES at all? Why not taking the limitations as a creative opportunity?

by on (#44716)
Some people think the fact that it's on the NES is enough of a limit. I kind of feel that way too. As long as you can stick it in the same machine without modifying it (with the exception of disabling the lockout chip), and it outputs good graphics, sound, and gameplay, do whatever you want. The NES itself has limits that no mapper/board can eliminate.

As for me, I am trying to achieve -really- great visuals and good sound. I want parallax effects, polygonal sequences, and other cool things which really are only achievable with CHR RAM. CHR RAM isn't exactly some NES-defying mapper specific cheat or anything. I also want people to save their games, so SRAM is a must. Other than that, just a good chunk of plain old PRG ROM will do. That's not exactly over-the-top. I'm still abiding by the 8 sprites per scanline, 64 sprites on screen, 2x2 attribute blocks, etc. rules that pretty much any other game has to stick to as well.

I do agree that you don't need a complex mapper to make a fun game. Sometimes, there just isn't enough to add as much complexity/variety in graphics or level layouts as one would like. In this instance, choosing a more powerful mapper is good.

by on (#44719)
bunnyboy wrote:
But most people aren't looking for a super mapper, they are looking for something around MMC3 level.


That's what I meant. If we could have a MMC3/VRCIV/VI clone, that would be nice. I would stop cracking up game and just use that instead.

I don't want to "improve" the graphics, just want to be able to have more space and more functionality to be able to make some fan game similar to the platformer in the end of the nes era. Which mean some split screen, some extra counter if the need arise. But it doesn't mean that I will use all the features: it just good to know that I have access to them if I feel like it.

I have no interest in a one screen game with almost no artwork at all, I always personally hated those ones. I started to like nes games when stories and side scrolling elements were added.

by on (#44723)
Banshaku wrote:
I don't want to "improve" the graphics, just want to be able to have more space and more functionality to be able to make some fan game similar to the platformer in the end of the nes era.

Exactly. We don't want to go all the way back to...
Quote:
a one screen game with almost no artwork at all

We want to pick it up where developers back then left off. When the NES died commercially, MMC3-level mappers were the standard, it's just fair that we have that kind of functionality available today without having to butcher old carts.

Having features available can't hurt, because nobody will be forced to use them. People will forever be able to continue making NROM games, but it's just fair that the ambitious fellows have a chance to play with more advanced features.

by on (#44725)
tokumaru wrote:
We want to pick it up where developers back then left off. When the NES died commercially, MMC3-level mappers were the standard, it's just fair that we have that kind of functionality available today without having to butcher old carts.


This explain my motive even better than what I said. I'm glad to see that someone agree with me :wink:

tokumaru wrote:
Having features available can't hurt, because nobody will be forced to use them. People will forever be able to continue making NROM games, but it's just fair that the ambitious fellows have a chance to play with more advanced features.


Right on. Once I can make something interesting with a MMC3 mapper, I wouldn't mind to see how much we could push the nes further with a more advanced mapper but right now I don't have that interest.

by on (#44737)
tokumaru wrote:
Banshaku wrote:
I don't want to "improve" the graphics, just want to be able to have more space and more functionality to be able to make some fan game similar to the platformer in the end of the nes era.

Exactly. We don't want to go all the way back to...
Quote:
a one screen game with almost no artwork at all



Does this apply to SMB? Excitebike?

Quote:
We want to pick it up where developers back then left off.


If that would be the case, your next game AT LEAST would have to match the artwork, gameplay and fun of the best NES games released. Like SMB3 or Kirby's Adventure for example.

Quote:
When the NES died commercially, MMC3-level mappers were the standard, it's just fair that we have that kind of functionality available today without having to butcher old carts.


And perhaps that's exactly the problem: you just want to go the direct route and skip everything which experienced NES developers had to work with before, and yet still made fantastic games while they were at it. Or in other words: starting with MMC3 will automatically make you more experienced and skilled than the people who designed SMB, Contra and Zelda.

Quote:
Having features available can't hurt, because nobody will be forced to use them. People will forever be able to continue making NROM games, but it's just fair that the ambitious fellows have a chance to play with more advanced features.


Ambitious developers most certainly don't rely on mappers. They work on the gameplay until it's perfected.

That, and only that, was the reason the NES was successful. From a technical standpoint, the NES never was impressive. It was a lot weaker than other systems at the time, like the Sega Master System. The Sega system however, did never take off because it lacked the killer games the NES had.

Like I said, people gravely underestimate how much work went into the DESIGN of those games. Adjusting the game mechanics until they play perfectly.

That's exactly why I have to chuckle when people here talk about impressive graphics, streaming megabytes of PCM sound, building all kinds of gadgets into the cartridge etc, because it's a 2009 mentality struggling with the realities of a 1983 hardware design. You guys picked the NES because it is cool, yet secretly you want to work on something else, because you feel intimidated by the restrictions.

The NES is an old and weak system. So seize the opportunity, free yourself from the burden of technology, and try to find the essence of great games. Just what the developers did back in the day.

THIS is MUCH HARDER than designing any kind of super mapper. But also far more impressive. And I know most of you are now being pissed at me for pointing that out, because you know it is true.

by on (#44738)
I agree with 6502 freak entierely. If you want to experience what the developpers back then did, you must first do some smaller game with low ROM/RAM/Mapper, and then make sequels with higher features, and eventually switch to a better system. Starting at MMC3 level isn't "Bad", but not what the developpers did back then. Altough it's true that even by starting with NROM, you could say you skip the atari crap, so this is agruable. I still feel presonally more like starting to do 1986-area games and then doing 1990-area games only once I'm able to do the former.

by on (#44739)
What I am trying to hint at is the following schizophrenic situation: thinking first in terms of technology and the gameplay comes second. People will buy my game because it looks and sounds more realistic than the previous games before. This is the 2009 way of doing games on modern systems.

The schizophrenic part is in applying this way of thinking to an obsolete and inherently weak hardware system.

If you want to pick up where developers left, code on the SNES. You don't like the slow CPU of the SNES, because you can not realize your brilliant game ideas?? Pick a Playstation 1. Damn, if I just had more texture memory, I could make WAYYY much better games... and so on.

I think it totally avoids the spirit and feeling why the NES is a great system. It is ONLY because of its games. They may look primitive and dated, but the best of them are all-time classics, because the designers constantly confronted themselves with the question: what MAKES a great game? How do I motivate the player? How can I make him/her come back to get some more?

They HAD to do it. No safety net. No way of fooling the player with graphics and sound.

If you realize that, you don't need a fancy mapper. Heck, you don't even need a system as advanced as the NES. ;)

by on (#44741)
I don't know where your trying to get to with your point but you know, if Celius want to use X mapper because he feel like it, it's raining with some wind coming from the north-east or I don't know, the sky is green with purple dots that day, I guess it his own business after all :P

It's pretty obvious from the get go that no mapper you select will make a better game and I'm sure that everybody knows it. Just look at the popularity of the Wii as an example. Not the best hardware but the game sells (only Nintendo ones and third party get shafted but that another story...).

But what if you goal is not to make games but just experiment with the hardware? Some people like to fool around with those things you know.

As for my self, I never selected to program on the nes because it was cool, never came to my mind this way. I'm just a programmer that decided to fool around with a machine that I have been using as a kid and still remember about it. Always wanted to know how it worked, what kind of hardware they usede etc. I may or may never finish a game for it, who knows. But for now, I'm just happy to fool around with it.

If your goal is to make a great game to sell then you chose the wrong platform in the first place. Go try something on a Ipod touch and see what you can do with motion sensors and a touch screen then.

by on (#44742)
Banshaku wrote:
If your goal is to make a great game to sell then you chose the wrong platform in the first place.


Now how does it sound if I remove your carefully chosen "to sell" from your statement?

Quote:
Go try something on a Ipod touch and see what you can do with motion sensors and a touch screen then.


And again, do I need a touch screen and motion sensors to make a great game?

I guess I have a totally different POV on the NES than some people here. I always thought the everlasting appeal of the NES were the carefully designed games.

But it's kind of senseless to discuss this any further, because in the end, everyone chooses his/her own way of spending his/her free time. My posts are only adressed to people who think that a fancy mapper will make them create better homebrew games.

by on (#44743)
6502freak wrote:
If that would be the case, your next game AT LEAST would have to match the artwork, gameplay and fun of the best NES games released. Like SMB3 or Kirby's Adventure for example.

Ah, Kirby 's Adventure... I hardly beleived it was a NES game, only the sound really tells it's a NES game

by on (#44744)
6502freak wrote:
Now how does it sound if I remove your carefully chosen "to sell" from your statement?


I Didn't carefully chose that word for any specific reason.

6502freak wrote:
And again, do I need a touch screen and motion sensors to make a great game?


My point it that you had a limited interface to interact with the hardware so you need to use your creativity to do something interesting. Same thing for the Wii. The nes has similar limitation but on a hardware based level, not input device level.

6502freak wrote:
I guess I have a totally different POV on the NES than some people here. I always thought the everlasting appeal of the NES were the carefully designed games.


I guess every one has a right to their own opinion and we cannot change that.

I won't comment any more in this thread since it seems to just add oil on an open fire. I don't know your purpose but It almost feels like your main goal is to provoke people for some reason (?). What are you looking for? Can we move back to the original topic?

If you don't agree to this topic you can start a new one and I'm sure many people will be more than happy to discuss about it. This thread is getting way too much off topic now.

by on (#44745)
Banshaku wrote:
If your goal is to make a great game to sell then you chose the wrong platform in the first place.

I'm almost willing to suggest that one chose the wrong industry. Every open (PC) or semi-open (iPod Touch) platform is flooded with free games, and the production values expected of a paid game (even one with NES graphics like Mega Man 9) usually need a team of at least half a dozen people.

by on (#44746)
The best way to enhance a NES cartridge is using an FPGA, map the dual port RAM to the PPU address and data bus so you can access it any time, and throw in a small blitter which can perform logical and arithmetical operations on the RAM. You could even generate new grahics modes that way, totally ignoring the PPU addressing scheme and just constantly feeding your own data to it, which is being displayed. This way, you can easily have things like 32 column vertical scrolling.

Alternatively, pick a fast and cheap microcontroller, program it with your game code, and make it interact with the 6502 CPU by using a very simple FIFO command interface, passing parameters between the 2 chips. You may even use it to count PPU address changes and generate NMI's for the 6502. That way, the 6502 is only a display data mover, which serves the PPU to feed data and change registers.

Alternatively, I think you could easily implement VROM/PRG ROM Mapping and a rastercounter in a 64 macrocell CPLD.

The big question is now: who would buy such a hardware? For what purpose?

by on (#44747)
6502freak wrote:
Alternatively, I think you could easily implement VROM/PRG ROM Mapping and a rastercounter in a 64 macrocell CPLD.

The big question is now: who would buy such a hardware? For what purpose?


Geeks like us that like to experiment with a dead platform like the nes? :P

by on (#44752)
I don't know what some of you are thinking, but yes a MMC3 class homebrew board is needed. Don't throw shit out there about how great some game was and it only used a UNROM or MMC1 board. What if someone is designing a game and it NEEDS 1KB CHR switching? What if they need more than a single 16K swappable bank? Who are you to tell them they don't need it? I think it's stupid to assume that everyone should just be happy with discrete mappers and the MMC1 clone. They are great choices to have, but some people will want something more capable. What if tomorrow someone popped up and they'd created some amazing game using MMC1 or NROM and they needed a MMC3 class mapper for the next game? Are you just gonna tell them you know best again?

Let people develop their ideas with whatever mapper they want. I think it's pretty clear that something a bit more powerful than the MMC1 clone will be needed by someone. Just my opinion that people shouldn't be telling other people what mapper they have to use.

by on (#44754)
6502freak wrote:
Does this apply to SMB? Excitebike?

In my opinion, those are the best NROM games, but they still lack on the graphics department.

Quote:
If that would be the case, your next game AT LEAST would have to match the artwork, gameplay and fun of the best NES games released. Like SMB3 or Kirby's Adventure for example.

I certainly hope so. Although I don't think much of SMB3, but Kirby is certainly a nice goal to aim for.

Quote:
And perhaps that's exactly the problem: you just want to go the direct route and skip everything which experienced NES developers had to work with before, and yet still made fantastic games while they were at it.

Yes I want to skip that. Do Playstation 3 programmers have to release Pong for it before moving on to Devil May Cry 4? They probably coded simpler stuff before, yes, but it didn't have to be on the PS3. I had my fair share of learning with simpler projects on other platforms, but now I picked the NES to host my "masterpiece". And we probably have different views on what "fantastic" means. I don't think SMB3 is fantstic at all. I think it was made by rushed programmers that didn't even get rid of scrolling artifacts properly. Kirby has the same artifacts.

Quote:
Or in other words: starting with MMC3 will automatically make you more experienced and skilled than the people who designed SMB, Contra and Zelda.

No, but previous experience with other platforms and languages could have put me at that level, and I'm ready to move forward, not backwards.

Quote:
That's exactly why I have to chuckle when people here talk about impressive graphics, streaming megabytes of PCM sound, building all kinds of gadgets into the cartridge etc, because it's a 2009 mentality struggling with the realities of a 1983 hardware design. You guys picked the NES because it is cool, yet secretly you want to work on something else, because you feel intimidated by the restrictions.

Not at all. The hardware is from 1983, but it did not remain static. The NES was upgraded by mappers within the cartridges. All we want is access to such an upgraded NES.

Quote:
The NES is an old and weak system. So seize the opportunity, free yourself from the burden of technology, and try to find the essence of great games. Just what the developers did back in the day.

I didn't fall in love with the NES because of Gyromite, I fell in love with it because of Bucky O'Hare, which is as good as the next generation's games. "Standard" is not fun, it's plain and boring. Pushing the system to its limits is the fun part.

Quote:
THIS is MUCH HARDER than designing any kind of super mapper. But also far more impressive. And I know most of you are now being pissed at me for pointing that out, because you know it is true.

Are you listening? Nobody wants a super duper MMC5 clone, we just want good bankswitching, some extra RAM and a decent scanline counter. If you look at the whole library of NES games, a great deal of them has those features, so there is no way you can tell me they take away the "spirit of the NES".

And what you say is not true. If we just wanted to make games without limitations, we'd code them for the PC where it'd be very hard to find limitations. We just like the NES and want to make the best out of it, and prove how cool it can be.

6502freak wrote:
I always thought the everlasting appeal of the NES were the carefully designed games.

You do know that "carefully designed games" and "good mapper hardware" are not mutually exclusive, right? I agree with you 100% that gameplay is way more important than visuals, but if we can have both, why not have the tools available? A crappy game is a crappy game, regardless of the graphics, and the same is true for good games.

by on (#44758)
Quote:
Nobody wants a super duper MMC5 clone

In fact me I'd like one, but maybe not immediately right now. :wink:

Well I guess I'm moving to Playstation dev so that this mapper crap flaming will not be there any longer, and I won't be told that I'm supposed to work with a limited system all the time and that I should switch to a more powerfull system if I'm unhappy.
Unfortunately I don't see any PS-dev communauty anywhere. :cry:

And back to NES dev, yes nobody should impose a mapper to someone else, so this whole thread has no point.

In summary from what people are saying (I'm not saying I agree) :

- If you want to make a game and want to limit you to what developpers did before use a common mapper the smaller possible to suit your project in this order : NROM, CNROM, GNROM, UNROM, AOROM, MMC1, MMC3, MMC5

- If you want to make a game that is freely downloadable care about emulator compatibility but want to have a powerfull mapper, use MMC5, FME-7, N106 or anything like that.

- If you want to make a game that will be on cartridge support only, make your own CPLD mapper and you don't mind emulator compatibilty (in practice it will be VERY hard to developp such a game without testing on an emulator, so you'd want to make it like a mapper mentionned above but slightly modified to fit a CPLD optimally)

- If you still feel too limited, move to another platform for developping games. You could still do a pseudo-NES game without getting it running on any NES hardware, like Mega Man 9.

by on (#44760)
tokumaru wrote:
I didn't fall in love with the NES because of Gyromite, I fell in love with it because of Bucky O'Hare, which is as good as the next generation's games. "Standard" is not fun, it's plain and boring. Pushing the system to its limits is the fun part.


That basically sums up my opinion too. I do plan to make an NROM platformer just to test my abilities in terms of consolidation and limits, and still have relatively good graphics, but I didn't fall in love with the NES because of 1 screen games that you play for 5 minutes and then get bored with (e.g. Golf or some shit), or even because of SMB1. Games like Battletoads, Bucky O'Hare, Batman: Return of the Joker, Final Fantasy 3, Castlevania 3, etc. are beautiful works of art on a very limited -system-. So long as the NES and a cartridge can work together to make a great game with great graphics and sound, that's enough. It's showing that it can happen that's fun; not necessarily making it happen within 32k of PRG and 8k of CHR.

by on (#44765)
6502freak wrote:
you just want to go the direct route and skip everything which experienced NES developers had to work with before, and yet still made fantastic games while they were at it. Or in other words: starting with MMC3 will automatically make you more experienced and skilled than the people who designed SMB, Contra and Zelda.

I agree with tokumaru's citation of Pong. "You should master all bytes $0000-$1FFF of the Atari 2600 and make a game better than Pitfall! before you even touch register $2000 of an NES." How does that sound?

6502freak wrote:
Ambitious developers most certainly don't rely on mappers. They work on the gameplay until it's perfected.

OK, how would I make a sim like Nobunaga's Ambition or Harvest Moon without a battery?

Bregalad wrote:
Starting at MMC3 level isn't "Bad", but not what the developpers did back then.

No, they started with Atari 2600, Odyssey 2, Apple II, Intellivision, ColecoVision, Commodore 64, and ZX Spectrum before the NES. Should everyone else have to follow the same path?

Mapper luddism will only get you so far. Let's please get back to a feasibility analysis of putting MMC3 in a CPLD.

by on (#44768)
tepples wrote:
Let's please get back to a feasibility analysis of putting FME-7 in a CPLD.

Fixed.

by on (#44769)
tepples wrote:
I agree with tokumaru's citation of Pong. "You should master all bytes $0000-$1FFF of the Atari 2600 and make a game better than Pitfall! before you even touch register $2000 of an NES." How does that sound?


haha. My Pitfall! clone used the mmc1. If you take out the custom mode, high score table, and optimize the chr usage, it'd be nrom. :D

A homebrew mapper like the mmc3 with more flexible PRG, CHR bank sizes/setup and supports SRAM would be nice. Scanline counter would be optional.

by on (#44771)
never-obsolete wrote:
tepples wrote:
I agree with tokumaru's citation of Pong. "You should master all bytes $0000-$1FFF of the Atari 2600 and make a game better than Pitfall! before you even touch register $2000 of an NES." How does that sound?


haha. My Pitfall! clone used the mmc1. If you take out the custom mode, high score table, and optimize the chr usage, it'd be nrom. :D

A homebrew mapper like the mmc3 with more flexible PRG, CHR bank sizes/setup and supports SRAM would be nice. Scanline counter would be optional.


What you describe is basically the FME-7 aka Sunsoft 5b. I like the FME-7, though it would be nice if the IRQ wasn't only a CPU Cycle timer, but also had a scanline PPU counting mode. Didn't somehow here suggest that you could use A13 rather than A12 and avoid the pattern table limitation with 8x16 sprites that MMC3 has?

by on (#44773)
tepples wrote:
I agree with tokumaru's citation of Pong. "You should master all bytes $0000-$1FFF of the Atari 2600 and make a game better than Pitfall! before you even touch register $2000 of an NES." How does that sound?


Sounds like a great idea. Atari 2600 coding is the perfect symbiosis of programming and game design. Because on this architecture, both aspects heavily influence each other. If you are a bad coder, you won't even get one single sprite properly displayed on the VCS. ;)

Quote:
6502freak wrote:
Ambitious developers most certainly don't rely on mappers. They work on the gameplay until it's perfected.

OK, how would I make a sim like Nobunaga's Ambition or Harvest Moon without a battery?


The better question is: is your sim worth the extra cost for having a battery?

Quote:
Bregalad wrote:
Starting at MMC3 level isn't "Bad", but not what the developpers did back then.

No, they started with Atari 2600, Odyssey 2, Apple II, Intellivision, ColecoVision, Commodore 64, and ZX Spectrum before the NES. Should everyone else have to follow the same path?


No. Just skip it, and start doing some stuff on the PS3. At least you have access to the Cell processor under Linux.

Quote:
Mapper luddism will only get you so far. Let's please get back to a feasibility analysis of putting MMC3 in a CPLD.


I personally think you can do the following on an inexpensive CPLD with 64 macrocells:

- PRG ROM banking at 16K for 512K ROMS
- RASTER counter, which works as a one-shot timer. No fancy extra registers, just write the value and BAM.
- No VROM. Instead, use a cheap 32KB SRAM as VRAM the following way:
* 4K PPU Banking for EVERYTHING, dividing the PPU memory space in 4K segments which can contain one of the 8 VRAM banks. Which means, you can have multiple 4-screen nametables!

This design could fit into an Altera 3064 CPLD in PLCC44 package. Perhaps maybe a little more (but just a little).

by on (#44781)
6502freak wrote:
Bregalad wrote:
Starting at MMC3 level isn't "Bad", but not what the developpers did back then.


Quote:
No, they started with Atari 2600, Odyssey 2, Apple II, Intellivision, ColecoVision, Commodore 64, and ZX Spectrum before the NES. Should everyone else have to follow the same path?


No. Just skip it, and start doing some stuff on the PS3. At least you have access to the Cell processor under Linux.


By the way... Not that I want to start a fight or anything but before you start to make a lot of ruckus in a thread on how we should all go back to NROM and design is everything you should, at the least, eat your own dog food?

It looks to me like you like the MMC3 too...

Edit: And designing "a typical jump'n'run game"..

by on (#44789)
Banshaku wrote:
By the way... Not that I want to start a fight or anything but before you start to make a lot of ruckus in a thread on how we should all go back to NROM and design is everything you should, at the least, eat your own dog food?

It looks to me like you like the MMC3 too...

Edit: And designing "a typical jump'n'run game"..


So that's being on-topic for you... ;)

Yes, I was experimenting with the MMC3. And? How does that invalidate ANY of the things I have been saying in my previous threads? I just reiterate them for you:

- A great mapper won't help you much in making a great game
- People unable to make a fun game in NROM won't probably succeed with a MMC3 or better
- The NES was popular because of it's breakthrough games. As a hardware platform, it never was any impressive, even back in 1983/85

These are my claims here, and you guys are just being pissed off right now because you know they are true. That's exactly the reason you guys constantly feel the need to put words in my mouth I never claimed. ;)

Besides, I figured out that what I did with the MMC3 could have easily been scaled down to a much more simple mapper. Many shitty games have been developed on the MMC3, and many of the best games on it could have been realised on a more simple mapper. Not because they exactly wasted the MMC3, but the fact that they are being great games because of their game design and idea.

One of the VERY FEW games which REALLY use the MMC3 effciently for gameplay is Bucky O' Hare. For example, in the stages which have the screen full of horitontal scrolling platforms.

Do I have to mention that Bucky O Hare is some kind of Konami's swan song for the NES, after almost a decade of NES experience and having released many great games on far simple mappers?

Great, now you have seduced me to go back off-topic, after pointing the moral finger at me for derailing the thread. Naughty boy. ;)

by on (#44791)
6502freak wrote:
These are my claims here, and you guys are just being pissed off right now because you know they are true. That's exactly the reason you guys constantly feel the need to put words in my mouth I never claimed. ;)


I don't think that anybody here are piss off because of that fact. Great games can be made with or without a mapper and nobody is arguing against that.

The reason that many people are adding to this off topic argument is maybe more the tone that was used in the message regarding it. To me, the way it was said was strong and didn't feel appropriate. This is more what I feel made people react than anything. People react strongly on the nesdev forum for these kind of message.

I think we are heading for a topic split and maybe this message will be the one. What maybe you want to argument is more about what make a great game? This is very subjective topic and everybody have their own opinion about it.

For myself, if you come with a NROM 1 screen game or atari 2600 one and you will lose my interest right away. Make some kind of platformer with a story, add a good sountrack with many songs with some nice graphics AND game play and you will start to get me interested (Ninja gaiden would be an example).

I guess discussing about this subject would be more constructive than "a mapper doesn't make a game". Maybe some good argumentation could come out of it.

Regarding my previous message, since the last message you wrote still felt not appropriate, I wrote very fast what I could call a "childish" message regarding the fact of showing by example make your argument stronger. I should have not posted since it didn't bring must to that thread and I'm sorry about that.

So, ready to talk about what makes a good game then? :lol:

by on (#44792)
THIS IS SPARTA!!!

by on (#44793)
Okay, I just figured something.
The "problem" is that with no mapper at all, the NES is really very limited and that about 90% of NES games (I just throw this, don't take that nunber as accurate) are using a mapper.

On platforms like the SNES where no mapper is the standard, everyone will be using no mapper unless they are doing complicated 3D stuff, so there is no real debate, a game will be using a mapper only if it would be impossible to make it without it.

For the NES, this is really different because most popular games DOES use a mapper. Also, the mapper used are very varied and there is no real standard, only a list of standard. Taking one over another means sacrifices, and this is highly agruable. It is understandable that one do not want to make sacrifices in order to keep to a simpler mapper when there is no real reason to do so other than challenge.

Altough, I'd say that you'd want to follow the rule "use the simplest mapper you need". That means that if you are not going to use most features of a mapper, you may as well not use that mapper and use a simpler instead. However, many effects can be achevied simply with a complex mapper or in a complex way on a simple mapper (the lack of a IRQ mapper will imply more complex sprite-zero hits and timed codes). So whenever you want to complicate your code or you mapper is up to personal choose, and nobody can answer to anyone whether it's good to complicate the software or the hardware.

When it comes to programming on a powerfull system, honnestly, if I had the tools and the skill to make a PS3 game that looks pretty much like other PS3 games, I would do that instead and not bother with nesdev. Unfortunatlely I'm a bad graphist and I throught that developping for a system with "bad" graphics would make my chances to build a game at a similar level to others on the same platform higher. By "similar to others on the same platform", it implicitely says that it also use similar hardware, so no super mapper that is able to do aditiooonal things never seen before on the NES.

Now people could be NESdeving for other reasons that I am, so it's some kind of personnal matter.

by on (#44804)
6502freak wrote:
- People unable to make a fun game in NROM won't probably succeed with a MMC3 or better

Besides, I figured out that what I did with the MMC3 could have easily been scaled down to a much more simple mapper. Many shitty games have been developed on the MMC3, and many of the best games on it could have been realised on a more simple mapper. Not because they exactly wasted the MMC3, but the fact that they are being great games because of their game design and idea.


Your first point is like most of your points, just some generalization. Making a fun game in NROM for one is harder to do than with a mapper because when you're limited to 32K PRG and 8K of CHR that severely limits what kind of game you can do. Don't bring up Super Mario Brothers because it is an exceptional game written by a professional company and god knows how long they spent making it. How about you mention some other NROM games that are as impressive? Or is that the only one you got?

I agree with part of what you are saying in the sense that someone probably shouldn't use MMC5 unless they really need it. Using the most common mapper for your needs certainly is the way to go. But it's not up to you to determine what they need. MMC3 is certainly what most people would prefer as the scanline counter is very desirable even if all you are doing is a status bar and maybe a few other effects with it. It's also very helpful if you need to bank smaller amounts of CHR for animation of BG or Sprites.

Your point about scaling down games makes no sense either. Ofcourse you could scale down just about anything but that's just stupid unless you have a good reason to do so. You know I think most of us wouldn't even care about Castlevania 3 if they had "scaled down" the game to do it as CNROM.

Basically I think you're acting elitist and putting people down. I'd certainly find your comments offensive if I was working on a NES project with the MMC3.

by on (#44825)
MottZilla wrote:
Your first point is like most of your points, just some generalization.


It's a good way to teach me about generalizations by starting with a generalization yourself. ;)

Quote:
Making a fun game in NROM for one is harder to do than with a mapper because when you're limited to 32K PRG and 8K of CHR that severely limits what kind of game you can do.


Tell that an Atari 2600 homebrewer, who has even fewer resources to work with, yet still manages to put a fun game together.

Quote:
Don't bring up Super Mario Brothers because it is an exceptional game written by a professional company and god knows how long they spent making it. How about you mention some other NROM games that are as impressive? Or is that the only one you got?


How about start reading the posts you are responding to? ;)

Quote:
I agree with part of what you are saying in the sense that someone probably shouldn't use MMC5 unless they really need it.


Why is that so? I think people should have a right for free mapper choice! If I would be a MMC5 fan now, I would be offended by your statements!

Quote:
Using the most common mapper for your needs certainly is the way to go. But it's not up to you to determine what they need.


Are you telling me this to reassure yourself that I have no control over your mind?

Damn, and I thought I could make you all my slaves through a clever mix of mind control and charme... Guess I must have failed. :(

Quote:
MMC3 is certainly what most people would prefer as the scanline counter is very desirable even if all you are doing is a status bar and maybe a few other effects with it.


You know, on the SNES, you even have 4 seperate layers to stack on top of each other. Think how much you can accomplish with that!

Quote:
It's also very helpful if you need to bank smaller amounts of CHR for animation of BG or Sprites.


Yes, your sprites will animate much smoother. And you can even animate your clouds and flower tiles in the background...

You aren't quite exactly telling me any news here. But go ahead if it makes you feel better.

Quote:
Your point about scaling down games makes no sense either. Ofcourse you could scale down just about anything but that's just stupid unless you have a good reason to do so.


Lol, then why ARE YOU GUYS developing on the NES at all???? Working in 2009 on NES games is the ultimate way to scale down your ideas on a minimal hardware. The whole time, you are dealing with self-inflicted restrictions by choosing this kind of platform.

Now come on, and prove my point by saying: "But yeah, since the NES is sooo limited, we need to have good mappers to accomplish anything!"

Quote:
You know I think most of us wouldn't even care about Castlevania 3 if they had "scaled down" the game to do it as CNROM.


But you surely would agree with me that people cared a tiny little bit about Castlevania, right? THE game which started the whole franchise. THE essential game, to which Konami came back when they designed Castlevania 3 (NOT Simon's Quest)?

Sorry, but your "argumentation" is just hilarious.

Quote:
Basically I think you're acting elitist and putting people down.


And I think you are not elitist, yet still putting people down.

Quote:
I'd certainly find your comments offensive if I was working on a NES project with the MMC3.


And yet, that doesn't hold you back either to deliberately misread and ignore ALL my arguments I previously wrote. But it felt good to let that anger out, right? ;)

Anyone else who wants to "discuss" what makes a great game?

And btw, I commend the moderator for the creative choice of the thread topic, so that NOTHING will be misunderstood. ;)

by on (#44830)
6502freak wrote:
And btw, I commend the moderator for the creative choice of the thread topic, so that NOTHING will be misunderstood. ;)

I had to write something; it wouldn't let me split the topic with a blank subject. Feel free to edit your first post in this topic to change the subject.

by on (#44832)
tepples wrote:
6502freak wrote:
And btw, I commend the moderator for the creative choice of the thread topic, so that NOTHING will be misunderstood. ;)

I had to write something; it wouldn't let me split the topic with a blank subject. Feel free to edit your first post in this topic to change the subject.


I changed it now. Thanks.

by on (#44835)
hmm, I don't see anything wrong with using mappers to work on a game. Personally, I don't feel as though I'm ready for more advanced mappers right now. CNROM seems to suit my needs for the time being, but when I feel better about my coding (and maybe even plan things ahead of time a bit better), I'm sure I will go ahead and venture that way hehe

If someone wants to REALLY limit themselves, there is always the minigame compo. I feel like I pulled off a decently fun game for being limited to 4k, so anything is possible really. I mean hell, Memblers pulled off a 1023 byte game, and I'm not lying when I tell you that my son and I have a blast playing that game two-players haha Button mashing goodness ; )

So, the way I see it, if you want to challenge yourself, limit yourself to the extreme, otherwise, do as you please.

by on (#44836)
I don't care about Atari 2600 homebrew. The system is garbage. I don't consider any 2600 game to be "fun". Why don't you go hang out at a 2600 forum instead.

I'm not developing on the NES right now. But if I were it would be for the fun of making a NES game. You know, enjoyment. I'm not going to "prove your point" by saying whatever crap makes you feel smart. You're acting like a huge ass. And believe it or not no one here including me has to prove anything to you. Infact if you know better than all of us why don't you go start your own forum.

by on (#44838)
6502freak wrote:
Tell that an Atari 2600 homebrewer, who has even fewer resources to work with, yet still manages to put a fun game together.


There's a game as enjoyable/fun/good looking as say, Battletoads on Atari 2600?

I personally find the Atari to be so limited that games really aren't that fun on it. Pitfall! is a miracle (basically the equivalent to, if not superior than SMB1's condensation) but I don't enjoy it as much as I could. Honestly, if they had another 4k to work with that would be great.

And I don't agree with the argument that SMB1 is a special case and we can't replicate that. True, it's an extraordinary display of compression, but your muscles are capable of moving in a manner that would type that same code, so you can do something similar in terms of compression. But it's true, this is basically the only NROM game that did anything that impressive, and there comes a point where you can't do anything else.

I like to think, what would someone who has a NES really like to see? Someone who doesn't have an SNES or anything else; what would they like to see? Well, I know they don't want to see anything boring like Lunar Pool or the other NROM games; not even another SMB1. They want to see something with good graphics, sound, and gameplay. Something that's not so limited/repetitive. I can take this MMC3 board which allows for more PRG space and a scanline counter; that, I could really work with. The only thing is that it -has- to be for the NES. The rest is up to me.

Most of us do NESDev to see what the NES can do. That's the one thing that we don't expand on/change. We all love our NES, and we don't want to move away from the system itself. That's why we come up with all these things to improve the quality of games we make on the NES. And it's a fact: certain things cannot be done with NROM that can be done with other mappers. A powerful mapper does improve the possibilities in terms of overall quality for a NES game. It does not improve the quality by itself.

I do find this argument rather thought-provoking. It does make sense to ask: why develop with super mappers when you can just develop for a more powerful system? Doesn't that take the fun/point away? But the answer is: because we love the NES.

by on (#44843)
The most impressive CNROM games are Dragon Quest 1, and Egypt. That's about as close to NROM as you can get without being NROM. (Don't bother mentioning those games that are NROM with "bankswitching" for copy protection)

by on (#44846)
6502freak wrote:
Ambitious developers most certainly don't rely on mappers. They work on the gameplay until it's perfected. [...] Like I said, people gravely underestimate how much work went into the DESIGN of those games. Adjusting the game mechanics until they play perfectly.

It's very similar with Apple, whose products distinguish themselves in the subtle refinements of their designs. Other companies make mp3 players that look similar, but subtle elements are left out. For a sidescroller, perhaps the single biggest thing that affects the feel of the game is the physics of the jump. Even within Nintendo games, a Mario jump is very different from a Samus jump (especially Super Metroid), where Mario is much more constrained by momentum.

by on (#44850)
blargg wrote:
For a sidescroller, perhaps the single biggest thing that affects the feel of the game is the physics of the jump.

But there's still no reason to prevent those that can program good jumps from also using background animations, more detailed sprites and an occasional parallax effect. A good community mapper would give homebrew programmers all the possibilities.

I think that the programming of a game is mostly mapper-independent. No mapper will do your physics for you, or the scrolling engine, or anything else. Sure, having extra RAM or ROM might have an impact on how you design the code, but the logic will be basically the same with or without a mapper. Mappers simply allow you to polish things up a bit, something that shouldn't be frowned upon. The NES is still doing all the work.

by on (#44856)
^^I think the limit is pretty much mmc5. At that point the nes is stil doing most of the work (mmc5 uses tricks to enhance this). After that point you're basicly taking over control.

by on (#44870)
Okay everyone seems on a NROM vs MMC5 debate here.
Personally most of my favourite NES games (Castlevania, Mega Man 1-2, Contra, Dragon Warrior, Final Fantasy, Battletoads) uses mapper 1, 2 or 7, which uses some mapper, but not some extremely powerful mapper and managed to be absolutely awesome IMO.

Sure I love Castlevania 3 and Just Breed which uses MMC5, and I like many games (SMB3 and Mega Man 3-6) using MMC3 pretty much, but they are sequels to games with simpler mappers.

But if I were to nominate a "super mapper" to be available, I'd definitely vote for a MMC5 clone. It has everything, all bankswitching modes thinkable, graphics extansions, large amounts of SRAM and a fast multiplier. Everyone seems to think about the MMC3 but I think it's over-rater, altough it can allow smaller CHR bankswitching than the MMC1, and have a crappy scanline counter, it's about all it can do better than it. If you use CHR-RAM you don't want to worry about the former, and the scanline counter is really limiting, but it's true that in some case it would still be better than sprite zero hits if you are having more than one split point on your screen during gameplay. Also it lacks MMC1's 32kb PRG bankswitching and one-screen mirroring, which could be essential.

Anyway it is pretty clear that the mapper have a strong influence on :
- Amount of different graphics
- Amount of levels and music present in the game
- Whether or not you can save your game
- The presence of on-boad SRAM can increase possibilities of the game engine

The following elements are not affected by the mapper :
- How detailed the graphics are drawn
- How good your levels are
- How good the music is
- How fun it is to play the game

Hironically, today the most advanced homebrew NES game ever made is NESnake 2 by Matrixz which uses NROM !! It has very great music, very good gameplay and is very fun. It does also have quite good graphics for a snaks game. It looks better than SMB in fact, but it doesn't make any sense to compare a snake game and a platformaer.

Know why so many of us can't release a really great NES game ? We really are thinking too hard instead of just throwing creative ideas straight in a programm ! Everyone on this thread is thinking way too hard. I want's until I came here, but now I'm thinking too hard instead of acting.

by on (#44871)
Bregalad wrote:
But if I were to nominate a "super mapper" to be available, I'd definitely vote for a MMC5 clone.

Do they even make 5 V CPLDs that will hold an MMC5 clone?

Quote:
Everyone seems to think about the MMC3 but I think it's over-rater, altough it can allow smaller CHR bankswitching than the MMC1, and have a crappy scanline counter, it's about all it can do better than it.

That and the MMC3's bankswitching of $C000-$DFFF, useful for sample playback.

Quote:
the [MMC3] scanline counter is really limiting, but it's true that in some case it would still be better than sprite zero hits if you are having more than one split point on your screen during gameplay.

Or if you don't want the complexity of having to ensure that a tile near the split point is opaque if you want a split point within a complex scrolling section.

Quote:
Also it lacks MMC1's 32kb PRG bankswitching and one-screen mirroring, which could be essential.

Essential for what? Multicarts?

Quote:
Anyway it is pretty clear that the mapper have a strong influence on :
- Amount of different graphics
- Amount of levels and music present in the game

Perhaps 6502freak's point is that few hobbyists have the dedication to make enough "different graphics" and "levels and music present in the game" to justify more than NROM.

Quote:
- Whether or not you can save your game

An eight-character password can reliably save 32 bits plus an 8-bit checksum. Perhaps 6502freak's point is that few hobbyists have the dedication to make a game that needs more than 32 persistent bits.

Quote:
The following elements are not affected by the mapper :
- How detailed the graphics are drawn

In the case of CHR RAM boards, some compression formats have an easier time with less complex tiles than with more complex tiles.

Quote:
- How good the music is

You need an MMC3 or FME-7 class mapper to use lots of samples.

Quote:
Know why so many of us can't release a really great NES game ? We really are thinking too hard instead of just throwing creative ideas straight in a programm ! Everyone on this thread is thinking way too hard. I want's until I came here, but now I'm thinking too hard instead of acting.

I think 6502freak's point is that a hobbyist's creative idea is likely to fit in 24,592 bytes.

by on (#44872)
Bregalad wrote:
Know why so many of us can't release a really great NES game ?

Lack of time, probably. I always see people saying how great NES programmers back at the time were, but there was nothing special about them. In fact, today we have more knowledge and better tools to make games better than they could. The big difference is that they were paid to do it, and invariably dedicated quite a few hours of each day to their projects. I have to squeeze a couple of hours at the end of the day whenever I can to work on my project, and that's why it's so hard to finish it.

IMO, the difference in complexity between NROM games and mapped games (that are similar in gameplay) is negligible. You'll just spend more time generating data for the project with more CHR and PRG, but getting the basic engine working shouldn't take longer. It's simply not true that games running on simpler mappers are easier to make. It can in fact be the opposite, such as in cases where more ROM and RAM would allow you to store data in a more friendly way for your code, and without it you have to use tricks to achieve the same goal (do more with less).

by on (#44873)
tokumaru wrote:
It can in fact be the opposite, such as in cases where more ROM and RAM would allow you to store data in a more friendly way for your code, and without it you have to use tricks to achieve the same goal (do more with less).


This is one thing. For example, I'm lucky if I can even put 1h per day on my project. These days, that hour is mostly taken for making my map editor and I can't wait that it can get out of the alpha stage. But once it will be stable enough to generate data, I may not have the time to make some crazy code to compress everything since I don't have the luxury of time.

The extra space may allow me to make a game without worrying too much about fitting everything in the smallest space possible. Of course it may sound sloppy and I do agree with that but since I'm doing it for fun and today we don't have to worry about the cost of chips and everything: who cares? My goal is to have fun making a nes game first and I'm more than happy with that. If I can optimize it latter to fit in a smaller space then I will be even happier. And if this allow me to take a simpler mapper or even none then even better.

by on (#44877)
Oh you are right about the lack of time, partially right at least.
A professional would do that 40 hours per week, but I'm glad if I do 2 or 3 in average, so we would take about 20 times more time to complete the project than a professional.

But also each time you wonder which of the many possible way you'll handle something instead of just handling it the first way that comes to mind is lost.

by on (#44882)
Not to mention also people had a staff dedicated to a project. And each person worked on a specific part, knowing exactly what they had to do. We have to come up with all the concepts behind a game, program it, and create all the data ourselves. Coming up with the concepts behind a game (for example, how maps are to be stored, how AI works, the math behind game logic, etc.) might prove to be the hardest part. Once you know what to program, it's not as hard. Though programming is usually tedious.

by on (#44886)
Programming is not as tedois as doing graphics IMO. And I'm supposed to be part of a team, but I admit I do 95% of the work currently. Altough I'm very happy when my friend gives me ideas, I don't always have good ideas for game (altough I can come with programming skills and do some music), but he always have incredibly fun and original ideas I would never have. Because he knowns absolutely nothing about computers this is even more funny how he can make up stuff. If I don't want to do it because the idea is just too crazy I just say it's technically impossible, and he's forced to belive me because he knowns nothing technical.

by on (#45012)
blargg wrote:
6502freak wrote:
Ambitious developers most certainly don't rely on mappers. They work on the gameplay until it's perfected. [...] Like I said, people gravely underestimate how much work went into the DESIGN of those games. Adjusting the game mechanics until they play perfectly.

It's very similar with Apple, whose products distinguish themselves in the subtle refinements of their designs. Other companies make mp3 players that look similar, but subtle elements are left out. For a sidescroller, perhaps the single biggest thing that affects the feel of the game is the physics of the jump. Even within Nintendo games, a Mario jump is very different from a Samus jump (especially Super Metroid), where Mario is much more constrained by momentum.


That's exactly what I mean.

I am always fascinated when I pop in Super Mario Bros, because although the graphics are highly outdated, the carefully tweaked game mechanism still manages to hook you up immediately. What I mean with game mechanism is how Mario responds to the controls, and how it interacts with the environment. What also works really great in enhancing playability are the sound effects, which are VERY closely timed to match the action. The result is a tight stream of action and feedback between the player and the game, which really is one of the most imortant aspects why SMB still today plays so well.

This is one of those timeless aspects which I am talking about. Of course there are many other aspects wich can be discussed here, but the point is to stress the fact that great games on the NES and other consoles rely SOLELY on their game mechanics, and not the amount of rastersplits or CHR banks you can swap in.

Of course we can only develop NES games in our free time. That's probably one more reason to cut down your game idea to its bare bones, and see how well it still plays. But on the other hand, we have no deadlines to fulfil. We can take as much time as we want to spend into developing our projects.

by on (#45019)
6502freak wrote:
the point is to stress the fact that great games on the NES and other consoles rely SOLELY on their game mechanics, and not the amount of rastersplits or CHR banks you can swap in.

Unless the game mechanics revolve around one of these mapper features. Moving a huge boss around the screen in a platformer needs either a raster split (later Mega Man games) or so little floor detail that horizontal lines with sprites over them are enough (Contra).

Quote:
That's probably one more reason to cut down your game idea to its bare bones, and see how well it still plays.

By "cutting down to bare bones", does that include cutting down the state that persists between power cycles to no more than 64 bits, the comfortable maximum for a 16-character password? Some game genres usually have much more state than that.

Quote:
But on the other hand, we have no deadlines to fulfil. We can take as much time as we want to spend into developing our projects.

Not necessarily. Some people like to finish one project before starting another. (For me it's finish a music engine, then finish a game that uses it, then start another game.) So the deadline is "before one gets sick of the project".