• Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.
  • We have made minor adjustments to how the search bar works on ResetEra. You can read about the changes here.
Oct 27, 2017
153
Many of you kids might not realize this but N64 games were also expensive as fuck. Cartridge games were NOT cheap and videogames didn't really have a universal price like they do now. Pair that with Nintendo's draconian polices regarding licenses and it was only a matter of time before a viable platform came out that didn't treat third party developers like peons for them to steal away Nintendo's pie.

Hubris is quite a thing. We've seen the same happen to Sony during the PS3 era and then with XBOX during the XB1 era. Really curious to see if Sony will buck the trend and continue their streak of making MOSTLY the right moves into the next gen.
 
Dec 2, 2017
3,435
Panasonic did before with 3DO and failed miserably. The idea of Sony making the same mistake was pretty normal at the time.

Like I said, that notion was thrown around initially when they announced in 93 but people stopped saying it really quickly. PS1 hit Japan in 94 - only 2 weeks after Donkey Kong Country hit and the SNES started to peak - and was a success right away. The games were jaw dropping, blowing away the Jaguar and 3DO. By the time the N64 launched 2 years later Square had already jumped ship to Sony, Ridge Racer, Wipeout and Tekken were hits, the Saturn had faltered and while nobody was writing Nintendo off (though they took a fair amount of heat for the cartridge decision) there was a fair amount of talk that they were David up against a financial Goliath. Yes, Nintendoom was a thing even back then.

You don't have to take my word for it, scans of Next Generation, Electronic Gaming Monthly and Gamefan are all floating around online, as well as usenet archives.
 

Slam Tilt

Member
Jan 16, 2018
5,585
It doesn't really matter if Nintendo treated someone well or poorly, on a macro level you're supporting your biggest competitor if you make their platform more attractive by giving it your software.
Yep. This is why third parties will always hesitate to support Nintendo consoles to a degree that they won't with other manufacturers -- because the first-party efforts of Sony and Microsoft and whoever else aren't as attractive as Nintendo's stuff, which makes it easier for third parties to thrive in those ecosystems. Would you prefer to fight for sales against Mario or Knack?

The only way Nintendo can attract meaningful third-party support is to have a runaway success, and even then the support won't be as strong as you'll see elsewhere.
 

Deleted member 19702

User requested account closure
Banned
Oct 27, 2017
1,722
Yamauchi went nuts and screw up. That's why.

Major reason why Nintendo refused to use CDs for N64 was because they would loose the control over manufacturing demand and the profits over this scheme. Namco complained about this control during the NES days and tried to bypass it (Tengen as well), shaking relations with Nintendo afterwards, that's why they heavily supported the PC-Engine and went all in with Sony after PSX came out.

This also was a major reason why Nintendo and Square relations shaked, as Nintendo refused to provide cartdriges with enough storage necessary for their games, reason why Secret of Mana had content cut at last instance. In contrast, Enix had larger storage for their games as Nintendo knew how important Dragon Quest was (and still is) for the japanese market. Despite Square games being hugely succesful and even worked together for Super Mario RPG, Nintendo always treated them as a stepchild and never thoughr FF (or anything else from them) would be something to be missed. Super Mario RPG was also another reason to shake Nintendo/Square relations. It was initally settled that Square would publish it, but Nintendo changed their mind and decided to publish themselves, frustrating Square in return. Final Fantasy VII was initially a 64DD project, an add-on that Nintendo developed especially for the RPG devs, mostly with Enix's Dragon Quest in mind. Nintendo was aware that the default N64 hardware and cartdriges didn't had the enough storage data to carry on the RPG genre, something huge in Japan, so they decided to use the 64DD to keep them at their side. Unlike the official story, Square didn't left because of the N64 lacking CDs, but rather because how badly mistreated they were by Nintendo. Sony decided to give them the red carpet treatment and couldn't refuse, so they decided to jump away from Nintendo and support the PSX as an AAA dev, something they would never have with Nintendo. This decision started a domino effect, causing pretty much every single big japanese dev, with few exceptions, to leave Nintendo and work for the competition. After Nintendo lost both Square and Enix, their 64DD project got a shot in the heart and Nintendo pretty much said "fuck this, I don't need it anymore", and treated it like shit when it came out.

Of course, excluding japanese devs from their "Dream Team" was also a huge factor for why they got alienated on them. It was a clear message that Nintendo was snubbing them.

Control over cartdrige manufacturing demand is a key reason why Nintendo never abandoned this type of media, was always reluctant to use CDs/DVDs as media and tried to use an "alternative" media, such as the GCN's mini-DVD. They have huge profits with this policy and still is a reason why it scare away some devs, despite how huge seller they are. You can see Wii and Switch selling like hot cakes and still struggling to get support.

Nintendo realized how toxic this decision was and how severely shaked their position on the japanese market and their relations with third-parties, so they decided to amend the damage on the GCN/GBA era. This is how Nintendo managed to bring Square back, get the Capcom 5 exclusive deal, work more closely with Namco... pretty much every single major japanese dev they lost to the N64 screw up.

On the other hand, Nintendo handled the N64 nicely in the western market. There wasn't the third-party mass exodus in this region and the system got solid support from western devs, some of then big exclusiviness, such as Star Wars, Turok, WCW vs NWO/WWF, Mission Impossible, Duke Nukem Zero Hour, Doom 64, etc. Pretty much every single big western third-party, with a very few exceptions, supported the N64. That's the reason why N64 was a significant success in the western market and sold almost the same as the SNES. Some tag the N64 as a failure. In Japan, yes, but no in the west, and it's great example from when Nintendo and western third-parties had good relations and how important it was to keep their pace on the western market. They consequently lost it after Nintendo decided to repair the damage from N64's japanese mass exodus and tried to bring them back, alienating western third-party support in return, as it was significantly lost during the GCN days and afterwards. Only now, with Switch, it seems they are starting to come back or treating Nintendo more seriously, again.

N64 took a heavy hit on Nintendo's management logistics and changed them completely. It was a key reason why Nintendo got skeptical toward third-party support and tried to depend mostly on their own efforts, alienating most of third-parties in return to support Nintendo systems. Nintendo decided to change this after how bad this backfired on them, culminating into two huge busts on every single region (GCN and Wii U), or even terminating a life cycle prematurely (Wii), resulting into an unprecedent three years in a row in the red situation. This aftermath made Nintendo reevaluate their strategy and logistics.

Every dumb decision Nintendo made from 1996 onward have it's roots and can be linked to the N64's japanese failure.
 
Last edited:

Hayama Akito

Member
Oct 25, 2017
1,326
Like I said, that notion was thrown around initially when they announced in 93 but people stopped saying it really quickly. PS1 hit Japan in 94 - only 2 weeks after Donkey Kong Country hit and the SNES started to peak - and was a success right away. The games were jaw dropping, blowing away the Jaguar and 3DO. By the time the N64 launched 2 years later Square had already jumped ship to Sony, Ridge Racer, Wipeout and Tekken were hits, the Saturn had faltered and while nobody was writing Nintendo off (though they took a fair amount of heat for the cartridge decision) there was a fair amount of talk that they were David up against a financial Goliath. Yes, Nintendoom was a thing even back then.

You don't have to take my word for it, scans of Next Generation, Electronic Gaming Monthly and Gamefan are all floating around online, as well as usenet archives.

I'm talking about before the PlayStation release in Japan too. In 1994, before the launch, I had zero faith in PlayStation because of what happened with 3DO and CD-I (and I even like 3DO a bit at the time, Super Street Fighter II Turbo blows my mind). I agree with everything you said, but I can't blame anyone who doesn't believe in Sony before 1995.
 

Cheerilee

Banned
Oct 25, 2017
3,969
It's certainly noteworthy (and consistent with your point) that Miyamoto does not mince words, and explicitly throws SGI (by name) under the bus:
It's amusing because despite the complaints, SGI was the frontrunner to produce Nintendo's next console after the N64, and Nintendo became deeply concerned when SGI's best employees started dropping like flies, and breathed a sigh of relief when those same employees turned around and formed ArtX. ArtX (later absorbed by ATI) designed the first and best part of the GameCube, the graphics chip, while Nintendo built the rest of the GameCube around it, as a sort of graphics chip delivery system (which is how GameCube was able to deliver such impressive graphics for such low prices).

And then Nintendo duct taped two GameCubes together to make Wii.

I can't remember if WiiU was still milking that amazing graphics chip given to them by former SGI employees.

I've always thought this as well. The outcome of the N64 is something that Nintendo has never fully recovered from.

Not to go too far off topic but that's why when people say market leaders can change with a gen reset, I disagree. History has proven otherwise
Market leaders can change with the generation reset, but the previous leader usually has an advantage.

NES had an absolute lock on the 8-bit era, but as soon as the generation changed, Sega stole market leadership, and the SNES was playing catch up until basically the last year when Sega bowed out and Nintendo cruised to a belated victory over the Genesis.

And then the 32/64-bit generation saw Nintendo lose all the inherited power which had helped them close the gap on Sega. While Sega wasn't the one to lead, Sony broke in and kicked all of their asses.

The PS2 inherited significant momentum from PSone and (being a strongly positioned console by it's own measure) crushed the ambitions of Dreamcast/GameCube/Xbox.

Then PS3 dropped the ball and X360 blew past Sony, leaving Sony to play a significant game of catch-up (which they eventually did). Meanwhile, Nintendo said "Let's try this GameCube thing again, this time with an exciting gimmick and some fresh marketing" and the second-last-place company shot into first place.

And then Nintendo dropped it all by betting on the wrong gimmick and confusing messaging with WiiU.

And then Nintendo cleaned the mud off that gimmick and tried again, and this time it worked. Meanwhile PS4 has been doing great and Xbone just can't keep up, largely due to a bad first impression with draconian anti-piracy measures they couldn't possibly turn off, until they could.

The generation reset has been a huge factor in all of these huge shifts in the market.
 

MistaTwo

SNK Gaming Division Studio 1
Verified
Oct 24, 2017
2,456
Seriously. We thought PlayStation was kind of a joke at the time. "Sony? Making a games console?"

And then it arrived, and it was more powerful than anyone expected, more reasonably priced than anyone expected, and it had some good ass games (Ridge Racer and Tekken basically, although I'm still fond of Jumping Flash). I wanted one instantly.

I was living out in the countryside during the time (Metropolis, IL) and was more into sports/go-karts/etc. than games at that point in life.
I remember after school one day a bunch of kids invited me over to check out the new "PlayStation" and I had no idea wtf they were on about
but tagged along anyway. For some reason in my head I imagined it as some sort of mini McDonalds Play area for the house or something.

Then I sat around playing a demo of that Extreme Sports game (whatever it was called) for hours and I didn't shut up about the damn thing until I got one for Christmas.
 

Yoshichan

The Fallen
Oct 25, 2017
3,045
Sweden
I'm reading this thread, and all I can think of is how much more drama things seemed to be back in the days? Like, Yamauchi, holy shit.
 

watdaeff4

Banned
Oct 27, 2017
2,451
Market leaders can change with the generation reset, but the previous leader usually has an advantage.

NES had an absolute lock on the 8-bit era, but as soon as the generation changed, Sega stole market leadership, and the SNES was playing catch up until basically the last year when Sega bowed out and Nintendo cruised to a belated victory over the Genesis.

And then the 32/64-bit generation saw Nintendo lose all the inherited power which had helped them close the gap on Sega. While Sega wasn't the one to lead, Sony broke in and kicked all of their asses.

The PS2 inherited significant momentum from PSone and (being a strongly positioned console by it's own measure) crushed the ambitions of Dreamcast/GameCube/Xbox.

Then PS3 dropped the ball and X360 blew past Sony, leaving Sony to play a significant game of catch-up (which they eventually did). Meanwhile, Nintendo said "Let's try this GameCube thing again, this time with an exciting gimmick and some fresh marketing" and the second-last-place company shot into first place.

And then Nintendo dropped it all by betting on the wrong gimmick and confusing messaging with WiiU.

And then Nintendo cleaned the mud off that gimmick and tried again, and this time it worked. Meanwhile PS4 has been doing great and Xbone just can't keep up, largely due to a bad first impression with draconian anti-piracy measures they couldn't possibly turn off, until they could.

The generation reset has been a huge factor in all of these huge shifts in the market.

Respectfully disagree.

8-Bit Nintendo was the market leader
16-Bit Nintendo, despite a healthy headstart froM Sega.....was still the market leader
5th (?) gen - Sony took over
6th gen - Still Sony
7th gen - Nintendo was the leader based over insane initial sales and Sony's slow start. Despite all this Sony still was in 2nd overtaking a head start from MS and got with 20 million of Nintendo
8th gen - Sony

We can say all we want about a generation reset but outside of the initial Wii phenomenon history shows that market leadership in this industry stays solid. It takes a big event for it to change at the generation reset, not merely because we have a new generation and everyone starts from "scratch"
 

Raiden

Member
Nov 6, 2017
2,922
Everyone i knew here was copying games for PSX, and cartridges were like 80 dollars even back then. Im sure that didnt help.
 

Yoshichan

The Fallen
Oct 25, 2017
3,045
Sweden
It's amusing because despite the complaints, SGI was the frontrunner to produce Nintendo's next console after the N64, and Nintendo became deeply concerned when SGI's best employees started dropping like flies, and breathed a sigh of relief when those same employees turned around and formed ArtX. ArtX (later absorbed by ATI) designed the first and best part of the GameCube, the graphics chip, while Nintendo built the rest of the GameCube around it, as a sort of graphics chip delivery system (which is how GameCube was able to deliver such impressive graphics for such low prices).

And then Nintendo duct taped two GameCubes together to make Wii.

I can't remember if WiiU was still milking that amazing graphics chip given to them by former SGI employees.


Market leaders can change with the generation reset, but the previous leader usually has an advantage.

NES had an absolute lock on the 8-bit era, but as soon as the generation changed, Sega stole market leadership, and the SNES was playing catch up until basically the last year when Sega bowed out and Nintendo cruised to a belated victory over the Genesis.

And then the 32/64-bit generation saw Nintendo lose all the inherited power which had helped them close the gap on Sega. While Sega wasn't the one to lead, Sony broke in and kicked all of their asses.

The PS2 inherited significant momentum from PSone and (being a strongly positioned console by it's own measure) crushed the ambitions of Dreamcast/GameCube/Xbox.

Then PS3 dropped the ball and X360 blew past Sony, leaving Sony to play a significant game of catch-up (which they eventually did). Meanwhile, Nintendo said "Let's try this GameCube thing again, this time with an exciting gimmick and some fresh marketing" and the second-last-place company shot into first place.

And then Nintendo dropped it all by betting on the wrong gimmick and confusing messaging with WiiU.

And then Nintendo cleaned the mud off that gimmick and tried again, and this time it worked. Meanwhile PS4 has been doing great and Xbone just can't keep up, largely due to a bad first impression with draconian anti-piracy measures they couldn't possibly turn off, until they could.

The generation reset has been a huge factor in all of these huge shifts in the market.
Where is the Wii here?
 

Crazymoogle

Game Developer
Verified
Oct 25, 2017
2,884
Asia
I'd be very interested to hear any additional discussion of this point, if you have the time, though you've made the point quite clearly in your post already.

Don't get me wrong. Compared to the CPUs of the day - Hitachi's SH-2 and MIPS R3000 - the R4300 was kind of a monster. Nintendo literally went from 3 MHz (SNES) to 93 MHz (N64).

The problem is the platform architecture as a whole was a bad idea for real-time development, because almost all of the decisions they made were in favor of "high quality pixels" as opposed to just rendering more of them.
  • 4kb Texture Cache: In order to avoid hits to main memory, you store the texture for a given object in the texture cache. The problem is, it was so damn small that you had to decide between 16x16x24bit or larger at 16-bit or even 8/4bit color. This is why a lot of N64 games have poor textures - it's not so much having the storage space (although that was a problem!) but to avoid the performance hit of using main memory, textures had to be cut into a bunch of 4kb pieces...which incurred a pretty severe performance hit of its own. A lot of the time you would end up storing textures uncompressed on the cartridge because doing a direct read from ROM was faster than copying to main memory and reading from there. (But then...you waste cart space quickly!)

  • RAMBUS (RDRAM): 4kb as a cache size is not unheard of from that era. But the N64 had a drawback other consoles didn't - RDRAM. The transfer rate is incredible, but the latency was awful. So who cares if you can transfer 800MB/s or whatever if your CPU is stuck doing nothing for 20 cycles? Most N64 coding was about trying to keep the CPU busy, knowing that any main memory hit was like waiting for the bus after work. Other devices would use DMA to access RAM independent of the CPU, but here, that's not possible. It's like having a V12 engine that stalls often. So games like Mario 64 excelled because they just didn't bother with textures most of the time.

  • Microcode: N64 was fill-rate limited, not geometry (the R4300 was a monster for geo). So optimizing rendering means microcode...but the catch was huge. No debugger and if you need particular information on how the CPU works, you needed to make an engineering request to Nintendo. That's right: part of the hardware documentation was confidential (maybe patented?) and Nintendo would not share even in part to authorized developers without a specific ask. So you'd waste time submitting requests, if you had the programmers good enough to work with it. F5 did their own microcode and I believe Nintendo supplied a "fast" mode, but in general fill rate was never enough, hence the allure to change the microcode on a per game basis.
And this isn't even getting into how you'd need an expensive SGI machine to compile, or the amount of time needed to flash a cart. Music was a problem too, because keeping high quality stuff running meant...again, reading from memory using the CPU.
 

Cheerilee

Banned
Oct 25, 2017
3,969
Respectfully disagree.

8-Bit Nintendo was the market leader
16-Bit Nintendo, despite a healthy headstart froM Sega.....was still the market leader
5th (?) gen - Sony took over
6th gen - Still Sony
7th gen - Nintendo was the leader based over insane initial sales and Sony's slow start. Despite all this Sony still was in 2nd overtaking a head start from MS and got with 20 million of Nintendo
8th gen - Sony

We can say all we want about a generation reset but outside of the initial Wii phenomenon history shows that market leadership in this industry stays solid. It takes a big event for it to change at the generation reset, not merely because we have a new generation and everyone starts from "scratch"
I agree that a new generation never starts from scratch, and there's a clear advantage to the previous winner. It's never just about "may the best console win". You can bring your A-game and still lose.

But the generation reset is where major opportunities present themselves (Sega breaking Nintendo's unbreakable monopoly, Sony's introduction, MS earning some respect with X360, Nintendo's two successful requests for a do-over), major blunders happen (N64's carts, Saturn's botched American launch, Saturn's abrupt American cancellation, WiiU, always-on DRM) and chickens come home to roost (NES arrogance, SNES arrogance, PS2 arrogance). These are all seismic events that are fundamentally linked to the generation reset.

And even still, the generation reset isn't everything. SNES managed to climb out of it's own hole. Xbox managed to claw it's way past GameCube. PS3 managed to climb out of it's own hole.

Where is the Wii here?
The unnamed "Let's try this GameCube thing again, this time with an exciting gimmick and some fresh marketing" first-place entry in the PS360 paragraph.
 

Zan

One Winged Slayer
Member
Oct 25, 2017
9,431
What was cut from Secret of Mana?

And yea, N64 games cost a fortune in Canada.......some games were over 100 bucks =/

My mom had to do two jobs for N64 gifts. When time travel becomes a thing, i'm going back and giving SmolZan a PS1 instead. (And convince him to ask for a PS2)

I can basically only think of 5 good exclusive N64 Japanese 3rd party titles: Harvest Moon, Goemon 1, Snowboard Kids, Bomberman 2nd Attack and... maybe one of the Aki WcW titles? Wouldn't even know if that'd count.
 

5pectre

Member
Nov 16, 2017
2,237
And yet all those 3rd party sans Square (Who Nintendo banned and came crawling back 6 years later) still made games for the game boy. And stop with your fanboy Nintendo screwed Sony nonsense.

I'm not one to easily get triggered but calling me a Sony fanboy surely does that! I tolerate Sony and that's about as much as I'll say. And it's not nonesense. Nintendo skipped out of the deal and did another behind their backs and on top of that tried to make them look bad.
 
Nov 4, 2017
7,377
Don't get me wrong. Compared to the CPUs of the day - Hitachi's SH-2 and MIPS R3000 - the R4300 was kind of a monster. Nintendo literally went from 3 MHz (SNES) to 93 MHz (N64).

The problem is the platform architecture as a whole was a bad idea for real-time development, because almost all of the decisions they made were in favor of "high quality pixels" as opposed to just rendering more of them.
  • 4kb Texture Cache: In order to avoid hits to main memory, you store the texture for a given object in the texture cache. The problem is, it was so damn small that you had to decide between 16x16x24bit or larger at 16-bit or even 8/4bit color. This is why a lot of N64 games have poor textures - it's not so much having the storage space (although that was a problem!) but to avoid the performance hit of using main memory, textures had to be cut into a bunch of 4kb pieces...which incurred a pretty severe performance hit of its own. A lot of the time you would end up storing textures uncompressed on the cartridge because doing a direct read from ROM was faster than copying to main memory and reading from there. (But then...you waste cart space quickly!)

  • RAMBUS (RDRAM): 4kb as a cache size is not unheard of from that era. But the N64 had a drawback other consoles didn't - RDRAM. The transfer rate is incredible, but the latency was awful. So who cares if you can transfer 800MB/s or whatever if your CPU is stuck doing nothing for 20 cycles? Most N64 coding was about trying to keep the CPU busy, knowing that any main memory hit was like waiting for the bus after work. Other devices would use DMA to access RAM independent of the CPU, but here, that's not possible. It's like having a V12 engine that stalls often. So games like Mario 64 excelled because they just didn't bother with textures most of the time.

  • Microcode: N64 was fill-rate limited, not geometry (the R4300 was a monster for geo). So optimizing rendering means microcode...but the catch was huge. No debugger and if you need particular information on how the CPU works, you needed to make an engineering request to Nintendo. That's right: part of the hardware documentation was confidential (maybe patented?) and Nintendo would not share even in part to authorized developers without a specific ask. So you'd waste time submitting requests, if you had the programmers good enough to work with it. F5 did their own microcode and I believe Nintendo supplied a "fast" mode, but in general fill rate was never enough, hence the allure to change the microcode on a per game basis.
And this isn't even getting into how you'd need an expensive SGI machine to compile, or the amount of time needed to flash a cart. Music was a problem too, because keeping high quality stuff running meant...again, reading from memory using the CPU.
This was a great write up. I liked that you threw audio in at the end too... Audio was the hardest thing to get used to on the N64 after playing PS1. It sounded like muffled, low-bit fuzz. The N64 was such a workhorse with a few minor design issues that had significant impact (E.g. no DMA for the CPU, and having to access main memory via the RPC using a 250MB/s bus. Insane.).

Thanks for the interesting and informative post.
 

Coloursheep

Member
Oct 31, 2017
185
Yes, except it was technically Nintendo and Sony screwing each other over. The contract that Sony had Nintendo sign was extremely advantageous to Sony to the point that Nintendo would basically lose the rights to their own software.

With that said Nintendo acted in a similar shitty way and backstabbed them back by announcing the partnership with Philips without telling Sony beforehand.

So both of them acted terrible in this relationship. There's no good guy in this story.

Nintendo agreed to the deal (unless there is something to what you said about Sony "had them sign" the deal)then changed their minds and instead of trying to renegotiate with Sony or just tell them straight up they don't like the deal and are going to go a different way, they went behind their partner's back, signed a deal with their rival (a non-japanese rival at that) and let Sony make their announcement knowing full well that they wouldn't go through with it AND that they would be announcing their new deal the next day.

It's hard to imagine a way that they could have screwed Sony over more with this deal or done it in more of a fuck you way.
And all because of a deal that they agreed to. So unless there's something I'm missing I don't see how Sony screwed Nintendo over at all let alone to the same level that Nintendo did to them.
 
Oct 25, 2017
6,099
I'm not one to easily get triggered but calling me a Sony fanboy surely does that! I tolerate Sony and that's about as much as I'll say. And it's not nonesense. Nintendo skipped out of the deal and did another behind their backs and on top of that tried to make them look bad.
Companies always have out clauses if they feel the partnership isn't working it a part of business.
 

sir_crocodile

Member
Oct 25, 2017
23,506
Yamauchi was abusing Nintendo's market position in the sfc era (so developers had little ingrained loyalty to them) and insisted on cartridges for n64 as they would make more money on licensing.
 

Dr. Caroll

Banned
Oct 27, 2017
8,111
Microcode: N64 was fill-rate limited, not geometry (the R4300 was a monster for geo). So optimizing rendering means microcode...but the catch was huge. No debugger and if you need particular information on how the CPU works, you needed to make an engineering request to Nintendo. That's right: part of the hardware documentation was confidential (maybe patented?) and Nintendo would not share even in part to authorized developers without a specific ask. So you'd waste time submitting requests, if you had the programmers good enough to work with it. F5 did their own microcode and I believe Nintendo supplied a "fast" mode, but in general fill rate was never enough, hence the allure to change the microcode on a per game basis.
And this isn't even getting into how you'd need an expensive SGI machine to compile, or the amount of time needed to flash a cart. Music was a problem too, because keeping high quality stuff running meant...again, reading from memory using the CPU.
The microcode situation is really complicated. Nintendo released updated versions of the official ucodes with each SDK update. The newer ucodes typically traded precision for performance. You had Fast3D, then Fast3DEX, then Fast3DEX2. There was a 2D microcode. A line drawing one. And so on. As an example, Rogue Squadron's is an extensively modified version of Fast3DEX. Naboo/Indiana Jones's are extensively modified versions of Fast3DEX2. There are no fully custom microcodes as far as I'm aware. They're modified versions of Nintendo ones. No need to reinvent the wheel, I guess. But many N64 games have partially custom microcodes. The Turok games, for instance, have a custom version of Fast3DEX with a custom lighting solution. (N64 devs had a massive dynamic lighting fetish.)

Where it gets kinda interesting is that towards the end of the system's life, relatively speaking, there was a ucode named ZSort. The BOSS racing games (WDC/Stunt Racer) use a custom version of ZSort.

The vanilla version was used in soccer games and... nothing else.

Z-sort significantly alleviates fill rate problems. It sorts based on depth and draws back to front, as the name suggests. The fact a whopping 3 or so games used this ucode is really indicative of how poorly Nintendo managed 3rd party development efforts. Here's a serious solution for RDP fill rate problems, and Nintendo didn't really bother fostering its use. They just seemingly didn't care.
 

D.Lo

Member
Oct 25, 2017
4,348
Sydney
It's amusing because despite the complaints, SGI was the frontrunner to produce Nintendo's next console after the N64, and Nintendo became deeply concerned when SGI's best employees started dropping like flies, and breathed a sigh of relief when those same employees turned around and formed ArtX. ArtX (later absorbed by ATI) designed the first and best part of the GameCube, the graphics chip, while Nintendo built the rest of the GameCube around it, as a sort of graphics chip delivery system (which is how GameCube was able to deliver such impressive graphics for such low prices).

And then Nintendo duct taped two GameCubes together to make Wii.

I can't remember if WiiU was still milking that amazing graphics chip given to them by former SGI employees.


Market leaders can change with the generation reset, but the previous leader usually has an advantage.

NES had an absolute lock on the 8-bit era, but as soon as the generation changed, Sega stole market leadership, and the SNES was playing catch up until basically the last year when Sega bowed out and Nintendo cruised to a belated victory over the Genesis.

And then the 32/64-bit generation saw Nintendo lose all the inherited power which had helped them close the gap on Sega. While Sega wasn't the one to lead, Sony broke in and kicked all of their asses.

The PS2 inherited significant momentum from PSone and (being a strongly positioned console by it's own measure) crushed the ambitions of Dreamcast/GameCube/Xbox.

Then PS3 dropped the ball and X360 blew past Sony, leaving Sony to play a significant game of catch-up (which they eventually did). Meanwhile, Nintendo said "Let's try this GameCube thing again, this time with an exciting gimmick and some fresh marketing" and the second-last-place company shot into first place.

And then Nintendo dropped it all by betting on the wrong gimmick and confusing messaging with WiiU.

And then Nintendo cleaned the mud off that gimmick and tried again, and this time it worked. Meanwhile PS4 has been doing great and Xbone just can't keep up, largely due to a bad first impression with draconian anti-piracy measures they couldn't possibly turn off, until they could.

The generation reset has been a huge factor in all of these huge shifts in the market.
Great post.

Seriously, what kind of crazy company would want to be involved in this market. Every five years you get a market reset!

And even still, the generation reset isn't everything. SNES managed to climb out of it's own hole. Xbox managed to claw it's way past GameCube. PS3 managed to climb out of it's own hole.
Some of these are a bit distorted. SNES took a while to overtake the Genesis in the west, but worldwide it was ahead very quickly thanks to Japan, and it wasn't quite as simple as SNES vs Mega Drive anyway as the NES was still selling lots of software for several years.

Xbox and PS3 didn't really crawl out, they were bought out, by their companies throwing billions at them to make and keep them relevant. All Microsoft consoles and the PS3 were huge market distortions in that financially they did not compete on fair terms, as they were subsidised so heavily to force some success. Other consoles have lost money, but typically in a market a product like that is let to run out quietly until a new product launch (examples in consoles are the Saturn and Wii U, which could have sold more if their companies threw money at them, but instead were left around to limp to death). PS3 in particular got an irrational, abnormal amount of reinvestment on bad money. Some may look at the PS4 building on that (purchased) momentum and so it was worth it, but if Microsoft didn't screw up so much with forced kinect and DRM who knows what would have happened.
 

Narasumas

Banned
Nov 27, 2017
952
Melbourne, Florida
There was a lot of upside to the Gamecube's mini-discs.

Just out of curiosity, what are those upsides? They still packaged them in "standard" sized boxes. I will say though, Nintendo had some of the best disc art out of that generation. And I always (and still do) think it's awesome whenever I get a CD or Soundtrack sampler that comes on a mini disc. I always feel nervous that my regular disc reader will eat them though lol.
 

Metallix87

User Requested Self-Ban
Banned
Nov 1, 2017
10,533
Just out of curiosity, what are those upsides? They still packaged them in "standard" sized boxes. I will say though, Nintendo had some of the best disc art out of that generation. And I always (and still do) think it's awesome whenever I get a CD or Soundtrack sampler that comes on a mini disc. I always feel nervous that my regular disc reader will eat them though lol.
The upsides mostly revolve around faster load times, which were the goal of the design from the get-go. It also avoided licensing fees to the DVD forum, and made it much more difficult to pirate Gamecube games. These were all by design, as Nintendo approached Panasonic to design the miniDVD optical discs with reduced load times, increased difficulty in piracy, and avoiding the fees as the express goals.
 

Nightside

Member
Oct 28, 2017
625
Don't get me wrong. Compared to the CPUs of the day - Hitachi's SH-2 and MIPS R3000 - the R4300 was kind of a monster. Nintendo literally went from 3 MHz (SNES) to 93 MHz (N64).

The problem is the platform architecture as a whole was a bad idea for real-time development, because almost all of the decisions they made were in favor of "high quality pixels" as opposed to just rendering more of them.
  • 4kb Texture Cache: In order to avoid hits to main memory, you store the texture for a given object in the texture cache. The problem is, it was so damn small that you had to decide between 16x16x24bit or larger at 16-bit or even 8/4bit color. This is why a lot of N64 games have poor textures - it's not so much having the storage space (although that was a problem!) but to avoid the performance hit of using main memory, textures had to be cut into a bunch of 4kb pieces...which incurred a pretty severe performance hit of its own. A lot of the time you would end up storing textures uncompressed on the cartridge because doing a direct read from ROM was faster than copying to main memory and reading from there. (But then...you waste cart space quickly!)

  • RAMBUS (RDRAM): 4kb as a cache size is not unheard of from that era. But the N64 had a drawback other consoles didn't - RDRAM. The transfer rate is incredible, but the latency was awful. So who cares if you can transfer 800MB/s or whatever if your CPU is stuck doing nothing for 20 cycles? Most N64 coding was about trying to keep the CPU busy, knowing that any main memory hit was like waiting for the bus after work. Other devices would use DMA to access RAM independent of the CPU, but here, that's not possible. It's like having a V12 engine that stalls often. So games like Mario 64 excelled because they just didn't bother with textures most of the time.

  • Microcode: N64 was fill-rate limited, not geometry (the R4300 was a monster for geo). So optimizing rendering means microcode...but the catch was huge. No debugger and if you need particular information on how the CPU works, you needed to make an engineering request to Nintendo. That's right: part of the hardware documentation was confidential (maybe patented?) and Nintendo would not share even in part to authorized developers without a specific ask. So you'd waste time submitting requests, if you had the programmers good enough to work with it. F5 did their own microcode and I believe Nintendo supplied a "fast" mode, but in general fill rate was never enough, hence the allure to change the microcode on a per game basis.
And this isn't even getting into how you'd need an expensive SGI machine to compile, or the amount of time needed to flash a cart. Music was a problem too, because keeping high quality stuff running meant...again, reading from memory using the CPU.

That's why they released the memory extension?
 

javac

Member
Oct 28, 2017
3,153
or even terminating a life cycle prematurely (Wii)
The Wii had a normal ass lifecycle tho, Sony and Microsoft just expanded their time because they only started making money late into the cycle, Wii had decent support until the end. Unlike the 3DS I admit once Wii U came out the Wii was shelved but like the DS that was probably due to them wanting to mitigate any further confusion between the two systems (Wii and Wii U, DS and 3DS), the DS however had Pokémon which gave it extra life, even so the Wii lasted 6 years.
 

logash

Member
Oct 27, 2017
4,739
It's a shame because I think the N64 had more innovative games than the PlayStation but Nintendo really screwed up and did so in the way only a market leader could have. Their arrogance with third parties alienated them and their decision to stick with cartridges put the final nail in their coffen. Sony treated developers better and the PS1 supported a cheaper to make storage media that could hold vastly more space. The N64 never stood a chance.
 

Orioto

Member
Oct 26, 2017
4,716
Paris
I think there was also a problem with their demographics, clearly. Nintendo didn't know how to evolve with their ips and their audience. The whole 16bit era was kinda stagnating with a teen market, and the PlayStation could extend that to young adults. Even the Saturn, when you think about it, had a more mature, hardcore profile. Games like Virtua Fighters, but even games like sakura Taisen (one of the big hitters at that time on Saturn) was looking for a more mature crowd with its sim+tactical style. Also fighting games from SNk and Capcom were pretty hardcore, arcade games at the time, and not something a 12yo would play.
Also, you know, paradoxically, the choices that killed the N64 in japan are kinda the reverse of what killed Sega in the west.. So that wasn't such a bad idea.
I mean, N64 was more 3D oriented, multiplayer, good for sport games etc... It was a big success in the west and that's what allowed Nintendo to stay in touch. The gamecube would have been a way bigger failure if there wasn't a solid "multiplayer party type" game made popular by the N64. Mario party, Mario Kart etc.. Smash Bros to.. All of this finds its root on N64.
While Sega banked on hardcore, more typical 2d type of games, that killed their international audience and hey would never recover.
 

Treasure Silvergun

Self-requested ban
Banned
Dec 4, 2017
2,206
A lot of reasons can explain the N64 debacle in Japan, but as far as games are concerned, the answer is most likely: JRPGs.

Notice how the PSOne era is known as the golden age of JRPGs? JRPGs had grown a lot since the first Dragon Quest days, and in the Super Famicom era they were already very, very big in Japan. JRPGs, shooters and action games were the most represented genres on the Super Famicom. Japan wanted JRPGs.

I remember a 1995 interview where Miyamoto clearly stated that CDs were too slow, and not the ideal medium for video games. At the time, he wasn't wrong. Remember the earliest PSOne games? You could play Galaga while the whole game was being loaded into the console's RAM. Having to wait for the game to load was such a stark departure from the instant loading provided by cartridges. We got used to it, but you can't blame Nintendo for thinking that for their own first-party games, fast loading was of the essence. It still was in the GC era, hence the smaller discs compared to DVDs.

Of course, proprietary cartridges were also crucial for Nintendo to control (bully) third parties, and to get some hefty royalties.

The truth is, the storage space provided by CDs was all but necessary in the early 32-bit era. A lot of developers put unnecessary FMVs and voice acting in their games because they could - and because the games would look much cooler that way. That kind of storage space was a solution looking for a problem. FMVs and voice acting were unquestionably the elements of those games that took the most time to render and record, and very likely the most expensive too.

Quite a lot of JRPGs (and other genres as well) were still 2D even after FF7 hit and changed things forever. And even FF7 would fit entirely on a single CD, were it not for the FMVs. It's been known for a long time that if you swap discs while an FMV is playing, all the areas from the game will load on any of the three CDs. The only difference between the three discs is the FMVs on each disc. Even a 2-CD wonder like Resident Evil 2 could fit on a N64 cartridge, and in high resolution, even if the FMVs took a hit in video quality.

So yeah, at least in the early days of the 32-bit era, games didn't require CDs, and CDs just introduced loading times. For a lot of those games, CD was actually the inferior medium. But it was way cheaper than Nintendo's carts, so devs went the way of the CD and never looked back. And when the JRPGs came in, with FF at the forefront, it was game over for the N64. Japan isn't interested in cutting-edge tech as much as they're interested in game genres they're accustomed to, and that cater to their taste and culture.

Also, let's not forget that by the time the N64 launched in Japan, the PSOne had already been on the market for a year and a half. That's a lot of time.
 

Narasumas

Banned
Nov 27, 2017
952
Melbourne, Florida
The upsides mostly revolve around faster load times, which were the goal of the design from the get-go. It also avoided licensing fees to the DVD forum, and made it much more difficult to pirate Gamecube games. These were all by design, as Nintendo approached Panasonic to design the miniDVD optical discs with reduced load times, increased difficulty in piracy, and avoiding the fees as the express goals.
Nice! Thanks for the info. That definitely makes sense, since Ninty is uber-paranoid about piracy loopholes. Smaller radius also would imply smaller travel distance for the laser reader, i.e.- faster read (load) times. Also explains the partnership with Panasonic where they made the Panasonic-Q! I think it's such a weird (but bad-ass) looking console!
pq5.jpg
 

Metallix87

User Requested Self-Ban
Banned
Nov 1, 2017
10,533
Nice! Thanks for the info. That definitely makes sense, since Ninty is uber-paranoid about piracy loopholes. Smaller radius also would imply smaller travel distance for the laser reader, i.e.- faster read (load) times. Also explains the partnership with Panasonic where they made the Panasonic-Q! I think it's such a weird (but bad-ass) looking console!
pq5.jpg
Yep. The partnership with Panasonic to develop the Gamecube discs made perfect sense from Nintendo's perspective, especially since, while compared to DVD it was a downgrade in space, it was still a major upgrade from Saturn and PS1 CDs.
 

FormatCompatible

One Winged Slayer
Member
Oct 25, 2017
12,071
Nintendo agreed to the deal (unless there is something to what you said about Sony "had them sign" the deal)then changed their minds and instead of trying to renegotiate with Sony or just tell them straight up they don't like the deal and are going to go a different way, they went behind their partner's back, signed a deal with their rival (a non-japanese rival at that) and let Sony make their announcement knowing full well that they wouldn't go through with it AND that they would be announcing their new deal the next day.

It's hard to imagine a way that they could have screwed Sony over more with this deal or done it in more of a fuck you way.
And all because of a deal that they agreed to. So unless there's something I'm missing I don't see how Sony screwed Nintendo over at all let alone to the same level that Nintendo did to them.
I agree.
 

Deleted member 19702

User requested account closure
Banned
Oct 27, 2017
1,722
The Wii had a normal ass lifecycle tho, Sony and Microsoft just expanded their time because they only started making money late into the cycle, Wii had decent support until the end. Unlike the 3DS I admit once Wii U came out the Wii was shelved but like the DS that was probably due to them wanting to mitigate any further confusion between the two systems (Wii and Wii U, DS and 3DS), the DS however had Pokémon which gave it extra life, even so the Wii lasted 6 years.

Wii did have a premature death, yes. You can see how the system became a hollow wasteland in it's late life. Never before in the gaming history the top selling system of it's generation passed through this situation, making Wii and exception and an outlier. PS3/360, even bleeding money in it's first years and recovering afterwards, managed to get most of the third party support, something Wii never did even in it's prime. They could have lasted longer than it did if Nintendo didn't abruptely pulled off the plug in it's late life in order to start support for 3DS and Wii U, consequently scaring away third party support as well. Wii latelife situation wasn't any different from any Nintendo console post-N64 when they had to choose where to concentrate most of their support because they lacked in house devs to properly support two system at once, leaving the detracted system with huge droughts and shortages. Wii couldn't get away from it despite the huge sales.

This "Wii had a normal life" rethoric in order to make apologism for how Nintendo mismanagent doesn't make sense if you actually take a look at many factors that caused to die early.
 
Last edited:

javac

Member
Oct 28, 2017
3,153
Wii did had a premature death, yes. You can see how the system became a hollow wasteland in it's late life. Never before in the gaming history the top selling system of it's generation passed through this situation, making Wii and exception and an outlier. PS3/360, even bleeding money in it's first years and recovering afterwards, managed to get most of the third party support, something Wii never did even in it's prime. They could have lasted longer than it did if Nintendo didn't abruptely pulled off the plug in it's late life in order to start support for 3DS and Wii U, consequently scaring away third party support as well. Wii latelife situation wasn't any different from any Nintendo console post-N64 when they had to choose concentrate most of their support because they didn't enough in house devs to properly support two system at once, leaving the detracted system with huge droughts and shortages. Wii couldn't get away from it despite the huge sales.

This "Wii had a normal life" rethoric in order to make apologism for how Nintendo mismanagent doesn't make sense if you actually take a look at many factors that caused to die early.
I don't really agree but interesting perspective, I just never saw it that way.
 

TaterTots

Member
Oct 27, 2017
12,964
The only thing I disliked about the N64 was the controller. I didn't care that it used cartridges. The console had a lot of great games. Mario 64, OoT, Goldeneye, Banjo-Kazooie, Perfect Dark, Mario Kart 64, Banjo-Tooie, Star Fox 64 and the first Mario Party.
 
Oct 25, 2017
27,849
My mom had to do two jobs for N64 gifts. When time travel becomes a thing, i'm going back and giving SmolZan a PS1 instead. (And convince him to ask for a PS2)

I can basically only think of 5 good exclusive N64 Japanese 3rd party titles: Harvest Moon, Goemon 1, Snowboard Kids, Bomberman 2nd Attack and... maybe one of the Aki WcW titles? Wouldn't even know if that'd count.

WCW vs NWO World Tour was over 100 bucks and was extremely difficult to find, only one store had it
 
Oct 27, 2017
995
Don't get me wrong. Compared to the CPUs of the day - Hitachi's SH-2 and MIPS R3000 - the R4300 was kind of a monster. Nintendo literally went from 3 MHz (SNES) to 93 MHz (N64).

The problem is the platform architecture as a whole was a bad idea for real-time development, because almost all of the decisions they made were in favor of "high quality pixels" as opposed to just rendering more of them.
• 4kb Texture Cache: In order to avoid hits to main memory, you store the texture for a given object in the texture cache. The problem is, it was so damn small that you had to decide between 16x16x24bit or larger at 16-bit or even 8/4bit color. This is why a lot of N64 games have poor textures - it's not so much having the storage space (although that was a problem!) but to avoid the performance hit of using main memory, textures had to be cut into a bunch of 4kb pieces...which incurred a pretty severe performance hit of its own. A lot of the time you would end up storing textures uncompressed on the cartridge because doing a direct read from ROM was faster than copying to main memory and reading from there. (But then...you waste cart space quickly!)
• RAMBUS (RDRAM): 4kb as a cache size is not unheard of from that era. But the N64 had a drawback other consoles didn't - RDRAM. The transfer rate is incredible, but the latency was awful. So who cares if you can transfer 800MB/s or whatever if your CPU is stuck doing nothing for 20 cycles? Most N64 coding was about trying to keep the CPU busy, knowing that any main memory hit was like waiting for the bus after work. Other devices would use DMA to access RAM independent of the CPU, but here, that's not possible. It's like having a V12 engine that stalls often. So games like Mario 64 excelled because they just didn't bother with textures most of the time.
• Microcode: N64 was fill-rate limited, not geometry (the R4300 was a monster for geo). So optimizing rendering means microcode...but the catch was huge. No debugger and if you need particular information on how the CPU works, you needed to make an engineering request to Nintendo. That's right: part of the hardware documentation was confidential (maybe patented?) and Nintendo would not share even in part to authorized developers without a specific ask. So you'd waste time submitting requests, if you had the programmers good enough to work with it. F5 did their own microcode and I believe Nintendo supplied a "fast" mode, but in general fill rate was never enough, hence the allure to change the microcode on a per game basis.
And this isn't even getting into how you'd need an expensive SGI machine to compile, or the amount of time needed to flash a cart. Music was a problem too, because keeping high quality stuff running meant...again, reading from memory using the CPU.

The microcode situation is really complicated. Nintendo released updated versions of the official ucodes with each SDK update. The newer ucodes typically traded precision for performance. You had Fast3D, then Fast3DEX, then Fast3DEX2. There was a 2D microcode. A line drawing one. And so on. As an example, Rogue Squadron's is an extensively modified version of Fast3DEX. Naboo/Indiana Jones's are extensively modified versions of Fast3DEX2. There are no fully custom microcodes as far as I'm aware. They're modified versions of Nintendo ones. No need to reinvent the wheel, I guess. But many N64 games have partially custom microcodes. The Turok games, for instance, have a custom version of Fast3DEX with a custom lighting solution. (N64 devs had a massive dynamic lighting fetish.)

Where it gets kinda interesting is that towards the end of the system's life, relatively speaking, there was a ucode named ZSort. The BOSS racing games (WDC/Stunt Racer) use a custom version of ZSort.

The vanilla version was used in soccer games and... nothing else.

Z-sort significantly alleviates fill rate problems. It sorts based on depth and draws back to front, as the name suggests. The fact a whopping 3 or so games used this ucode is really indicative of how poorly Nintendo managed 3rd party development efforts. Here's a serious solution for RDP fill rate problems, and Nintendo didn't really bother fostering its use. They just seemingly didn't care.


Thanks for the additional detail and discussion, really interesting stuff! I suppose someone like Hiroshi Kawai, doing his hardware tests for FF7 back in 1995, would have been able to get much better results if either SGI or Nintendo had been able to supply a variety of official, performance-oriented microcodes (as opposed to just the 'standard', precision-oriented one), that could have helped optimize performance for the particular rendering demands of his unique benchmark tests:

https://www.polygon.com/a/final-fantasy-7
Hiroshi Kawai: I kind of had a suspicion that things weren't going too well for the 64 at that point, because … one of my responsibilities … was to write performance applications that compared how well the 64 fared against the prototype [PlayStation]. And we'd be running parallel comparisons between the [PlayStation] where you'd have a bunch of 2D sprites bouncing off the screen and see how many polygons you could get within a 60th of a second. And even without any kind of texturing or any kind of lighting, it was less than 50% of what you would be able to get out of the [PlayStation]. Of course, the drawback of the [PlayStation] is it didn't really have a z-buffer, so you'd have these overlapping polygons that you'd have to work around so that you wouldn't get the shimmering [look]. But on the other hand, there was no way you'd be able to get anything close to what FF7 was doing [on PlayStation] on the 64 at that time….

But it sounds like the newer microcodes (as implemented later in the N64's life by Rare, Iguana, Factor 5, BOSS, et al.) were worked out with the benefit of time and 'hands-on' developer experience with the architecture. Do you think it would've been feasible for SGI and Nintendo to have a variety of official, performance-oriented microcodes available to developers like Kawai prior to launch (i.e., could they have offered PS1-level performance to someone with Kawai's particular needs, at a time when they didn't have the benefit of 'hands-on' experience with the hardware)?

In either case, whether it was feasible or not to have made multiple performance-oriented microcodes available prior to launch, the fact that they weren't available prior to launch (resulting in developer struggles) ultimately does not reflect well on the deal Nintendo made with SGI, and hence Crazymoogle's earlier statement ("...Nintendo's deal with SGI was basically a lemon...") still seems quite accurate, to me.
It's amusing because despite the complaints, SGI was the frontrunner to produce Nintendo's next console after the N64, and Nintendo became deeply concerned when SGI's best employees started dropping like flies, and breathed a sigh of relief when those same employees turned around and formed ArtX. ArtX (later absorbed by ATI) designed the first and best part of the GameCube, the graphics chip, while Nintendo built the rest of the GameCube around it, as a sort of graphics chip delivery system (which is how GameCube was able to deliver such impressive graphics for such low prices).

And then Nintendo duct taped two GameCubes together to make Wii.

I can't remember if WiiU was still milking that amazing graphics chip given to them by former SGI employees.

True, that's definitely important context to add here, in defense of the ArtX folks. I think our overall perspective changes slightly, if a "lemon" with "big technical issues" (as identified not only by devs like Miyamoto in 1999, and Crazymoogle here in this thread, but also Hiroshi Kawai speaking to Polygon in 2016/2017) is redefined in such a way that the architecture is merely the subject of "complaints," as you put it, but I think everyone would agree that the folks formerly known as SGI (and specifically, the particular folks who were already known as 'ArtX' by the time of Miyamoto's 1999 comment about 'SGI') learned from their 'mistakes'/experience on the N64, and did much better with the GameCube.
With the amount of control sony wanted. Any sane company would have turned sony down.

As mentioned earlier, that's essentially what industry observers were saying, at the time this was all playing out in public (see NYT articles from 1991, here and here). And those observers were openly wondering how exactly Nintendo had managed to put themselves into such a disadvantageous position. When I asked this exact question over at the old forum, Cheerilee cited material from the book Game Over by David Sheff, and also offered some interesting points for speculation:
Cheerilee said:
David Sheff in Game Over muses whether Hiroshi Yamauchi might have been intimidated by Sony at the time, but I think it was much more likely due to Sheff's other theory, that Yamauchi/Nintendo didn't recognize/understand the threat. Also, lawyers heavily involved in corporate Japan supposedly wasn't a thing at the time, and the contract was probably hammered out by Nintendo and Sony's CEOs personally over drinks...

Ken Kutaragi went behind Sony's back and designed the SNES sound chip without permission. He asked for permission to work on the SNES only after he already had Nintendo standing there with a purchase order in their hands.

For Kutaragi's next trick, he wanted to bring the greatness of CD to the SNES, and Yamauchi gave him permission. I think Yamauchi saw this as nothing but a good thing, and didn't recognize how this could have dire consequences for Nintendo.

The SNES CD contract was signed (according to Game Over) in 1988, two years before the Japanese launch of the SNES (probably around the time Nintendo bought Sony's sound chip, maybe even in the same contract). According to Wikipedia, CD had just overtaken vinyl as a second-place audio storage medium in 1988 (with cassette tapes holding the lead). I can't see Yamauchi realizing CD's potential.

I don't believe game approval rights and royalty fees were negotiated into the contract. Nintendo simply "gave permission" for Kutaragi to go crazy and build something great, assuming it would work out as well for Nintendo as the sound chip did. After Kutaragi built the Nintendo PlayStation, Nintendo expected to simply be handed the keys to it (along with a better explanation of what "it" was), and Sony said "What? We made this, with your permission. These keys are ours." In the absence of a deal specifying who owns the backdoor which Kutaragi legally created, possession is 9/10ths of the law.

Nintendo was only able to extract themselves from this unexpectedly terrible position (of Sony/Kutaragi being primed to usurp Nintendo on their own Super Nintendo) by threatening to go nuclear, and prepping the bombs to back their words up.
 

Encephalon

Member
Oct 26, 2017
5,856
Japan
Also, lawyers heavily involved in corporate Japan supposedly wasn't a thing at the time, and the contract was probably hammered out by Nintendo and Sony's CEOs personally over drinks...

This sounds like the sort of thing that is based on partial information and a misunderstanding of how things work. It doesn't make any sense.
 

Deleted member 19702

User requested account closure
Banned
Oct 27, 2017
1,722
https://www.polygon.com/a/final-fantasy-7
Hiroshi Kawai: I kind of had a suspicion that things weren't going too well for the 64 at that point, because … one of my responsibilities … was to write performance applications that compared how well the 64 fared against the prototype [PlayStation]. And we'd be running parallel comparisons between the [PlayStation] where you'd have a bunch of 2D sprites bouncing off the screen and see how many polygons you could get within a 60th of a second. And even without any kind of texturing or any kind of lighting, it was less than 50% of what you would be able to get out of the [PlayStation]. Of course, the drawback of the [PlayStation] is it didn't really have a z-buffer, so you'd have these overlapping polygons that you'd have to work around so that you wouldn't get the shimmering [look]. But on the other hand, there was no way you'd be able to get anything close to what FF7 was doing [on PlayStation] on the 64 at that time….

Excellent post.

This info puts to rest whatever claim by Sony fanboys that FF7 was never being developed for N64 in the first place, which actually was but Square changed it's mind later for multiple reasons.

As mentioned earlier, that's essentially what industry observers were saying, at the time this was all playing out in public (see NYT articles from 1991, here and here). And those observers were openly wondering how exactly Nintendo had managed to put themselves into such a disadvantageous position. When I asked this exact question over at the old forum, Cheerilee cited material from the book Game Over by David Sheff, and also offered some interesting points for speculation:

This, too. It boogles my mind that some people keep insisting on the biased and one-sided view that Nintendo "stabbed Sony in the back" without proper knowlege of what actually happened. This deal was a death wish for Nintendo's autonomy as a hardware maker. They did the right thing by jumping off at time.
 

SuperBlank

Attempted to circumvent ban with alt account
Banned
Oct 31, 2017
1,591
Failure to read the market + hubris seemed like the main problems.
 
Oct 27, 2017
995
This sounds like the sort of thing that is based on partial information and a misunderstanding of how things work. It doesn't make any sense.

That particular claim does sound a bit odd, and with regard to that claim I'm not sure if Cheerilee is simply characterizing/summarizing David's Sheff's argument, or offering speculation of their own.
 

Cheerilee

Banned
Oct 25, 2017
3,969
That particular claim does sound a bit odd, and with regard to that claim I'm not sure if Cheerilee is simply characterizing/summarizing David's Sheff's argument, or offering speculation of their own.
Probably partial information and misunderstanding.

I remember I had heard people saying that the legal/corporate culture in Japan was different back in the 80's, but then other people contested that. *shrug* I'm definitely not an expert in these matters, I just try to absorb and process events that fascinate me.
 

Crazymoogle

Game Developer
Verified
Oct 25, 2017
2,884
Asia
I liked that you threw audio in at the end too... Audio was the hardest thing to get used to on the N64 after playing PS1. It sounded like muffled, low-bit fuzz.)

It's not that the N64 necessarily had bad sound hardware - there was a DSP on there - but keeping the CPU fed was impossibly hard, and even if it was possible to just store chunks of uncompressed music, then storage was a huge issue given the pressure to make an 8MB cart vs. 64 or whatever. When we worked with Neil Voss (Tetrisphere) he had to go through so many hoops to give us even 1 or 2 workable tracks, and then a channel too many and suddenly it's hurting game performance.

The microcode situation is really complicated. Nintendo released updated versions of the official ucodes with each SDK update.

Thanks for the recap - it's been what, 19 years since I last was involved in this stuff? :lol But I distinctly remember our Lead GPs submitting technical papers to Nintendo just to gain access to certain microcode instructions, which we needed to tune the best one at the time to improve performance. Turns out making a 4 player action RPG was a huge fillrate strain. ;) I really believe they lacked the engineering department necessarily to fully understand their platform, and that they inevitably thought they could rely on SGI as their support mechanism.

I'm not surprised a ucode came out that addressed the massive performance penalty on the ZBuffer, but by 1999 Nintendo themselves was shedding contracts and trying to move everyone to either the Gamecube waiting list (early devkits in america were not available until at least Q3 2000 at the absolute earliest) or the GBA early kits (the motherboard with SNES controllers). The moment DD was canned, they were looking for an out.

That's why they released the memory extension?

Adding more memory would have been incredible for something like the PlayStation, but for the N64 it was adding a fuel tank when there wasn't enough gas at the pump. We were using the memory pak because our original spec was N64DD - which required it - but it was basically only useful for three things:
  1. Debug (because a game in development never fits on the target platform, even today)
  2. Decompressing very big files on initial game load (game geometry, usually)
  3. 640x480 mode
If you asked any developer at the time if they would prefer more RAM or more cart space, they would always say cart space, simply because the access time hit to RAM was so bad. (And I believe it was slightly worse to address the expansion memory.) Even on the GBA, it was much easier to just store uncompressed assets on cart and access them directly.

But it sounds like the newer microcodes (as implemented later in the N64's life by Rare, Iguana, Factor 5, BOSS, et al.) were worked out with the benefit of time and 'hands-on' developer experience with the architecture. Do you think it would've been feasible for SGI and Nintendo to have a variety of official, performance-oriented microcodes available to developers like Kawai prior to launch (i.e., could they have offered PS1-level performance to someone with Kawai's particular needs, at a time when they didn't have the benefit of 'hands-on' experience with the hardware)?

0%, because SGI had no experience with real-time hardware. We have to remember that even to Nintendo, 3D was undiscovered country. A lot of the structure and functions that you'd use in Unity today just weren't known, understood or even supported back then. The PS1 didn't ship with a Z buffer! So they had all of these great engineers who were working to understand how to deliver pixel accuracy that could decrease render times from hours to minutes, and here's Nintendo needing to do 30-60 fps with a Zbuffer. Nintendo didn't have the time or expertise to learn how to fix it by launch let alone realize the critical flaws of their platform.