Hacker News Reader

Comment by delusional | Hacker News Reader

I'm always dubious of these claims. For some reason the marketing guys are always the key to the success, and nothing else barely matters. In these "documentaries" the hardware could have been terrible, hell they could have been selling bananas, but if the marketing would have just been tweaked a bit, it would all have worked out.

It's not surprising that the marketing guys market themselves the best and therefore end up with the more compelling story. It's just boring and trite. Tell me the story of their hardware. The ups and downs of developing games for a platform that wasn't done. The stories of turmoil, as the executives started demanding more features to match the competition.

View on HN

You're not wrong. The Genesis shipped with fairly 'boring' but powerful hardware for the time - an M68K, a Yamaha synth, and a PPU-style video scanout chip. In contrast the Saturn was a kitchen sink of novel ideas that developers struggled to work with. Two CPUs, a polygon rasterizer[0], a sound system with an M68K inside, AND a DSP with a VLIW instruction set.

At the same time I'm not convinced that these failures are entirely orthogonal. Developer marketing is still marketing. At some point during the Saturn's development someone should have said "hey, maybe we shouldn't make developers juggle five CPUs to get decent performance." Letting the system engineers go wild with power is how you wind up shipping a system that looks really powerful on paper but nobody can actually develop games for. See also: the PlayStation 3.

[0] Which itself wasn't even as functional as the PlayStation's. If I remember correctly, rendering to a texture was either impossible or merely very difficult. SEGA also stuck a PPU in there as well which you could use for some effects but it didn't make up for those missing features.

I worked with the DreamCast dev hardware for a time back then. The rendering hardware was definitely rather different than typical triangle-based approaches. IIRC, polygonal shapes were defined by means of multiple intersecting planes. I also seem to recall that it was capable of unusually high frame rates (for the time).

Are you perharps mixing up the Dreamcast and the Saturn? The Dreamcast had a fairly standard PowerVR GPU.

Yes, the Saturn was a different beast. The Saturn also suffered from the PS2 problem, not having a decent developer toolkit. I think the PS2 was floundering hard but the hype was unreal and the dreamcast could not match the PS2 hardware. The dreamcast did offer Windows CE and as such, Microsoft's DirectX APIs.. but at the time PC gaming wasn't as big and console-only developers did not see value in porting between the two. IIRC only a handful of third party games relied on CE. There's probably a lot more I'm glossing over.

This was in 1999, and definitely the DreamCast. I was using one of the Katana dev kits. But I may be misremembering about the the plane-based rendering; I can't seem to find any references to that now.

The Saturn hardware was a mess because of last-minute additions welded on in response to their “Oh Crap!” moment of seeing what the PS1 could do.

The Saturn was made to be the ultimate 2D games machine. It had incredibly powerful 2D sprite hardware. But, then the PS1 announced it was all about 3D! Suddenly the hardware engineers were ordered to retrofit 3D into their 2D beast. So, they tacked on a second CPU and some other chips as fast as they could. If you open one up, you’ll see wires haphazardly stretched across the boards connecting the extra hardware to the original design.

The Saturn’s “3D rendering” was a clever desperation move. Turns out, with powerful enough 2D sprite rendering hardware, you can scale, rotate and shear a square 2D sprite into the distorted configuration that is equivalent to an 3D transformed rectangle viewed in perspective…. So, just tell the game developers to build their 3D scenes out of sprites instead of triangles!

As far as I know no Saturns have "wires stretched haphazardly across the board", at least mine don't. Perhaps really early Japanese models, but random bodge wires tend to get fixed pretty quickly for later revisions of PCBs, and like most games consoles the Saturn has quite a few variants through its production run.

For example very first revision of the Mega Drive in Japan did have an extra PCB and wires to fix a bug though. But only the first revision, it went away by the second.

Indeed. The Saturn didn't REALLY render "quads", but just transformed and translated sprites. That lead to all sorts of strange concepts. You didn't have UVs; a "quad" "texture" covered the entire quad from corner to corner and you couldn't change that. That lead to situations like https://www.youtube.com/watch?v=WDJgeuoaSvQ where the programmer wanted a simple environment mapping effect, but it was impossible with the Saturn hardware so he ended up writing a 3D rasterizer in software to enable the effect on the start screen.

If you want to succeed in consumer electronics (or consumer anything), you need a great product AND great marketing. There are plenty of examples of failed consoles that lacked one or both.

For example: The Turbografx-16 was a pretty nice system with awesome games that did well in Japan but had next to no presence in the US. Very few people owned one as their only system, so it was mostly bought by video game aficionados who had to have every system. (And I guess reviewers for video game magazines.)

The 3DO by contrast had great marketing (lots of commercials, ads in every magazine) but had expensive, lackluster hardware and only a few decent titles. Although to be fair it had higher ambitions than just being a video game console, so that identity crisis was one of its major struggles.

Success needs great vision of what can be in the near future, and carefully pushing that as far as possible but not letting "feature bloat" push it too far out. You can find this vision in both marketing and in engineering teams. If you are great at engineering (likely if you are reading this here), then you can hire people in marketing to sell whatever you create; while if you are great in marketing you can hire engineers to create your vision. Either works.

That the vision seems to have been in marketing in Sega's case shouldn't take away from all the engineering effort that went in as well, but the vision, and thus credit, is on marketing making the right decisions. There are plenty of other engineering companies that likewise have a good vision.

The argument is that Sega's global success with the Genesis came from its products - not the marketing (despite what the marketers say). One piece of support for that is that the console succeeded against Nintendo in most territories, despite all being under different marketing leadership and following different strategies.

When the Sega of America marketers were put in charge of product development and stopped relying on Japanese products, they were unable to replicate their success. They followed questionable practices derived from the toy industry (from where CEO Tom Kalinske emerged), such as a heavy emphasis on expensive licensed titles over original franchises.

>For some reason the marketing guys are always the key to the success,

I think this has a lot to do with the personality types that are attracted to sales/marketing vs hardware/software developers. One group is typically outgoing, bombastic, in your face, talk your ear off, no problems bending the truth to fit the narrative while the other group is not. (care to guess my sentiments on marketing/sales types ;-)

The number of passionate developers that can talk about their project in compelling way to make the average person interested is pretty small compared to the number sales people that will make you absolutely want to buy something while at the same time have little to no actual understanding of what they are selling or care to.

> I'm always dubious of these claims. For some reason the marketing guys are always the key to the success, and nothing else barely matters. In these "documentaries" the hardware could have been terrible, hell they could have been selling bananas, but if the marketing would have just been tweaked a bit, it would all have worked out.

IMHO, there can be multiple "keys" to success. The hardware has to be at least adequate (and IIRC, the Genesis beat the SNES to market by a couple years), and so does the marketing.

I've never seen that documentary, but it's possible Sega's marketing for the Genesis was relatively better than average, compared to the hardware.

Reply on news.ycombinator.com