There were two really "serious" plotlines from Transformers, IMHO: the movie plotline, and the return of Optimus Prime -> the end of the original Transformers (incl. the Headmasters thing...). The movie plotline I discussed elsewhere here: mainly, it's amazing for as much as they changed things in the first ten minutes, mainly (in my optimistic hope) to get to a solid footing with which to base the rest of the film.
The second plotline, the Headmasters plotline and the return of Optimus Prime has a few interesting choice ideas in it. Daniel (Spike's son) is paralyzed in the Headmasters plotline, and they have to carry him around and take care of him. In the end, he ends up being invaluable (as he gets fitted into an exo-suit allowing him to move again). The Return of Optimus Prime (first and second-fold) have some interesting dynamics as you see Rodimus Prime clearly hesitant about Optimus Prime returning, even though everyone else is overjoyed.
There WERE some interesting "first half" original Transformers episodes, but many were one shot, never hear from it again, nothing changes in the world plotlines as seen in so many cheesy anime flicks of the time (AND in Voltron, which was even worse than Transformers for this - how many Robeasts did they fight, for chrissakes?). I personally thought the "second half" episodes were much more intriguing (there was one episode where Galvatron had to "reform" in order to escape a planet. That was damn funny).
You're also forgetting Eric Idle (I think...?) - whoever was the voice of Wreck-Gar, that is.
My personal opinion about TF: The Movie was that the people who wrote the script looked at the cartoons, and said "how the HELL do we write a coherent plot with decent dynamics with these characters?" I mean, seriously, Optimus Prime has little/no flaws, other than being overly trusting (though that always seems to benefit him in the end), and most of the other characters have very little unique personality (other than muttering a few choice phrases unique to their character).
So, what do we do? In the first 10 minutes of the film, we kill EVERYONE off except a few people we leave around for continuity, and we replace them with a cast of characters that has a much more dynamic set of personalities.
Optimus got replaced with Hot Rod (Rodimus) who is naive, eager, a little too hotheaded, but generally good-natured. I always liked Rodimus and I was really upset when Optimus came back and took the Matrix from Rodimus...
Megatron got replaced with Galvatron, who didn't have a whiny voice (big plus) and showed a BIT of personality: megalomania, mainly, but a bit of humor ("Coronation, Starscream? This is bad comedy.") - seriously, listen to the way Megatron speaks ("Die, Autobot!") and the way that Galvatron speaks ("It's a pity you Autobots die so easily... if it were harder, it might be enjoyable." - or something like that).
The new characters added all had quite a bit of personality to them as well, which is what made it interesting. There was weird sexual tension between Arcee, Springer, and Hot Rod, there was a wizened father figure with Kup, and Ultra Magnus was the perfect example of an uncertain leader. Grimlock mainly stayed in for comic relief along with a bunch of new guys (Wreck-Gar, Blurr - god, how I wanted him to die..., and Wheelie) whose presence was solely for amusement. Plus Unicron, which was a very good overarching all-powerful evil force. It was really impressive to see Galvatron basically wallowing in misery as he realized that there was something far more evil and destructive in the universe than his own quest for power.
Note that I'm not saying that there wasn't some dynamism in the old characters: it's just that certain characters were just there to serve functions and to show up from time to time, rather than actually have some sort of personality, whereas the cast in the movie was almost solidly there for personality or comic relief (save Blurr. Kill Blurr. I hated Blurr.)
Of course, that's the optimist in me. Everybody knows the real reason they axed everyone was so they could make more money with new toys.
If you added a "blanket" filter which removed an equal amount of light across the visible band, and brought the amount of light down a significant amount (probably about one ten-millionth of the current amount of light) yah, it'd look blue-green, but only really really barely. The sun's spectrum is pretty flat in the visible, but the Sun puts out more energy in the blue-green band than any other (you can figure this out from Wien's law: take the temperature of the chromosphere, 5800 K roughly (I'm within 100 K here), and then the wavelength is 0.29 cm/T and that's roughly 500nm, which is dead-set blue-green.
Note that about 50% of physics books and Web pages get this wrong, but I'm right on this - do the math if you want to check it yourself. For instance, a quick search on Wien's Law actually returns http://scruffy.phast.umass.edu/a114/math1/node6.ht ml, which says that 500 nm is the wavelength of -yellow- light, which is, of course, wrong. The Sun appears yellow because all of the blue is spread out along the sky. And the sky appears blue because the Sun puts out more blue than violet.
Now, as to what 'color' the Sun is, honestly, the best answer is "white" - you MIGHT be able to get it to look blue-green just barely if you knock down the light level a LOT, but you wouldn't be able to do it enough that your eyes could see it.
I doubt that's true: it's definitely true that the average life span was easily half what it is today: probably worse. But that's because infant mortality really lowered the average quite significantly. If you correct for that, and only take into account deaths from non-infectious diseases, I'd actually guess the average lifespan isn't THAT significantly increased: probably 5-10 years would be my guess.
Not arguing the conclusions, just arguing your facts - most of the herbal remedies were targeted at generally increasing health, which is a difficult thing to prove or disprove using large-scale averages. The herbal remedies, I'm sure, are mostly all crap. I just wish more of them actually underwent clinical trials - even if they failed them, they could still push the remedy for its nutritional values (weak as they are).
(score +1 bonus dropped because I know it's offtopic:) )
Blue sky? It's a result of all the water in the air.
Actually, the sky isn't blue. It's purple. We just happen to be looking at it through a blue/green light.
(whee, proof for those who'll jump on me: the sky bends purple/violet light more than blue light, so if equal amounts of every visible wavelength was emitted into the atmosphere, the sky would look purple more than blue. However, the sun peaks in the blue-green - yes, the Sun is blue-green, not orange - so there ends up being far more blue light scattered in the atmosphere than purple light)
Yah, same here in PA on car sales. I think the law's kindof ambiguous on sales tax, to be honest, because you probably don't earn large amounts of money doing it, so there's no way they can find out if you did sell something. I imagine that people who sell large numbers of things on eBay may have to report it and eventually may have to pay sales tax (if they ever started enforcing state sales tax on the Internet). The difference with cars is that since you need to transfer over the title and change the registration, the state knows when cars are bought and sold (for the most part) and so they force the tax issue then. In any case, the car transfer thing is just because states have a tax on the sale of all cars. I think the same thing would apply to anything that's resold if there's a tax on it (though there aren't many of those: houses and cars, basically).
I do agree with you that his arguments are a bit weak - considering they don't flow directly from the facts, he probably should've supported them a bit better. To a reader, it looks more like fluff than the actual truth.
If you sell something taxable 10 times, you still pay 10 sets of sales tax (in theory - you would with used books, at least). The DMV has a tax on "the selling of cars". Every time the car is sold, it's taxed. The publishers can't tax, because, well, they're not the government. If they were the government, they could declare a "book tax". Yah. Like that would fly. The publishers sell something: they sell a book. They don't "license" the book. People buy the book. They then have the right to do with it whatever they want. Sorry, Author's Guild, but fair use for books has been around for a long long time, and no court would ever take it away.
Bezos's arguments are valid, though a little vague because he didn't want to get into the monetary details. Any economist will tell you that expanding a market can only have positive effects for those providing the product being sold in the market. If you make books more accessible to people, and get people to buy books, there are more people out on the market to buy books. If the publishers complain "well, people will only buy used books, and not new books!" it's because in the market, the publishers are being beaten out price-wise. Tough. Suck it up and deal. Lower your profit margins, work on efficiency, get the price down.
So while I'll agree that Bezos's arguments are a little, well, public-relationized, they're still valid. It's not an argument that'd stand in court, but with a economists on the stand afterward, you could prove it.
Then again, Napster could've made the same argument, and that didn't fare that well.:)
Because if you're trying to contact other civilizations, the 21 cm line makes sense. It's a scientifically interesting wavelength (Neutral hydrogen has a forbidden line at 21 cm) and the Universe is basically transparent to it.
The reverse argument, that we don't transmit at 21 cm, why should they, which implies if we aren't trying to contact them, why are they trying to contact us, is a valid criticism. I think SETI has basically acknowledged that unless an alien civilization is trying to contact us (or any other civilization), we won't contact them. It's fairly evident that within a few years, "leaking" radiation from Earth will be indistinguishable from noise, so we can expect other civilizations did the same.
This is why if SETI fails, it does not imply there aren't any other civilizations out there. It just implies that none of them are far more gung-ho about contacting other civilizations than we are.
Just a few bits... (glad to know someone else hates a UMA design as well)
The main problem with the pixel shader is the fact that it's a graphical tweak, like antialiasing, for the most part. If a game is lacking it, while you might notice that a few of the textures look a little polygonal, and don't bend right, it's not a fatal flaw of the game. I think that the UMA architecture will provide more problems to a game architect than not having a pixel shader. Though I do agree with you that in time, they may get very good at using them. I just don't think that there's as MUCH room for growth on the X-Box as there is on the PS2 and the GC. The PS2 is so bloody complicated that figuring out how everything works could take a lifetime.
It's just that if you look at the three systems, here's the way it works out, IMHO:
PS2: Huge vector processing capabilities. There's a lot which can be done here, but you need to work at it. Apparently developers aren't having that much of a problem...
GC: Unique memory design. As developers push more and more polygons on screen, the interesting bandwidth-conserving and low-latency functions of the GC may be extremely helpful.
X-Box: Extremely conservative design: nothing really unique. Strengths are the hard drive (not a strength, IMHO: a stupid idea. Hard drives fail irrepairably, and people do NOT want dying consoles like dying PCs), built in Ethernet, and the high quality Dolby sound (I can't remember what it is...) "Modern" graphics chipset gives it a leg up on the other two consoles, but only if the features on its chipset are hugely noticeable and can't be done on other consoles. There is no way in hell I'll give X-Box a nod for a "faster processor", it's an x86, and it has all the problems that that architecture brings with it.
Personally, I think the X-Box processor choice exemplifies what a bad job Microsoft did - it was a lazy choice. A good software console SDK will entirely hide the architecture choice of the processor. Microsoft's SDK for the X-Box is basically DirectX, and the fact that they used an x86 just means they didn't want to port DirectX to another architecture. x86 is not an ideal choice for multimedia applications that stream a large amount of data.
Also, I think MusyX is a great example of exactly what I'm talking about. Factor 5 originally developed MusyX for the N64 (for Rogue Squadron, I think) several years into the N64's lifespan, and suddenly the audio capabilities of the N64 jumped a huge notch. Somehow I don't see that happening with the X-Box.
By the way, just to add, that "microwave frequencies are absorbed by water" bit is just junk - GHz-ish frequencies are roughly foot-long in wavelength (1 GHz = 1 ns = 1 foot, roughly. 2 GHz = 1/2 ns = 1/2 foot), and water molecules are surely not that long at all. Microwaves work because they're flooding a compartment with radiation, not because they're using frequencies that water alone absorbs. Put a piece of metal in a microwave and turn it on - you'll get a heckuva light show because metal is very conductive and absorbs the energy very quickly.
Anyway, I'll buy an exponential decay if you've got a bunch of absorbers in line (houses, etc.) but it's definitely 1/r^2 in unimpeded space.
Just wanted to make sure that the 'water myth' doesn't get propagated much more.:)
For the most part, I was talking about very wide band stuff (I guess "extreme spread spectrum" would be a better term for frequency hopping over a large range?), so that's what I meant by that. I'm mainly against the idea of using signals which have a wide frequency range over a blanket area of the environment (i.e. no 'radio pollution' to compound light pollution). If you're going to use a very directional antenna to link two areas, that's... probably fine. Radio antennas aren't THAT directional (even good Yagi antennas still have what, a 60 degree front lobe or something like that?) and so I think wires over long distances are probably best.
So if you're saying "spread spectrum is great, should be limited to personal/home use, rather than big huge towers broadcasting" yah, I agree with you totally.
I think an ad-hoc network will actually spring up quite naturally over time - if I like my neighbor, and trust him, I might actually talk with him and set up a wireless point in my house that can touch a wireless point in his house, and possibly share bandwidth, or if both of us have an ISP, at least be connected. Granted, you have to worry slightly about security and things like that, but it's dealable.
So here, you'd have a spread spectrum 802.11b network over an entire neighborhood, but the noise floor wouldn't be touched, except for 2.4GHz using stuff, and that's all personal devices. Outside of people's homes, it probably would be negligible, because you'd want the inter-home links to be minimum power links. For that you could do something like a hybrid directional/omnidirectional antenna (ones that have a strong lobe in one direction, but still broadcast in all directions) so you would be okay anywhere in your own home, and you'd be able to link to a neighbor.
Here's a thought. We already do know that raising the noise floor by careless use of devices happens. It's called light pollution. Hadn't thought of it that way before...
The reason I said that wireless priority isn't possible is although you could prioritize things inside the actual protocol itself, if you're using an ad-hoc type network, you have to make sure everyone in the network is using the same type of protocol, or if there's enough hops, that you can route enough ways to get to where you want to go. It's doable that way, but I think most emergency services would prefer a dedicated frequency.
What's causing the exponential decay in GHz frequencies? It should be 1/r^2 with distance unless something's absorbing it, and I don't know anything in the atmosphere that's got a diameter of a few inches to a foot (GHz-ish wavelengths). If you put a tower up on a mountain, and broadcast at GHz frequencies, you should get 1/r^2 with distance.
The thing I don't like about spread spectrum is that it sucks up a large portion of the spectrum, rather than individual frequencies. That means that for scientists observing in those bands, if you allow stuff like cell towers using really spread spectrum stuff, it's going to not be good for scientists.
But then, in my opinion, the US is getting way too "radio loud" in areas where we don't need to be, just like light pollution is getting way out of hand. People who put up anything that radiates really should be taking care to ensure that it only goes where it needs to go, and nowhere else.
The funny thing is that if you do this correctly, it saves money. You know all of those gas stations and parking lots that use these HUGE bulbs that broadcast light everywhere? Why don't they just use a lower power light, and direct the light where they want it to go? That and use a surface that's not very reflective, and boom, instant light pollution reduction. Same idea for radio transmission. If you're trying to get from one point to another, you should definitely be using very directional antennas.
Don't get me wrong. I'm not against all usage of the EM spectrum like the "EM radiation is going to kill you!" wackos. I'm just of the opinion that people should be a hell of a lot more careful than we're even being now.
You're exactly correct. This is a fundamental question that I have. SETI would not find the Earth.
We've tried to contact other planets, but we've contacted them for only a ridiculously short period of time (like, 5 years or something like that).
My personal opinion is that we're not going to find life and civilizations out there until we GO out there and look. And then I think we're going to find out that "where there's water, there's life."
"Light" usually implies "electromagnetic radiation". If you meant visible light, you should've said visible light. I was just saying that your terminology was a little ambiguous.
Interstellar absorption is VERY important on local star group scales! Anything smaller than about 1 micron is basically unusable on space scales: it's a power concern. For anything less than 1 micron, you have to transmit MORE power to get the same signal to the receiver. In that case, you'd be an idiot to use that frequency and not the clearer, low power one.
There's a reason we use radio to communicate with satellites and space probes - because it's intelligent, and far easier than visible light (aka lasers). While they MIGHT use lasers, they'd probably only do it for point-to-point connections (planets) and not for spacecraft.
Radio will never go away, not until we find a new medium to broadcast through. And in other civilizations, it won't go away either. It's fundamentally a good region of the frequency spectrum to use.
Personally I think the solution is tightly collimated beams and making sure that you're not leaking radiation everywhere. Broadcast spread spectrum is annoying - it does raise the noise floor. If you're talking about for IP services, and you have highly directional antennas, yah, spread spectrum should be OK. Less interference, more stable signal, should be good. But if you're talking about for cell phone use, no way, and unfortunately, I think that's in general what people want. Put a big tower broadcasting spread spectrum, and that WILL raise the noise floor. Broadcast from inside your house, and no, that won't.
You can't give wireless things priority, unfortunately. Wireless is wireless, and the only way to make certain things a priority is to use a different frequency. The problem with spread spectrum is that if you're allowed to broadcast it (the cell tower thing I was talking about) then you futz EVERYONE's frequencies, and it really sucks.
This only applies to ultra spread spectrum, in any case - the stuff that's spread over many GHz of frequency space. Ugh.
Be careful - the raising of the noise floor idea is real, and it will suck. But you are right that for personal things, it's fine - even an ad-hoc wireless over an entire city would be fine, because probably the connections between each "house" would be weak, but inside the house they'd be fine.
The worry is for commercial services, and people broadcasting (that is, intention omnidirectional broadcasting) - that will raise the noise floor.
If we set up an ad-hoc wireless network across a city, that's cool. If everyone decides to broadcast their own spread-spectrum 'SSFM' radio station, that would suck.
Um. Radio is light is electromagnetic radiation. But radio has a benefit, because it is not absorbed by air. Why? Because it's long-wavelength. Basically any gaseous atmosphere is going to be transparent to radio waves, so they are probably using that - probably a liquid environment will also be transparent to it.
You're also wrong that any form of EM radiation travels well in space - dust really sucks, and it preferentially absorbs higher wavelengths (because the dust can absorb them). Radio travels well in space, light does not.
Besides, SETI is looking at an 'intelligent' portion of the spectrum (I believe... they may have switched) - the 21 cm line of hydrogen. We can't look at any large portion of the spectrum - that's really friggin' difficult. So we assume that if they're trying to contact us, they're using an intelligent wavelength.
SETI isn't looking for stray communications, in any case. They're looking for a signal intended for us to notice. If SETI fails, that doesn't mean that there isn't anything out there - it simply means they aren't trying to contact us in the way we think they are, OR they aren't trying to contact us at all.
That's the point. And you honestly think that JSRF will remain an X-Box exclusive? Sega says it will. But I'd bet large amounts of money that Sega got money to make it an X-Box exclusive, and it was probably for a set amount of time. How long will you consider waiting to buy an X-Box plus JSRF? A few months? In a few months, they'll probably announce a Jet Set Radio sequel (probably not 'Future', probably 'Cubed' or 'Extreme', for GC and PS2 respectively:) ) for another system. Then what do you buy? For the PS2, well, um, you probably already have one of those. For the GC, you KNOW games that come out on it won't come out on the X-Box, but it certainly looks like most games that are worth buying that come out on the X-box also come out on the GC - so what do you buy? Probably the GC.
It's not how you support the X-Box when its hype is high and it's the bee's knees. It's how you support the X-Box when developer support is starting to lag, you've got competing consoles out there, and the system's starting to show it's age.
Fast forward a year or so. Sony announces the successor to the PS2, the PS3, and is generating a lot of hype for it. Pressure's on Nintendo and Microsoft to announce a successor. Do you do it? Or don't you? And then, if you do, how many dev teams do you pull from your current system to work on the next system?
If you're Nintendo, you don't announce a successor. You hint at it, but keep the current system your focus. You shift the top dev teams over, but really push for second-party and third-party support, possibly with one or two last real kick-ass games for the system. With the N64, this was Zelda: Majora's Mask, the Mario Party set, Super Smash Brothers, and the Pokemon series. (Yes, SSB was produced by a different group, but they probably got the license for Nintendo characters for a song and dance). THIS is how you keep/win back developer support (the reason they lost dev support with the N64 was the cart-based format).
If you're Microsoft, who knows? Microsoft has a tendency to ditch projects when they suck too much (see also early Windows CE devices, and the whole WebTV/UltimateTV fiasco). They need to ditch this attitude, really bad.
If you're Sega, you abandon the system, say no more will be made, and the system dies a horrible screaming death as many developers line up to strangle CEOs. And then you abandon being in the hardware market altogether.
If you're Sony, you do the same thing Nintendo did (for the most part - they announce the system, but they always announce backwards compatibility. They better not ditch this for the PS3). See? This is how the market works. I worry that Microsoft won't see that, and in the next generation, we'll see them fold back into being a games developer only.
He was saying that the X-Box and PS2 compete, not the PC competes with the PS2.
My personal opinion is the X-Box and the PC are too similar for them NOT to compete - games aren't significantly different between the two (at least not yet) and thanks to Halo and a couple other of the really good games for X-Box, most people believe that games that come out for the X-Box will eventually come out for the PC. Most gamers will wait the few months if they KNOW it's coming out rather than plunk down $300.
So I personally think the X-Box has more to worry about from PCs than it does from the PS2. That's what he was arguing with.
$300 will almost buy you a PC with better specs than the X-box. I know I could probably do it with $500, $400 I'd probably have to compromise a few things. In half a year or so, I could do it with $400 definitely. Right now I could almost definitely UPGRADE a recent (2-4 years) PC to X-box status with $300.
As for online, here's the problem though: how many people are still playing Half-Life, a game that's what, 3 years old at this point? This is NOT what you want for a game system. You need to sell new games, keep games being sold, etc. to get money flowing in, because it takes a while for the system to be profitable.
That being said, MS could probably work out a strategy to make online gameplay work, so whatever. But anyway, I don't know - I personally think single-player games will garner a LOT of the market and still drive things more than multiplayer online games. The highest selling PC games are still single player games (I think).
I'd also like to point out that online games basically require a keyboard to be fun. The benefit is that you're playing against other people, rather than the computer, so communication, strategy become important. Keyboards will suck with a console (IMHO) - I'd just not bother.
See the other posts. As was pointed out (not even by me, as I'm not a financial expert) - that's not cash. That's marketable securities, so while they have 'book value' of $38B, they don't actually have $38B cash in hand.
That said, you can't buy certain companies. You can buy the company, but the talent may or may not come with it, especially if it's an unwanted buyout. With game developers, you're MUCH better off saying you'll give them lots and lots of money "if they don't develop for other systems" - but the developers can still CHOOSE, and if X-Box is dead in the water, they probably won't take any amount of money, because they want to see their game succeed, not flounder.
MS is not going to win this with money. The only way they can win this is with 1) good games, and 2) exclusive games (lots of games doesn't help: if the PS2 has the games already, the PS2's got the mindshare, and people will buy the PS2) and 3) a LOT of hype. They don't have 1, they don't have 2, and they're losing 3. Game over, Microsoft. You lost this round. Stick in the game, learn the rules, and maybe you'll play better in the second round.
Yah, I was mainly talking about graphics bandwidth, which, for the GC, is guaranteed, whereas for the XBox, you'll get stutters and hiccups when you start doing many things in the system.
Latency for 1T-SRAM is also much better than standard DDR-SDRAM. DDR-SDRAM (if memory serves - correct me here, but I think I'm within about 10%) is typically roughly 10-clock latency (maybe 7 or so - I know definitely what it is for SDRAM, but not for DDR) which is going to be about 50 ns or so for a 133MHz bus (is it 133 or 100 in the X-Box?). That's not actually what the latency is though, because it's actually a gaussian with multiple peaks when you measure it: one at "normal", one at "crossing row boundaries" (probably about +1 or 2 clock cycles), and one at "refresh cycle during read) - probably about +8 or more clock cycles. That means that your latency is only "guaranteed" probably about 100 ns: ow! 1T-SRAM, however, has a guaranteed latency of 10 ns (there's probably a tail for "crossing row boundaries" as well - not sure, MoSys is a little guarded to the general public about how things work - but not for refreshes, which don't happen). Which is, well, way cool. The benefit here is that a system with 1T-SRAM *knows* that a memory access is going to be within 10 ns, which is damn impressive. I might be off a bit here (MoSys doens't list maximum worst-case latency, but Nintendo quotes approximate sustainable latency - my guess would be that the worst-case latency is probably about 30-40 ns, but I'll assume Nintendo meant worst-case latency), but the lack of memory refresh really helps. This means that you can really push the limit of the CPU and graphics chips - much more than on the X-Box.
Personally I think a UMA was an amazingly stupid architecture for a video game system - since NUMA is already used for something else, I think "anti-UMA": that is, a system in which basically NONE of the memory is unified (that is, you have CPU RAM, sound RAM, graphis RAM, and then huge buffers on both the CD-ROM and any network adapter - large amounts of texture RAM and frame buffer RAM, etc. - you get the idea) is probably more intelligent - which is basically what the GC is. The reason is because UMA is basically a simplicity and compromise architecture. You're saying "well, in general, accesses are going to mainly be from CPU to RAM, and graphics doesn't need memory all that much, so you're okay just using all the same memory" - but you DON'T want that in a console. The reason they had to have a processor with so much higher clock speed is because they're using such an inefficient architecture that they need to give themselves a much higher headroom to get the same performance. It's really bizarre, in my opinion - UMA doesn't make any sense for a game console, which you want to be able to use all of its resources maximally, otherwise you're wasting money - which is what the X-Box is doing (wasting money).
I wasn't trying to say whether or not the X-Box is more powerful than the GC or not: what I was trying to say was whether or not developers will need time to figure out the system. In the GC's case, I think that's definitely true - developers have probably never worked with a system that's so disjoint like this - they may need time to figure out which accesses go at what time to get optimum speed. But the X-Box is a standard system: it's a P3 on a UMA, with a GeForce 3 graphics card (roughly). They're used to systems like this, especially Microsoft and all the PC developers who port games over. I don't think there will BE any 'tricks' that they can learn over time to make the system faster.
Remember, with the N64, Nintendo did bizarre stuff like allow the game to rewrite microcode in certain chips on the board: this kindof stuff is really weird, because you can say "ok, I'm NEVER going to use this function, so I'm going to use the space it was taking to streamline and speed up this other function" and things like that - that's what Factor 5 did with their MusyX engine for the N64, and it ended up producing incredible sound with little CPU overhead and really good compression (factors of 10 or better were quoted) - which is how you ended up with games like Conker's Bad Fur Day, which had literally tons of voice in the game, in a cart! I don't know if Nintendo did this with the GC, but I'd bet they did.
My point basically is that I think Microsoft made a very stupid decision creating the X-Box: using standard off-the-shelf components is how you build a general purpose PC, NOT a specific-function game console. I think that developers may figure out how to put a Web browser into their games, but I don't think they'll figure out how to make their games 10x faster and better looking.
I could be wrong. We'll see what 2nd and 3rd generation X-Box games look like compared to 2nd and 3rd generation GC games (well, see the "% difference" comparison).
I definitely agree with you on the price point and games as well, though. The GC has more chance of coming down in price lower than the X-Box, IMHO: the Pentium III that the X-Box is using is probably already as dirt cheap as it's ever going to be. Nintendo knows what it's doing. It debuted a system at its normal price point (I think? N64 debuted at $199, right? that's what I bought it at when it debuted, I'm pretty sure...) and in time, will drop the price to $175, then $150, then $99, and by then the system will be quite good with a lot of games. The X-Box is going to have a harder time coming down in price, and I think that's going to hurt Microsoft just as much as anything else. (I think I saw somewhere that the GC in a few months will be break-even on its price for Nintendo - dunno, though)
Sustained, not theoretical maximum. The main reason for this is that the GC uses MoSys 1T-SRAM, so when it wants bandwidth, it gets it, sustained, low latency. X-Box is just a normal UMA, and people've been working with that a long time. 1T-SRAM's new. I think in time people will realize that with the GC they can keep the pipe full for basically forever, and games will change to appreciate that. I'll be surprised if people "find things" in the X-Box they didn't think was there originally.
That's exactly what I said - they alienated large portions of the developers because they didn't follow the PSX's lead for the N64. Sorry if it sounded a little confusing.
There's a good chance that the N64 would've taken the lead from the PSX had it been a CD based system. It also wouldn't've been the same system, so I wouldn't've gotten several kickass games, so, not sure which one is better.:)
Slashdot Mirror
There were two really "serious" plotlines from Transformers, IMHO: the movie plotline, and the return of Optimus Prime -> the end of the original Transformers (incl. the Headmasters thing...). The movie plotline I discussed elsewhere here: mainly, it's amazing for as much as they changed things in the first ten minutes, mainly (in my optimistic hope) to get to a solid footing with which to base the rest of the film.
The second plotline, the Headmasters plotline and the return of Optimus Prime has a few interesting choice ideas in it. Daniel (Spike's son) is paralyzed in the Headmasters plotline, and they have to carry him around and take care of him. In the end, he ends up being invaluable (as he gets fitted into an exo-suit allowing him to move again). The Return of Optimus Prime (first and second-fold) have some interesting dynamics as you see Rodimus Prime clearly hesitant about Optimus Prime returning, even though everyone else is overjoyed.
There WERE some interesting "first half" original Transformers episodes, but many were one shot, never hear from it again, nothing changes in the world plotlines as seen in so many cheesy anime flicks of the time (AND in Voltron, which was even worse than Transformers for this - how many Robeasts did they fight, for chrissakes?). I personally thought the "second half" episodes were much more intriguing (there was one episode where Galvatron had to "reform" in order to escape a planet. That was damn funny).
You're also forgetting Eric Idle (I think...?) - whoever was the voice of Wreck-Gar, that is.
My personal opinion about TF: The Movie was that the people who wrote the script looked at the cartoons, and said "how the HELL do we write a coherent plot with decent dynamics with these characters?" I mean, seriously, Optimus Prime has little/no flaws, other than being overly trusting (though that always seems to benefit him in the end), and most of the other characters have very little unique personality (other than muttering a few choice phrases unique to their character).
So, what do we do? In the first 10 minutes of the film, we kill EVERYONE off except a few people we leave around for continuity, and we replace them with a cast of characters that has a much more dynamic set of personalities.
Optimus got replaced with Hot Rod (Rodimus) who is naive, eager, a little too hotheaded, but generally good-natured. I always liked Rodimus and I was really upset when Optimus came back and took the Matrix from Rodimus...
Megatron got replaced with Galvatron, who didn't have a whiny voice (big plus) and showed a BIT of personality: megalomania, mainly, but a bit of humor ("Coronation, Starscream? This is bad comedy.") - seriously, listen to the way Megatron speaks ("Die, Autobot!") and the way that Galvatron speaks ("It's a pity you Autobots die so easily... if it were harder, it might be enjoyable." - or something like that).
The new characters added all had quite a bit of personality to them as well, which is what made it interesting. There was weird sexual tension between Arcee, Springer, and Hot Rod, there was a wizened father figure with Kup, and Ultra Magnus was the perfect example of an uncertain leader. Grimlock mainly stayed in for comic relief along with a bunch of new guys (Wreck-Gar, Blurr - god, how I wanted him to die..., and Wheelie) whose presence was solely for amusement. Plus Unicron, which was a very good overarching all-powerful evil force. It was really impressive to see Galvatron basically wallowing in misery as he realized that there was something far more evil and destructive in the universe than his own quest for power.
Note that I'm not saying that there wasn't some dynamism in the old characters: it's just that certain characters were just there to serve functions and to show up from time to time, rather than actually have some sort of personality, whereas the cast in the movie was almost solidly there for personality or comic relief (save Blurr. Kill Blurr. I hated Blurr.)
Of course, that's the optimist in me. Everybody knows the real reason they axed everyone was so they could make more money with new toys.
If you added a "blanket" filter which removed an equal amount of light across the visible band, and brought the amount of light down a significant amount (probably about one ten-millionth of the current amount of light) yah, it'd look blue-green, but only really really barely. The sun's spectrum is pretty flat in the visible, but the Sun puts out more energy in the blue-green band than any other (you can figure this out from Wien's law: take the temperature of the chromosphere, 5800 K roughly (I'm within 100 K here), and then the wavelength is 0.29 cm/T and that's roughly 500nm, which is dead-set blue-green.
t ml, which says that 500 nm is the wavelength of -yellow- light, which is, of course, wrong. The Sun appears yellow because all of the blue is spread out along the sky. And the sky appears blue because the Sun puts out more blue than violet.
Note that about 50% of physics books and Web pages get this wrong, but I'm right on this - do the math if you want to check it yourself. For instance, a quick search on Wien's Law actually returns http://scruffy.phast.umass.edu/a114/math1/node6.h
Now, as to what 'color' the Sun is, honestly, the best answer is "white" - you MIGHT be able to get it to look blue-green just barely if you knock down the light level a LOT, but you wouldn't be able to do it enough that your eyes could see it.
I doubt that's true: it's definitely true that the average life span was easily half what it is today: probably worse. But that's because infant mortality really lowered the average quite significantly. If you correct for that, and only take into account deaths from non-infectious diseases, I'd actually guess the average lifespan isn't THAT significantly increased: probably 5-10 years would be my guess.
Not arguing the conclusions, just arguing your facts - most of the herbal remedies were targeted at generally increasing health, which is a difficult thing to prove or disprove using large-scale averages. The herbal remedies, I'm sure, are mostly all crap. I just wish more of them actually underwent clinical trials - even if they failed them, they could still push the remedy for its nutritional values (weak as they are).
(score +1 bonus dropped because I know it's offtopic :) )
:)
Blue sky? It's a result of all the water in the air.
Actually, the sky isn't blue. It's purple. We just happen to be looking at it through a blue/green light.
(whee, proof for those who'll jump on me: the sky bends purple/violet light more than blue light, so if equal amounts of every visible wavelength was emitted into the atmosphere, the sky would look purple more than blue. However, the sun peaks in the blue-green - yes, the Sun is blue-green, not orange - so there ends up being far more blue light scattered in the atmosphere than purple light)
Good points, though.
Yah, same here in PA on car sales. I think the law's kindof ambiguous on sales tax, to be honest, because you probably don't earn large amounts of money doing it, so there's no way they can find out if you did sell something. I imagine that people who sell large numbers of things on eBay may have to report it and eventually may have to pay sales tax (if they ever started enforcing state sales tax on the Internet). The difference with cars is that since you need to transfer over the title and change the registration, the state knows when cars are bought and sold (for the most part) and so they force the tax issue then. In any case, the car transfer thing is just because states have a tax on the sale of all cars. I think the same thing would apply to anything that's resold if there's a tax on it (though there aren't many of those: houses and cars, basically).
I do agree with you that his arguments are a bit weak - considering they don't flow directly from the facts, he probably should've supported them a bit better. To a reader, it looks more like fluff than the actual truth.
If you sell something taxable 10 times, you still pay 10 sets of sales tax (in theory - you would with used books, at least). The DMV has a tax on "the selling of cars". Every time the car is sold, it's taxed. The publishers can't tax, because, well, they're not the government. If they were the government, they could declare a "book tax". Yah. Like that would fly. The publishers sell something: they sell a book. They don't "license" the book. People buy the book. They then have the right to do with it whatever they want. Sorry, Author's Guild, but fair use for books has been around for a long long time, and no court would ever take it away.
:)
Bezos's arguments are valid, though a little vague because he didn't want to get into the monetary details. Any economist will tell you that expanding a market can only have positive effects for those providing the product being sold in the market. If you make books more accessible to people, and get people to buy books, there are more people out on the market to buy books. If the publishers complain "well, people will only buy used books, and not new books!" it's because in the market, the publishers are being beaten out price-wise. Tough. Suck it up and deal. Lower your profit margins, work on efficiency, get the price down.
So while I'll agree that Bezos's arguments are a little, well, public-relationized, they're still valid. It's not an argument that'd stand in court, but with a economists on the stand afterward, you could prove it.
Then again, Napster could've made the same argument, and that didn't fare that well.
Because if you're trying to contact other civilizations, the 21 cm line makes sense. It's a scientifically interesting wavelength (Neutral hydrogen has a forbidden line at 21 cm) and the Universe is basically transparent to it.
The reverse argument, that we don't transmit at 21 cm, why should they, which implies if we aren't trying to contact them, why are they trying to contact us, is a valid criticism. I think SETI has basically acknowledged that unless an alien civilization is trying to contact us (or any other civilization), we won't contact them. It's fairly evident that within a few years, "leaking" radiation from Earth will be indistinguishable from noise, so we can expect other civilizations did the same.
This is why if SETI fails, it does not imply there aren't any other civilizations out there. It just implies that none of them are far more gung-ho about contacting other civilizations than we are.
Just a few bits... (glad to know someone else hates a UMA design as well)
The main problem with the pixel shader is the fact that it's a graphical tweak, like antialiasing, for the most part. If a game is lacking it, while you might notice that a few of the textures look a little polygonal, and don't bend right, it's not a fatal flaw of the game. I think that the UMA architecture will provide more problems to a game architect than not having a pixel shader. Though I do agree with you that in time, they may get very good at using them. I just don't think that there's as MUCH room for growth on the X-Box as there is on the PS2 and the GC. The PS2 is so bloody complicated that figuring out how everything works could take a lifetime.
It's just that if you look at the three systems, here's the way it works out, IMHO:
PS2: Huge vector processing capabilities. There's a lot which can be done here, but you need to work at it. Apparently developers aren't having that much of a problem...
GC: Unique memory design. As developers push more and more polygons on screen, the interesting bandwidth-conserving and low-latency functions of the GC may be extremely helpful.
X-Box: Extremely conservative design: nothing really unique. Strengths are the hard drive (not a strength, IMHO: a stupid idea. Hard drives fail irrepairably, and people do NOT want dying consoles like dying PCs), built in Ethernet, and the high quality Dolby sound (I can't remember what it is...) "Modern" graphics chipset gives it a leg up on the other two consoles, but only if the features on its chipset are hugely noticeable and can't be done on other consoles. There is no way in hell I'll give X-Box a nod for a "faster processor", it's an x86, and it has all the problems that that architecture brings with it.
Personally, I think the X-Box processor choice exemplifies what a bad job Microsoft did - it was a lazy choice. A good software console SDK will entirely hide the architecture choice of the processor. Microsoft's SDK for the X-Box is basically DirectX, and the fact that they used an x86 just means they didn't want to port DirectX to another architecture. x86 is not an ideal choice for multimedia applications that stream a large amount of data.
Also, I think MusyX is a great example of exactly what I'm talking about. Factor 5 originally developed MusyX for the N64 (for Rogue Squadron, I think) several years into the N64's lifespan, and suddenly the audio capabilities of the N64 jumped a huge notch. Somehow I don't see that happening with the X-Box.
By the way, just to add, that "microwave frequencies are absorbed by water" bit is just junk - GHz-ish frequencies are roughly foot-long in wavelength (1 GHz = 1 ns = 1 foot, roughly. 2 GHz = 1/2 ns = 1/2 foot), and water molecules are surely not that long at all. Microwaves work because they're flooding a compartment with radiation, not because they're using frequencies that water alone absorbs. Put a piece of metal in a microwave and turn it on - you'll get a heckuva light show because metal is very conductive and absorbs the energy very quickly.
:)
Anyway, I'll buy an exponential decay if you've got a bunch of absorbers in line (houses, etc.) but it's definitely 1/r^2 in unimpeded space.
Just wanted to make sure that the 'water myth' doesn't get propagated much more.
For the most part, I was talking about very wide band stuff (I guess "extreme spread spectrum" would be a better term for frequency hopping over a large range?), so that's what I meant by that. I'm mainly against the idea of using signals which have a wide frequency range over a blanket area of the environment (i.e. no 'radio pollution' to compound light pollution). If you're going to use a very directional antenna to link two areas, that's... probably fine. Radio antennas aren't THAT directional (even good Yagi antennas still have what, a 60 degree front lobe or something like that?) and so I think wires over long distances are probably best.
So if you're saying "spread spectrum is great, should be limited to personal/home use, rather than big huge towers broadcasting" yah, I agree with you totally.
I think an ad-hoc network will actually spring up quite naturally over time - if I like my neighbor, and trust him, I might actually talk with him and set up a wireless point in my house that can touch a wireless point in his house, and possibly share bandwidth, or if both of us have an ISP, at least be connected. Granted, you have to worry slightly about security and things like that, but it's dealable.
So here, you'd have a spread spectrum 802.11b network over an entire neighborhood, but the noise floor wouldn't be touched, except for 2.4GHz using stuff, and that's all personal devices. Outside of people's homes, it probably would be negligible, because you'd want the inter-home links to be minimum power links. For that you could do something like a hybrid directional/omnidirectional antenna (ones that have a strong lobe in one direction, but still broadcast in all directions) so you would be okay anywhere in your own home, and you'd be able to link to a neighbor.
Here's a thought. We already do know that raising the noise floor by careless use of devices happens. It's called light pollution. Hadn't thought of it that way before...
The reason I said that wireless priority isn't possible is although you could prioritize things inside the actual protocol itself, if you're using an ad-hoc type network, you have to make sure everyone in the network is using the same type of protocol, or if there's enough hops, that you can route enough ways to get to where you want to go. It's doable that way, but I think most emergency services would prefer a dedicated frequency.
What's causing the exponential decay in GHz frequencies? It should be 1/r^2 with distance unless something's absorbing it, and I don't know anything in the atmosphere that's got a diameter of a few inches to a foot (GHz-ish wavelengths). If you put a tower up on a mountain, and broadcast at GHz frequencies, you should get 1/r^2 with distance.
The thing I don't like about spread spectrum is that it sucks up a large portion of the spectrum, rather than individual frequencies. That means that for scientists observing in those bands, if you allow stuff like cell towers using really spread spectrum stuff, it's going to not be good for scientists.
But then, in my opinion, the US is getting way too "radio loud" in areas where we don't need to be, just like light pollution is getting way out of hand. People who put up anything that radiates really should be taking care to ensure that it only goes where it needs to go, and nowhere else.
The funny thing is that if you do this correctly, it saves money. You know all of those gas stations and parking lots that use these HUGE bulbs that broadcast light everywhere? Why don't they just use a lower power light, and direct the light where they want it to go? That and use a surface that's not very reflective, and boom, instant light pollution reduction. Same idea for radio transmission. If you're trying to get from one point to another, you should definitely be using very directional antennas.
Don't get me wrong. I'm not against all usage of the EM spectrum like the "EM radiation is going to kill you!" wackos. I'm just of the opinion that people should be a hell of a lot more careful than we're even being now.
You're exactly correct. This is a fundamental question that I have. SETI would not find the Earth.
We've tried to contact other planets, but we've contacted them for only a ridiculously short period of time (like, 5 years or something like that).
My personal opinion is that we're not going to find life and civilizations out there until we GO out there and look. And then I think we're going to find out that "where there's water, there's life."
If you could see radio, you'd see a HELL of a lot more than just stars.
The wording (and context) obviously meant "radio travels better in space than light does" which is true.
It's easy to deal with absorption when you've got a transmitter as powerful as the Sun.
"Light" usually implies "electromagnetic radiation". If you meant visible light, you should've said visible light. I was just saying that your terminology was a little ambiguous.
Interstellar absorption is VERY important on local star group scales! Anything smaller than about 1 micron is basically unusable on space scales: it's a power concern. For anything less than 1 micron, you have to transmit MORE power to get the same signal to the receiver. In that case, you'd be an idiot to use that frequency and not the clearer, low power one.
There's a reason we use radio to communicate with satellites and space probes - because it's intelligent, and far easier than visible light (aka lasers). While they MIGHT use lasers, they'd probably only do it for point-to-point connections (planets) and not for spacecraft.
Radio will never go away, not until we find a new medium to broadcast through. And in other civilizations, it won't go away either. It's fundamentally a good region of the frequency spectrum to use.
Personally I think the solution is tightly collimated beams and making sure that you're not leaking radiation everywhere. Broadcast spread spectrum is annoying - it does raise the noise floor. If you're talking about for IP services, and you have highly directional antennas, yah, spread spectrum should be OK. Less interference, more stable signal, should be good. But if you're talking about for cell phone use, no way, and unfortunately, I think that's in general what people want. Put a big tower broadcasting spread spectrum, and that WILL raise the noise floor. Broadcast from inside your house, and no, that won't.
:) )
You can't give wireless things priority, unfortunately. Wireless is wireless, and the only way to make certain things a priority is to use a different frequency. The problem with spread spectrum is that if you're allowed to broadcast it (the cell tower thing I was talking about) then you futz EVERYONE's frequencies, and it really sucks.
This only applies to ultra spread spectrum, in any case - the stuff that's spread over many GHz of frequency space. Ugh.
Be careful - the raising of the noise floor idea is real, and it will suck. But you are right that for personal things, it's fine - even an ad-hoc wireless over an entire city would be fine, because probably the connections between each "house" would be weak, but inside the house they'd be fine.
The worry is for commercial services, and people broadcasting (that is, intention omnidirectional broadcasting) - that will raise the noise floor.
If we set up an ad-hoc wireless network across a city, that's cool. If everyone decides to broadcast their own spread-spectrum 'SSFM' radio station, that would suck.
(do I get kudos for using 'suck' a lot?
Um. Radio is light is electromagnetic radiation. But radio has a benefit, because it is not absorbed by air. Why? Because it's long-wavelength. Basically any gaseous atmosphere is going to be transparent to radio waves, so they are probably using that - probably a liquid environment will also be transparent to it.
You're also wrong that any form of EM radiation travels well in space - dust really sucks, and it preferentially absorbs higher wavelengths (because the dust can absorb them). Radio travels well in space, light does not.
Besides, SETI is looking at an 'intelligent' portion of the spectrum (I believe... they may have switched) - the 21 cm line of hydrogen. We can't look at any large portion of the spectrum - that's really friggin' difficult. So we assume that if they're trying to contact us, they're using an intelligent wavelength.
SETI isn't looking for stray communications, in any case. They're looking for a signal intended for us to notice. If SETI fails, that doesn't mean that there isn't anything out there - it simply means they aren't trying to contact us in the way we think they are, OR they aren't trying to contact us at all.
That's the point. And you honestly think that JSRF will remain an X-Box exclusive? Sega says it will. But I'd bet large amounts of money that Sega got money to make it an X-Box exclusive, and it was probably for a set amount of time. How long will you consider waiting to buy an X-Box plus JSRF? A few months? In a few months, they'll probably announce a Jet Set Radio sequel (probably not 'Future', probably 'Cubed' or 'Extreme', for GC and PS2 respectively :) ) for another system. Then what do you buy? For the PS2, well, um, you probably already have one of those. For the GC, you KNOW games that come out on it won't come out on the X-Box, but it certainly looks like most games that are worth buying that come out on the X-box also come out on the GC - so what do you buy? Probably the GC.
It's not how you support the X-Box when its hype is high and it's the bee's knees. It's how you support the X-Box when developer support is starting to lag, you've got competing consoles out there, and the system's starting to show it's age.
Fast forward a year or so. Sony announces the successor to the PS2, the PS3, and is generating a lot of hype for it. Pressure's on Nintendo and Microsoft to announce a successor. Do you do it? Or don't you? And then, if you do, how many dev teams do you pull from your current system to work on the next system?
If you're Nintendo, you don't announce a successor. You hint at it, but keep the current system your focus. You shift the top dev teams over, but really push for second-party and third-party support, possibly with one or two last real kick-ass games for the system. With the N64, this was Zelda: Majora's Mask, the Mario Party set, Super Smash Brothers, and the Pokemon series. (Yes, SSB was produced by a different group, but they probably got the license for Nintendo characters for a song and dance). THIS is how you keep/win back developer support (the reason they lost dev support with the N64 was the cart-based format).
If you're Microsoft, who knows? Microsoft has a tendency to ditch projects when they suck too much (see also early Windows CE devices, and the whole WebTV/UltimateTV fiasco). They need to ditch this attitude, really bad.
If you're Sega, you abandon the system, say no more will be made, and the system dies a horrible screaming death as many developers line up to strangle CEOs. And then you abandon being in the hardware market altogether.
If you're Sony, you do the same thing Nintendo did (for the most part - they announce the system, but they always announce backwards compatibility. They better not ditch this for the PS3). See? This is how the market works. I worry that Microsoft won't see that, and in the next generation, we'll see them fold back into being a games developer only.
He was saying that the X-Box and PS2 compete, not the PC competes with the PS2.
My personal opinion is the X-Box and the PC are too similar for them NOT to compete - games aren't significantly different between the two (at least not yet) and thanks to Halo and a couple other of the really good games for X-Box, most people believe that games that come out for the X-Box will eventually come out for the PC. Most gamers will wait the few months if they KNOW it's coming out rather than plunk down $300.
So I personally think the X-Box has more to worry about from PCs than it does from the PS2. That's what he was arguing with.
$300 will almost buy you a PC with better specs than the X-box. I know I could probably do it with $500, $400 I'd probably have to compromise a few things. In half a year or so, I could do it with $400 definitely. Right now I could almost definitely UPGRADE a recent (2-4 years) PC to X-box status with $300.
As for online, here's the problem though: how many people are still playing Half-Life, a game that's what, 3 years old at this point? This is NOT what you want for a game system. You need to sell new games, keep games being sold, etc. to get money flowing in, because it takes a while for the system to be profitable.
That being said, MS could probably work out a strategy to make online gameplay work, so whatever. But anyway, I don't know - I personally think single-player games will garner a LOT of the market and still drive things more than multiplayer online games. The highest selling PC games are still single player games (I think).
I'd also like to point out that online games basically require a keyboard to be fun. The benefit is that you're playing against other people, rather than the computer, so communication, strategy become important. Keyboards will suck with a console (IMHO) - I'd just not bother.
See the other posts. As was pointed out (not even by me, as I'm not a financial expert) - that's not cash. That's marketable securities, so while they have 'book value' of $38B, they don't actually have $38B cash in hand.
That said, you can't buy certain companies. You can buy the company, but the talent may or may not come with it, especially if it's an unwanted buyout. With game developers, you're MUCH better off saying you'll give them lots and lots of money "if they don't develop for other systems" - but the developers can still CHOOSE, and if X-Box is dead in the water, they probably won't take any amount of money, because they want to see their game succeed, not flounder.
MS is not going to win this with money. The only way they can win this is with 1) good games, and 2) exclusive games (lots of games doesn't help: if the PS2 has the games already, the PS2's got the mindshare, and people will buy the PS2) and 3) a LOT of hype. They don't have 1, they don't have 2, and they're losing 3. Game over, Microsoft. You lost this round. Stick in the game, learn the rules, and maybe you'll play better in the second round.
Yah, I was mainly talking about graphics bandwidth, which, for the GC, is guaranteed, whereas for the XBox, you'll get stutters and hiccups when you start doing many things in the system.
Latency for 1T-SRAM is also much better than standard DDR-SDRAM. DDR-SDRAM (if memory serves - correct me here, but I think I'm within about 10%) is typically roughly 10-clock latency (maybe 7 or so - I know definitely what it is for SDRAM, but not for DDR) which is going to be about 50 ns or so for a 133MHz bus (is it 133 or 100 in the X-Box?). That's not actually what the latency is though, because it's actually a gaussian with multiple peaks when you measure it: one at "normal", one at "crossing row boundaries" (probably about +1 or 2 clock cycles), and one at "refresh cycle during read) - probably about +8 or more clock cycles. That means that your latency is only "guaranteed" probably about 100 ns: ow! 1T-SRAM, however, has a guaranteed latency of 10 ns (there's probably a tail for "crossing row boundaries" as well - not sure, MoSys is a little guarded to the general public about how things work - but not for refreshes, which don't happen). Which is, well, way cool. The benefit here is that a system with 1T-SRAM *knows* that a memory access is going to be within 10 ns, which is damn impressive. I might be off a bit here (MoSys doens't list maximum worst-case latency, but Nintendo quotes approximate sustainable latency - my guess would be that the worst-case latency is probably about 30-40 ns, but I'll assume Nintendo meant worst-case latency), but the lack of memory refresh really helps. This means that you can really push the limit of the CPU and graphics chips - much more than on the X-Box.
Personally I think a UMA was an amazingly stupid architecture for a video game system - since NUMA is already used for something else, I think "anti-UMA": that is, a system in which basically NONE of the memory is unified (that is, you have CPU RAM, sound RAM, graphis RAM, and then huge buffers on both the CD-ROM and any network adapter - large amounts of texture RAM and frame buffer RAM, etc. - you get the idea) is probably more intelligent - which is basically what the GC is. The reason is because UMA is basically a simplicity and compromise architecture. You're saying "well, in general, accesses are going to mainly be from CPU to RAM, and graphics doesn't need memory all that much, so you're okay just using all the same memory" - but you DON'T want that in a console. The reason they had to have a processor with so much higher clock speed is because they're using such an inefficient architecture that they need to give themselves a much higher headroom to get the same performance. It's really bizarre, in my opinion - UMA doesn't make any sense for a game console, which you want to be able to use all of its resources maximally, otherwise you're wasting money - which is what the X-Box is doing (wasting money).
I wasn't trying to say whether or not the X-Box is more powerful than the GC or not: what I was trying to say was whether or not developers will need time to figure out the system. In the GC's case, I think that's definitely true - developers have probably never worked with a system that's so disjoint like this - they may need time to figure out which accesses go at what time to get optimum speed. But the X-Box is a standard system: it's a P3 on a UMA, with a GeForce 3 graphics card (roughly). They're used to systems like this, especially Microsoft and all the PC developers who port games over. I don't think there will BE any 'tricks' that they can learn over time to make the system faster.
Remember, with the N64, Nintendo did bizarre stuff like allow the game to rewrite microcode in certain chips on the board: this kindof stuff is really weird, because you can say "ok, I'm NEVER going to use this function, so I'm going to use the space it was taking to streamline and speed up this other function" and things like that - that's what Factor 5 did with their MusyX engine for the N64, and it ended up producing incredible sound with little CPU overhead and really good compression (factors of 10 or better were quoted) - which is how you ended up with games like Conker's Bad Fur Day, which had literally tons of voice in the game, in a cart! I don't know if Nintendo did this with the GC, but I'd bet they did.
My point basically is that I think Microsoft made a very stupid decision creating the X-Box: using standard off-the-shelf components is how you build a general purpose PC, NOT a specific-function game console. I think that developers may figure out how to put a Web browser into their games, but I don't think they'll figure out how to make their games 10x faster and better looking.
I could be wrong. We'll see what 2nd and 3rd generation X-Box games look like compared to 2nd and 3rd generation GC games (well, see the "% difference" comparison).
I definitely agree with you on the price point and games as well, though. The GC has more chance of coming down in price lower than the X-Box, IMHO: the Pentium III that the X-Box is using is probably already as dirt cheap as it's ever going to be. Nintendo knows what it's doing. It debuted a system at its normal price point (I think? N64 debuted at $199, right? that's what I bought it at when it debuted, I'm pretty sure...) and in time, will drop the price to $175, then $150, then $99, and by then the system will be quite good with a lot of games. The X-Box is going to have a harder time coming down in price, and I think that's going to hurt Microsoft just as much as anything else. (I think I saw somewhere that the GC in a few months will be break-even on its price for Nintendo - dunno, though)
Sustained, not theoretical maximum. The main reason for this is that the GC uses MoSys 1T-SRAM, so when it wants bandwidth, it gets it, sustained, low latency. X-Box is just a normal UMA, and people've been working with that a long time. 1T-SRAM's new. I think in time people will realize that with the GC they can keep the pipe full for basically forever, and games will change to appreciate that. I'll be surprised if people "find things" in the X-Box they didn't think was there originally.
That's exactly what I said - they alienated large portions of the developers because they didn't follow the PSX's lead for the N64. Sorry if it sounded a little confusing.
:)
There's a good chance that the N64 would've taken the lead from the PSX had it been a CD based system. It also wouldn't've been the same system, so I wouldn't've gotten several kickass games, so, not sure which one is better.