Semi designers have been trying to ditch polysilicon for years. Back when it was introduced, it was an improvement over metal gates, but that's not the case anymore. Since it's still a semiconductor, you get a depletion region on the poly itself. You can ignore this in large processes, but as things shrink it really starts to have an effect.
It works even better than that: if you can reduce the operating frequency of the chip, you can typically affoed to reduce the voltage. Whereas dynamic power usage is proportional to frequency, it is also proportional to the square of the voltage.
The voltage required for a device depends on the device complexity, and the frequency - for every device, you can find a sweet-spot in terms of voltage per unit frequency, after which you tend to get decreasing returns. By selling a device clocked at its "sweet spot," you can deliver high efficiency and decent performance. The 7600 GS, at 1.15v core, is an example of a chip in its sweet spot - at less than 25w, you can have performance for very little power. The 7600 GT, by comparison, offers only %40 more performance for %60-70 more power (~36w), thanks to its much higher core voltage (1.35v).
With every process generation, you can push this sweet spot a little further - for example, the sweet spot of the GeForce 6 series was the 6600, clocked at 300 MHz core. With the move from 11nm to 90nm (6600->7600), we get a more complex chip and a boost to 400 MHz for the same power envelope. I expect to see a similar "sweet-spot" chip available from Nvidia when the 8800 series makes its way to 65nm...probably a cut-down chip with only 48 shader units (8800GTX has 128).
I believe the core pack really only made sense for people who were not interested in xbox live and had no need for any storage besides game saves. Why anybody would pick up a 360 if they weren't interested in online gaming is another question though...
The online community tends to say the same of HDTVs - why anyone owning an SDTV would pick up a 360 is beyond them, but there certainly are lots of people doing this (remember the issue with the fonts in Dead Rising on SDTVs?).
Further, there are only a bit over 4 million gold Live users, both on Xbox and the 360. If you assume half are on the new console (there's probably less than that), that means only 1 in 5 360 owners has a gold account.
So, let's look at the assumptions and how they break down:
1. People only buy 360s to play online - NO. 2. People only buy 360s to play on their HDTVs - NO. 3. People really need the hard drive - again, NO.
I have to agree, OOO re-defined Oblivion and made it fun. Without mods, random mobs were leveled based on your character level, so you're not really challenged early on in the game. The problem is, unless you level perfectly, as you gain levels the random enemies get tougher and tougher. I ended up giving up after level 20, I just couldn't keep up.
But OOO is really all you need. Here's why:
OOO brings a bit of randomness back into the game: monsters are generated on a random level scale with a maximum cap. This means you can actually meet difficult monsters you cannot beat (so run away already!), but eventually you can surpass even the toughest monsters in an area. Thus, you don't really need the AF Leveling Mod.
In addition, OOO adds improved AI and brand-new enemies. There are also roving bands of warriors, or encounters with battles already in-progress, making the game feel more lively.
There's also less NPC "equipment escalation" than there was in un-modded Oblivion, which means you have to work harder to find rare items. You still find rare items on dead NPCs, but it's much later in the game, and much less frequent. I've found in this play-through that I've spent a lot more time browsing shops, and a bit more money actually PAYING for armor, which makes the game more challenging. Among other things, it has encouraged me to complete the Thieve's Guild quests that I ignored before, both because of the rare finds, and the extra cash.
I started over about a month ago, and now I cannot get enough of the game.
I would be interested to know if there is any provision for HDTVs handling higher resolution/bandwidth signals. I.e. is it possible for the broadcasters to transmit 2160p signals and for all the current (1080 and 720) systems to be able to receive it and down-scale? I imagine this isn't possible (decoding 1080p H.264+ takes enough CPU already, I dread to think how much CPU it takes to decode 2160p), but when broadcasters eventually want to go beyond 1080p it seems like an awful waste of bandwidth to transmit both versions independently. Maybe there is some scope for transmitting 1080p signals and then just transmitting some extra data in a separate stream which can be mathematically combined to produce the extra resolution?
I imagine the supported broadcast resolution is quite open-ended, since you can pretty much define any resolution you want using MPEG2.
The problem is bandwidth. Using the ATSC standard 8VSB delivers just over 19 Megabits of bandwidth in the 6 MHz channel size currently allocated for each station. With the limits of MPEG, you don't want to transmit 1080p 60, let alone higher resolutions, because you'll start to see more obvious compression artifacts.
If you could use a more advanced encoding scheme, or change over to.h264, you could almost certainly move up in resolution - but I doubt standards will change anytime soon. I'd say expect the next big change in 30-40 years.
1. Development of efficient compilers and high-end IDEs. Without having to see the mess that is x86 machine code, you can usually ignore it. People made a clamour for clean RISC machine code in the 80s, but within a decade very few people really cared anymore.
2. Total backward-compatibility of the API for the last 20+ years. Even Windows doesn't offer such amazing compatibility modes.
3. Every fundamental architectural improvement in CPU design has been integrated into the x86 family. Academics and designers alike said it was impossible, but x86 today enjoys all the benefits of RISC, pipelining, superscalar design, branch prediction, out-of-order execution, register renaming, symmetric multi-threading and multi-processing, real-time voltage and frequency adjustment...you name it, it has been implemented on an x68 processor.
These reasons are why everyone still uses x86. These reasons are why x86-64 is the predominant 64-bit architecture, and will be for some time. The die overhead for the compatibility and translation layers on modern processes is tiny, so why the hell not keep using it?
Why would anyone over the age of 25, that doesn't have children, and already has a mac, need (or even want?) an Xbox 360?
No children, over 25, already owns a mac - that's a relatively tiny market. In fact, I know of only two (including myself). But let's assume I'm wrong, for the moment. Those types are also the kind who have already purchased Tivos, or they're the type who have already toiled away at their 1337 HTPC / MythTV boxes. Typically, they've already filled the niche that Apple has been too lazy to fill for the last 5 years.
Apple has a long uphill battle creating a system as seamless as Tivo.
The other problem is FrontRow eating into the ITV's market: why would people spend $300 on a "dumb" ITV box when they can spend $600 on a much more capable Mac Mini HTPC that also serves as a "dumb" FrontRow box at the touch of a remote? There are some technophobes in your defined market, but a lot of those "childless professionals" are technically inclined, and like flexibility. A number of peoplehave already created a comunnity around this, and there are products like DistantDVD and EyeTV that make the Mac Mini even more fully-featured.
The iPod was actually an impressive product for the time, just taking into account the specs. It was almost as small as flash players of the day, with a hundred times the capacity. It had about 10x the transfer rate of any other portable thanks to the Firewire connector and disk storage.
The iPod was great for the time, and as a hardcore PC backer it was EMBARASSING for me to see how slowly the rest of the industry responded.
Today, however, a good chunk of the industry has caught up. The iPod has serious large-market competitors from Creative and Sandisk. The iTV already has to compete with an installed base of almost 10 million Xbox 360s, not to mention the industry push from Intel (VIIV) and AMD (Live) to create "living room" PCs.
Also, now that Apple is Intel's bitch, I think it is bad that Apple is so set on the "Macworld major release schedule." Apple pulled some strings with the Core launch so it coincided with the x86 OS X launch, but I doubt Intel will ever cooridinate like that again for such a small vendor. Intel did not wait for January to release their quad-core, so the quad-core Mac Pro will end up looking more like a "me too" product instead of a market leader.
Apple can still shake things up, but in their current situation they will need to pull off some serious magic. All the "obvious" stuff just isn't that impressive.
But you CAN buy inexpensive single-core processors. AMD still sells single-core Semprons and Athlon 64s for AM2, and they're damn cheap thanks to recent battles in the high-end dual-core market.
As for 64-bit, what's wrong with "overcomercializing" it? The whole purpose of flooding the market with 64-bit processors is two-fold:
1. The processor makers are forced to keep up with each other. Buzzwords like 64-bit are powerful, and eventually you have to make all your processors match your competiitor's specs, or come out with your own version. The same thing happened with big buzzwords like MMX and the battle between SSE and 3DNOW!.
2. Microsoft and the open-source world are lobbying the processor makers heavily to go full 64-bit because it is MUCH EASIER to drop support for 32-bit modes when all the processors sold on the market have 64-bit support. You might as well buy one now so the driver makers have an excuse to pour extra cash into making better drivers...by the time the (64-bit only) successor to Vista comes out, you'll have a nice stable platform.
I'm curious myself what builds of Vista hardware manufacturers will ship. Will they bite the bullet and go 64-bit, or will they ride it out until they run into the 4GB wall? There's only one thing I know for certain: hardware manufacturers WILL have to ship Vista 64 systems - their performance systems will hit the 4GB wall within the year, and even their low-end systems will be pushing 2GB within the next 5 years.
Furthermore, dual core is being advertised as being double as fast as single core processors, which is not true. At the most, you get a 50% increase.
We didn't go multi-core because "it's a fad." We've gone dual and quad core because (relatively) we've hit the wall on processing speed. We've gone with multiple simple cores because Instruction-Level Parallelism (the governor of out-of-order-execution) shows dimishing returns as you throw more hardware at it.
Do you think Intel made the Core 2 "just" 4-decode, 4-retire on a whim? It's a balance between hardaware requirements and performance: you get decreasing returns as you add more pipes, and each additional pipe requires more hardware than the last.
Yes, the increase in performance with more cores is not that impressive for most applications. However, that will HAVE to change, because ILP processing gains won't increase much over the Core 2 design unless we see a paradigm shift in chip design techniques. We have simply moved the problem from hardware to software.
Think about this: multi-threaded software is hard, but at least it's cheaper than a super-complex chip. One other key point: even the most complex software costs virtually nothing to make copies of it. The complex chip, on the other hand, costs more to design than the complex software: each complex chip built has a higher defect rate and requres significantly more test time than a pair of simpler cores (which can be tested in-parallel).
I've had Sprint for five years now, and although the coverage was spotty when I first signed-on (like every network), they have really improved the network. I can't remember the last time I got a dropped call - even with 2 bars I can talk while moving without dropping-out, and there are very few places outdoors where I only get one bar.
Through all of the Washington / Baltimore area, the phone gets great reception. I've taken the phone out to the middle of the the Chesapeake - still got two bars. The phone has worked fine through VA/MD/PA/NJ/NY, although I got a little spotty coverage in northern NJ.
The "no roaming" long-distance works great too. I was able to make calls from southern California last time I was there wwithout issue. I even got good signals when I went to Puerto Rico (the only exceptions being the island of Culebra, and the rainforest).
Of course, I have the benefit of using an older phone (TP 1100) with good reception. I'm hesitant to upgrade because I've heard such bad things about newer phones and reception.
Connection standards will also be a bottleneck. You think they're going to throw out SATA and USB once drives like these become available? Hell no, SATA and USB are cheap to implement.
So, your 500x performance improvement will be limited to a measly 4-6x (depending on your current drive) improvement over existing hard drives, thanks to a maximum transfer rate of 300 MByte/s. It's a sizeable improvement, but is 4-6x really worth the (most likely) extra cost and severly reduced capacity?
These new technologies will also show improvement in speed over USB flash drives, but they will be limited to the same 30MBytes/s most hard drives top-out at on USB2. In my experience, mainstream flash is delivering a read rate of around 5MByte/sec, but performance flash is already delivering 20MByte/s or more. So again, you have a relatively small improvement in performance of only 2-6x. Once again, density and price will be of paramount importance because the promised improvements are severely limited by the interface.
This is the same reason why voltatile memory drives are not so popular - the massively improved latency can't make up for the massvely decreased density, higher cost, and the limited thoroughput improvement (due to reliance on PATA or SATA).
Until we have significantly faster busses for mass-storage devices, or until these new technologies can deliver competitive density at a competitive price, they're all pipe-dreams.
Good RADAR sends out a directional beam that rapidly hops frequency in unpredictable ways.
And good EW systems can easily correlate radars featuring tight beams, frequency agility, jitter and PRI. What exactly do you think those multi-GHz signal-processing systems are for?
What, you think we EW guys have been sleeping since they introduced phased arrays with all these bells and whistles? Right.
Quicktime is a trojan. It hijacks your browser's media settings WITHOUT asking, even if you tell it not to associate with any of those files on startup. It starts up it's own little preloader app which is NOT necessary, and overrides you when you try to disable it.
Finally, those fucks have the audacity to insist YOU PAY TWENTY BUCKS just to get something every other general media player offers for free: full-screen video. And even if you refuse to pay, you get a nag screen every time you load the program.
Mind you, I own a Mac, and even though I can use an applescript hack to bypass the nagware, I still avoid using Quicktime as a rule when I can. If you must have your Quicktime files, VLC plays most of them without installing the trojan.
The Itanium looks great to hardware enginners until you realize that it executes BOTH paths from a branch (until it can verify the correct branch direction), then throws out the incorrect path. When you combine this with the fact that the compiler can not always schedule each VLIW with the full 3 instructions (due to dependencies between instructions), you've got a mediocre architecture for general-purpose code. Still, at least the Itanium is a scalar processor with multiple execution pipes - this is much simpler than programming for Cell.
The SPEs in the Cell run on independent threads, so that means getting the most out of your code is mindbogglingly complex, and you can't easily make a compiler to handle it. There are only a handful of general processing tasks on this planet that can make full use of multiple independent processing threads:
Rendering (an obvious parallel task) is already handled by the GPU. Physics is another obvious parallel task, but it can only add so much to the game (and is limited by the triangle rate of the video hardware anyway). AI is a much tougher boondoggle, and making it run in-parallel AND have it take full advantage of an SPE thread or two will be a difficult challenge. And sure, it's easy to parallelize tasks like physics, but once you have many threads running, real-time synchronization becomes a serious problem.
Unfortunately, moving the complexity to the software side is the way of the future, because we're seeing less and less returns moving the complexity to the hardware side. I like to think of Cell in one good light: it will give software designers a taste of what a pain real-time hardware design is like. Having to deal with timing and synchronization of 8 or more high-speed tasks is nothing compared to doing the same with millions of high-speed transistors on a processor.
Right, that's why Microsoft made silver Live memberships free on the 360 - in anticipation of the need for patch distribution.
The hardware (even with high-level middleware like XNA and OpenGL) has gotten VERY complicated. Game datasets, now measured in gigabytes, are 100 times larger than they were 10 years ago. I would expect nothing less than an increase in bugs, and patch downloads (plus a way to store those pathches locally) is a foregone conclusion.
I have to say though: as tight as development schedules are these days, I'm surprised by the relatively low number of patches required to make games "playable." There are some exceptions, but on both consoles and the PC most games are playable out-of-the-box, or need one or two patches at the most. That's commendable, considering the circumstances.
Yes, you have to tailgate just to keep traffic flowing - if you slow down enough to allow 2 seconds of space, you only tempt the "chronic lane changer" who just can't help themselves. These people darting in-and-out of traffic only perpetuate traffic backup by re-creating the same wave effect that started it in the first place. From another perspective, it's also wasteful to leave that big a gap in-between cars when traffic is heavy, because if everyone does it, you reduce the overall carrying capacity of the lane significantly.
So, how can you drive safely without a sizeable safety buffer? Here's my take:
Even if you're driving in thick traffic, you can usually see several cars beyond the one in front of you. I think of this as a "virtual" 1-2 seconds stopping distance. Yes, keep an eye on the guy in front of you, but pay attention to all the cars beyond that as far as you can see. This is your virtual buffer zone - if one of them hits the brakes, you should do the same and take advantage of your advanced warning. This makes sense when you consider the following:
* The possibility of the car in front of you going from 65 mph to 0 mph instantly is virtually zero. It would be a once-in-a-lifetime (or less) freak event. Most people can live with those odds (I certainly can).
* The car in front of you, even if they slam on their brakes, is not going to stop *significantly* faster than your car. All that really matters is a buffer zone ~10 feet to make up the braking differential, and a decent reaction time buffer so you have time to apply your brakes (your "virtual" 1-2 seconds stopping distance).
Sure, I drive with a REAL 1-2 second buffer whenever possible, but in heavy traffic it's just not possible. Using this method makes you a safer driver overall, because even when you have a large buffer you find yourself looking at more than just the car in front of you.
In many states, the onramp is about twenty feet long after a sharp righthand turn near the end. It may be intended to be an acceleration lane, but in practice, it rarely works. (I live in Pennsylvania and this could be a symptom of poor road design)
But this is an offshoot of Pennsylvania - you have poor road design because a lot of the roads are based off of old trails, and then you have the task of fitting roads through craggy mountain passes and narrow valleys.
The good part is (at least in the Allentown - Bethlehem area where I lived), people make-way for merging cars. They see a car attempting to merge, click their turn signal and get into the left lane, giving it much more merge space. The turn signal is a clue to people behind the lead car in the right lane to get over, so the car merging ends up with a pretty clear lane.
Now, this doesn't help when the highway gets backed-up, but neither does a longer merge lane (long merge lanes only slightly raise the capacity threshold at which a highway gets backed up). Once the highway reaches overcapacity, it will eventually get backed up by merging cars no matter how long the merge lane is. The only way to fix that is to add another traffic lane, or to artifically limit inflowing traffic.
Most definately... for me that has been Fire Emblem (both of 'em, but especially the second). The graphics surely don't make this a hit; for me it is the grind gameplay. There is enough choice IMHO to justify replaying this game many times. To each his/her own.
Absolutely - to each their own. I found Fire Emblem: Path of Radiance rather boring because its nothing but tactics. There's really no roleplay - you decide what units are on the field, and then you battle. I ended up quitting after Chapter 15 or so, because I felt like I was sleepwalking.
I guess I was just jaded because I had already played an excellent tactical RPG series (Shining Force) ten years ago. In fact, I was disappointed to see that you have very little tactical decision when it comes to NPC levels and death:
1. In Fire Emblem, when your characters are dead, they're gone. In Shining Force you have the choice of bringing your characters back, for a fee. This is sometimes a better choice than replaying the battle, and infinitely less tedious. Alternately, if you don't care for the character, you have the option of doing the same thing Fire Emblem forces you to: leave them dead. The penalties go up as levels go up, so you are forced to learn good tactics, but there's lots of leeway in the system.
With Fire Emblem character deaths, I just ended up replaying the battle because I don't want to lose characters who MIGHT be important later in the game. This forces me to either cheat with an online guide to see who makes the best cannon fodder, or play a perfect game. That kind of pressure and tedium gets boring fast.
2. In Fire Emblem, you can only fight battles once, and if you retreat you lose all your progress. In Shining Force, you can retreat and maintain your progress, and fight again. In fact, this makes it easier to level-up characters you previously neglected, but suddenly find you need to win a battle. You can either win the battle without the character or put in the time to beef them up, but at least you have the choice.
On the contrary, when you combine Fire Emblem's lost progress with the permadeath, it's like you're caught between a rock and a hard place managing player skills. Pumping tons of time and experience into a character just to watch them die permanantly is - depressing.
3. There's no freedom of movement. While I won't tout Shining Force 1 as the poster child for freedom of movemenet, it does allow you to interact on your own with every town you visit (prior to the next battle). Shining Force 2 gives you full freedom of movement throughout the world, which is an impressive improvement. With Fire Emblem, I see towns, but it's always "look but don't touch." Don't these people EVER get a break from battle? Don't these people EVER leave the barracks?
I think we sometimes forget that the $600 on the RETAIL is not what SONY really get per UNIT sold (retailer margin, Handling cost, etc.). And in EU, they pay more tax than in US. So, I am pretty sure that the loss per console sold is not that much different between region.
Yup. VAT in most regions is part of the advertised price ($599 Euros = 788 USD, as of today). The EU price is high because it has to take in account the HIGHEST VAT in the whole region (%25, you can thank Denmark and Sweden for that). If they charge $788 USD for it in Denmark with VAT, the pre-VAT price is $630 USD, VERY close to the US pre-tax price. The rest is just added to take into account currency fluctuations.
That's how it is with all highly-successful consoles. By the time the next-generation comes out, you have leftover momentum. Not using that momentum to your advantage would be foolish.
I remember buying Castlevania III for NES even with the Genesis already out, and the SNES (with Castlevania IV no less) just over the horizon. I even remember buying NES games after the SNES release, although none of them were AAA releases. Nintendo released a heavily revises NES at a low price to meet this demand.
Sega did the same thing with the Genesis. They had a couple of major revisions that shrunk the system down to a square half the size of the original. Sega also had several major releaes in 1993-94 (Eternal Champions, Gunstar Heroes, Phantasy Star IV, Sonic 3, Sonic & Knuckles), the 5th and 6th year of the console's life.
On the converse, with only 20-25 million Xboxes sold, I'd expect nothing less than to see MS drop the console cold. The only time I've seen a company support a dead-end product like that was Sega porting Sonic to the Master System. But it was more of a question of convenience - if you're already going to port it to the brand-new Game Gear, why not take 5 extra minutes to make it work on the Master System?
It apparently didn't work for the season finale, which IMO sucked monkey balls. It felt unfinished, like they forgot the punchline, and was only funny in a "eh" sort of way.
What, are you so unmotivated that you can't make one intuitive leap for yourself? Are you so retarded that you cannot grasp a simple punchline without having it stuffed in your face?
-spoilers-
Stan was obviously in a movie - the twist was he was in the WRONG movie. The losing team in a sports movie is always a shallow group of assholes that get about 5 seconds of screen time off the ice. The winning team is always portrayed as a slightly less-shallow group that you feel sorry for because they're really great guys, but they all suck, and someone is dying of [insert disease here]. The roles were reversed.
You RARELY see a sports movie from the loser's perspective, because it's just too depressing.
Moreover though, saying they should prove actual value lost is a nonsense and precisely that reason why a figure was legislated in the first place. You don't know how many people downloaded the song. You don't know how many of those people passed it on to others. You do know that the person put the music up for download onto a network where it was effectively available, without control, to millions of anonymous strangers.
Ever been to a library?
They let people TAKE HOME books, CDs and DVDs!
You don't know how many people copied the book / song / movie. You don't know how many of those people passed the copy on to others. The only difference is, at least one link in the chain has to have a library card, but it's not so different: on Kazaa, at least one link in the chain has to buy the CD.
This flat $750 fine per song is bullshit. If you can be prosecuted for assumed damages, you might as well prosecute libraries and their patrons. It's about time somebody challenged this damn law. I'm all for paying for what I enjoy, and I think these people who got caught should pay for their music...but the price is due for renegotiation.
Slashdot Mirror
It's not just a "compatibility" problem.
Semi designers have been trying to ditch polysilicon for years. Back when it was introduced, it was an improvement over metal gates, but that's not the case anymore. Since it's still a semiconductor, you get a depletion region on the poly itself. You can ignore this in large processes, but as things shrink it really starts to have an effect.
It works even better than that: if you can reduce the operating frequency of the chip, you can typically affoed to reduce the voltage. Whereas dynamic power usage is proportional to frequency, it is also proportional to the square of the voltage.
The voltage required for a device depends on the device complexity, and the frequency - for every device, you can find a sweet-spot in terms of voltage per unit frequency, after which you tend to get decreasing returns. By selling a device clocked at its "sweet spot," you can deliver high efficiency and decent performance. The 7600 GS, at 1.15v core, is an example of a chip in its sweet spot - at less than 25w, you can have performance for very little power. The 7600 GT, by comparison, offers only %40 more performance for %60-70 more power (~36w), thanks to its much higher core voltage (1.35v).
With every process generation, you can push this sweet spot a little further - for example, the sweet spot of the GeForce 6 series was the 6600, clocked at 300 MHz core. With the move from 11nm to 90nm (6600->7600), we get a more complex chip and a boost to 400 MHz for the same power envelope. I expect to see a similar "sweet-spot" chip available from Nvidia when the 8800 series makes its way to 65nm...probably a cut-down chip with only 48 shader units (8800GTX has 128).
I believe the core pack really only made sense for people who were not interested in xbox live and had no need for any storage besides game saves. Why anybody would pick up a 360 if they weren't interested in online gaming is another question though...
The online community tends to say the same of HDTVs - why anyone owning an SDTV would pick up a 360 is beyond them, but there certainly are lots of people doing this (remember the issue with the fonts in Dead Rising on SDTVs?).
Further, there are only a bit over 4 million gold Live users, both on Xbox and the 360. If you assume half are on the new console (there's probably less than that), that means only 1 in 5 360 owners has a gold account.
So, let's look at the assumptions and how they break down:
1. People only buy 360s to play online - NO.
2. People only buy 360s to play on their HDTVs - NO.
3. People really need the hard drive - again, NO.
I have to agree, OOO re-defined Oblivion and made it fun. Without mods, random mobs were leveled based on your character level, so you're not really challenged early on in the game. The problem is, unless you level perfectly, as you gain levels the random enemies get tougher and tougher. I ended up giving up after level 20, I just couldn't keep up.
But OOO is really all you need. Here's why:
OOO brings a bit of randomness back into the game: monsters are generated on a random level scale with a maximum cap. This means you can actually meet difficult monsters you cannot beat (so run away already!), but eventually you can surpass even the toughest monsters in an area. Thus, you don't really need the AF Leveling Mod.
In addition, OOO adds improved AI and brand-new enemies. There are also roving bands of warriors, or encounters with battles already in-progress, making the game feel more lively.
There's also less NPC "equipment escalation" than there was in un-modded Oblivion, which means you have to work harder to find rare items. You still find rare items on dead NPCs, but it's much later in the game, and much less frequent. I've found in this play-through that I've spent a lot more time browsing shops, and a bit more money actually PAYING for armor, which makes the game more challenging. Among other things, it has encouraged me to complete the Thieve's Guild quests that I ignored before, both because of the rare finds, and the extra cash.
I started over about a month ago, and now I cannot get enough of the game.
I would be interested to know if there is any provision for HDTVs handling higher resolution/bandwidth signals. I.e. is it possible for the broadcasters to transmit 2160p signals and for all the current (1080 and 720) systems to be able to receive it and down-scale? I imagine this isn't possible (decoding 1080p H.264+ takes enough CPU already, I dread to think how much CPU it takes to decode 2160p), but when broadcasters eventually want to go beyond 1080p it seems like an awful waste of bandwidth to transmit both versions independently. Maybe there is some scope for transmitting 1080p signals and then just transmitting some extra data in a separate stream which can be mathematically combined to produce the extra resolution?
.h264, you could almost certainly move up in resolution - but I doubt standards will change anytime soon. I'd say expect the next big change in 30-40 years.
I imagine the supported broadcast resolution is quite open-ended, since you can pretty much define any resolution you want using MPEG2.
The problem is bandwidth. Using the ATSC standard 8VSB delivers just over 19 Megabits of bandwidth in the 6 MHz channel size currently allocated for each station. With the limits of MPEG, you don't want to transmit 1080p 60, let alone higher resolutions, because you'll start to see more obvious compression artifacts.
If you could use a more advanced encoding scheme, or change over to
1. Development of efficient compilers and high-end IDEs. Without having to see the mess that is x86 machine code, you can usually ignore it. People made a clamour for clean RISC machine code in the 80s, but within a decade very few people really cared anymore.
2. Total backward-compatibility of the API for the last 20+ years. Even Windows doesn't offer such amazing compatibility modes.
3. Every fundamental architectural improvement in CPU design has been integrated into the x86 family. Academics and designers alike said it was impossible, but x86 today enjoys all the benefits of RISC, pipelining, superscalar design, branch prediction, out-of-order execution, register renaming, symmetric multi-threading and multi-processing, real-time voltage and frequency adjustment...you name it, it has been implemented on an x68 processor.
These reasons are why everyone still uses x86. These reasons are why x86-64 is the predominant 64-bit architecture, and will be for some time. The die overhead for the compatibility and translation layers on modern processes is tiny, so why the hell not keep using it?
Why would anyone over the age of 25, that doesn't have children, and already has a mac, need (or even want?) an Xbox 360?
No children, over 25, already owns a mac - that's a relatively tiny market. In fact, I know of only two (including myself). But let's assume I'm wrong, for the moment. Those types are also the kind who have already purchased Tivos, or they're the type who have already toiled away at their 1337 HTPC / MythTV boxes. Typically, they've already filled the niche that Apple has been too lazy to fill for the last 5 years.
Apple has a long uphill battle creating a system as seamless as Tivo.
The other problem is FrontRow eating into the ITV's market: why would people spend $300 on a "dumb" ITV box when they can spend $600 on a much more capable Mac Mini HTPC that also serves as a "dumb" FrontRow box at the touch of a remote? There are some technophobes in your defined market, but a lot of those "childless professionals" are technically inclined, and like flexibility. A number of people have already created a comunnity around this, and there are products like DistantDVD and EyeTV that make the Mac Mini even more fully-featured.
Oh, MS was busy selling away their Basic on the 6502 for a song.
From the same page, Apple also got a similar deal on 6502 Basic for the Apple II. MS sure didn't make that mistake again with IBM.
The iPod was actually an impressive product for the time, just taking into account the specs. It was almost as small as flash players of the day, with a hundred times the capacity. It had about 10x the transfer rate of any other portable thanks to the Firewire connector and disk storage.
The iPod was great for the time, and as a hardcore PC backer it was EMBARASSING for me to see how slowly the rest of the industry responded.
Today, however, a good chunk of the industry has caught up. The iPod has serious large-market competitors from Creative and Sandisk. The iTV already has to compete with an installed base of almost 10 million Xbox 360s, not to mention the industry push from Intel (VIIV) and AMD (Live) to create "living room" PCs.
Also, now that Apple is Intel's bitch, I think it is bad that Apple is so set on the "Macworld major release schedule." Apple pulled some strings with the Core launch so it coincided with the x86 OS X launch, but I doubt Intel will ever cooridinate like that again for such a small vendor. Intel did not wait for January to release their quad-core, so the quad-core Mac Pro will end up looking more like a "me too" product instead of a market leader.
Apple can still shake things up, but in their current situation they will need to pull off some serious magic. All the "obvious" stuff just isn't that impressive.
But you CAN buy inexpensive single-core processors. AMD still sells single-core Semprons and Athlon 64s for AM2, and they're damn cheap thanks to recent battles in the high-end dual-core market.
As for 64-bit, what's wrong with "overcomercializing" it? The whole purpose of flooding the market with 64-bit processors is two-fold:
1. The processor makers are forced to keep up with each other. Buzzwords like 64-bit are powerful, and eventually you have to make all your processors match your competiitor's specs, or come out with your own version. The same thing happened with big buzzwords like MMX and the battle between SSE and 3DNOW!.
2. Microsoft and the open-source world are lobbying the processor makers heavily to go full 64-bit because it is MUCH EASIER to drop support for 32-bit modes when all the processors sold on the market have 64-bit support. You might as well buy one now so the driver makers have an excuse to pour extra cash into making better drivers...by the time the (64-bit only) successor to Vista comes out, you'll have a nice stable platform.
I'm curious myself what builds of Vista hardware manufacturers will ship. Will they bite the bullet and go 64-bit, or will they ride it out until they run into the 4GB wall? There's only one thing I know for certain: hardware manufacturers WILL have to ship Vista 64 systems - their performance systems will hit the 4GB wall within the year, and even their low-end systems will be pushing 2GB within the next 5 years.
Furthermore, dual core is being advertised as being double as fast as single core processors, which is not true. At the most, you get a 50% increase.
We didn't go multi-core because "it's a fad." We've gone dual and quad core because (relatively) we've hit the wall on processing speed. We've gone with multiple simple cores because Instruction-Level Parallelism (the governor of out-of-order-execution) shows dimishing returns as you throw more hardware at it.
Do you think Intel made the Core 2 "just" 4-decode, 4-retire on a whim? It's a balance between hardaware requirements and performance: you get decreasing returns as you add more pipes, and each additional pipe requires more hardware than the last.
Yes, the increase in performance with more cores is not that impressive for most applications. However, that will HAVE to change, because ILP processing gains won't increase much over the Core 2 design unless we see a paradigm shift in chip design techniques. We have simply moved the problem from hardware to software.
Think about this: multi-threaded software is hard, but at least it's cheaper than a super-complex chip. One other key point: even the most complex software costs virtually nothing to make copies of it. The complex chip, on the other hand, costs more to design than the complex software: each complex chip built has a higher defect rate and requres significantly more test time than a pair of simpler cores (which can be tested in-parallel).
I live in Suburban Maryland.
I've had Sprint for five years now, and although the coverage was spotty when I first signed-on (like every network), they have really improved the network. I can't remember the last time I got a dropped call - even with 2 bars I can talk while moving without dropping-out, and there are very few places outdoors where I only get one bar.
Through all of the Washington / Baltimore area, the phone gets great reception. I've taken the phone out to the middle of the the Chesapeake - still got two bars. The phone has worked fine through VA/MD/PA/NJ/NY, although I got a little spotty coverage in northern NJ.
The "no roaming" long-distance works great too. I was able to make calls from southern California last time I was there wwithout issue. I even got good signals when I went to Puerto Rico (the only exceptions being the island of Culebra, and the rainforest).
Of course, I have the benefit of using an older phone (TP 1100) with good reception. I'm hesitant to upgrade because I've heard such bad things about newer phones and reception.
Connection standards will also be a bottleneck. You think they're going to throw out SATA and USB once drives like these become available? Hell no, SATA and USB are cheap to implement.
So, your 500x performance improvement will be limited to a measly 4-6x (depending on your current drive) improvement over existing hard drives, thanks to a maximum transfer rate of 300 MByte/s. It's a sizeable improvement, but is 4-6x really worth the (most likely) extra cost and severly reduced capacity?
These new technologies will also show improvement in speed over USB flash drives, but they will be limited to the same 30MBytes/s most hard drives top-out at on USB2. In my experience, mainstream flash is delivering a read rate of around 5MByte/sec, but performance flash is already delivering 20MByte/s or more. So again, you have a relatively small improvement in performance of only 2-6x. Once again, density and price will be of paramount importance because the promised improvements are severely limited by the interface.
This is the same reason why voltatile memory drives are not so popular - the massively improved latency can't make up for the massvely decreased density, higher cost, and the limited thoroughput improvement (due to reliance on PATA or SATA).
Until we have significantly faster busses for mass-storage devices, or until these new technologies can deliver competitive density at a competitive price, they're all pipe-dreams.
Crummy RADAR is like a homing beacon.
Good RADAR sends out a directional beam that rapidly hops frequency in unpredictable ways.
And good EW systems can easily correlate radars featuring tight beams, frequency agility, jitter and PRI. What exactly do you think those multi-GHz signal-processing systems are for?
What, you think we EW guys have been sleeping since they introduced phased arrays with all these bells and whistles? Right.
Quicktime is a trojan. It hijacks your browser's media settings WITHOUT asking, even if you tell it not to associate with any of those files on startup. It starts up it's own little preloader app which is NOT necessary, and overrides you when you try to disable it.
Finally, those fucks have the audacity to insist YOU PAY TWENTY BUCKS just to get something every other general media player offers for free: full-screen video. And even if you refuse to pay, you get a nag screen every time you load the program.
Mind you, I own a Mac, and even though I can use an applescript hack to bypass the nagware, I still avoid using Quicktime as a rule when I can. If you must have your Quicktime files, VLC plays most of them without installing the trojan.
The Itanium looks great to hardware enginners until you realize that it executes BOTH paths from a branch (until it can verify the correct branch direction), then throws out the incorrect path. When you combine this with the fact that the compiler can not always schedule each VLIW with the full 3 instructions (due to dependencies between instructions), you've got a mediocre architecture for general-purpose code. Still, at least the Itanium is a scalar processor with multiple execution pipes - this is much simpler than programming for Cell.
The SPEs in the Cell run on independent threads, so that means getting the most out of your code is mindbogglingly complex, and you can't easily make a compiler to handle it. There are only a handful of general processing tasks on this planet that can make full use of multiple independent processing threads:
Rendering (an obvious parallel task) is already handled by the GPU. Physics is another obvious parallel task, but it can only add so much to the game (and is limited by the triangle rate of the video hardware anyway). AI is a much tougher boondoggle, and making it run in-parallel AND have it take full advantage of an SPE thread or two will be a difficult challenge. And sure, it's easy to parallelize tasks like physics, but once you have many threads running, real-time synchronization becomes a serious problem.
Unfortunately, moving the complexity to the software side is the way of the future, because we're seeing less and less returns moving the complexity to the hardware side. I like to think of Cell in one good light: it will give software designers a taste of what a pain real-time hardware design is like. Having to deal with timing and synchronization of 8 or more high-speed tasks is nothing compared to doing the same with millions of high-speed transistors on a processor.
Right, that's why Microsoft made silver Live memberships free on the 360 - in anticipation of the need for patch distribution.
The hardware (even with high-level middleware like XNA and OpenGL) has gotten VERY complicated. Game datasets, now measured in gigabytes, are 100 times larger than they were 10 years ago. I would expect nothing less than an increase in bugs, and patch downloads (plus a way to store those pathches locally) is a foregone conclusion.
I have to say though: as tight as development schedules are these days, I'm surprised by the relatively low number of patches required to make games "playable." There are some exceptions, but on both consoles and the PC most games are playable out-of-the-box, or need one or two patches at the most. That's commendable, considering the circumstances.
Isn't it ironic... don't you think?
It's like RAAAAAA-EEE-AAAAIIINN, on your wedding day!
couldn't resist.
Oh, and just to keep this on-topic:
2 + 2 = 5, for large values of 2.
Yes, you have to tailgate just to keep traffic flowing - if you slow down enough to allow 2 seconds of space, you only tempt the "chronic lane changer" who just can't help themselves. These people darting in-and-out of traffic only perpetuate traffic backup by re-creating the same wave effect that started it in the first place. From another perspective, it's also wasteful to leave that big a gap in-between cars when traffic is heavy, because if everyone does it, you reduce the overall carrying capacity of the lane significantly.
So, how can you drive safely without a sizeable safety buffer? Here's my take:
Even if you're driving in thick traffic, you can usually see several cars beyond the one in front of you. I think of this as a "virtual" 1-2 seconds stopping distance. Yes, keep an eye on the guy in front of you, but pay attention to all the cars beyond that as far as you can see. This is your virtual buffer zone - if one of them hits the brakes, you should do the same and take advantage of your advanced warning. This makes sense when you consider the following:
* The possibility of the car in front of you going from 65 mph to 0 mph instantly is virtually zero. It would be a once-in-a-lifetime (or less) freak event. Most people can live with those odds (I certainly can).
* The car in front of you, even if they slam on their brakes, is not going to stop *significantly* faster than your car. All that really matters is a buffer zone ~10 feet to make up the braking differential, and a decent reaction time buffer so you have time to apply your brakes (your "virtual" 1-2 seconds stopping distance).
Sure, I drive with a REAL 1-2 second buffer whenever possible, but in heavy traffic it's just not possible. Using this method makes you a safer driver overall, because even when you have a large buffer you find yourself looking at more than just the car in front of you.
In many states, the onramp is about twenty feet long after a sharp righthand turn near the end. It may be intended to be an acceleration lane, but in practice, it rarely works. (I live in Pennsylvania and this could be a symptom of poor road design)
But this is an offshoot of Pennsylvania - you have poor road design because a lot of the roads are based off of old trails, and then you have the task of fitting roads through craggy mountain passes and narrow valleys.
The good part is (at least in the Allentown - Bethlehem area where I lived), people make-way for merging cars. They see a car attempting to merge, click their turn signal and get into the left lane, giving it much more merge space. The turn signal is a clue to people behind the lead car in the right lane to get over, so the car merging ends up with a pretty clear lane.
Now, this doesn't help when the highway gets backed-up, but neither does a longer merge lane (long merge lanes only slightly raise the capacity threshold at which a highway gets backed up). Once the highway reaches overcapacity, it will eventually get backed up by merging cars no matter how long the merge lane is. The only way to fix that is to add another traffic lane, or to artifically limit inflowing traffic.
Most definately... for me that has been Fire Emblem (both of 'em, but especially the second). The graphics surely don't make this a hit; for me it is the grind gameplay. There is enough choice IMHO to justify replaying this game many times. To each his/her own.
Absolutely - to each their own. I found Fire Emblem: Path of Radiance rather boring because its nothing but tactics. There's really no roleplay - you decide what units are on the field, and then you battle. I ended up quitting after Chapter 15 or so, because I felt like I was sleepwalking.
I guess I was just jaded because I had already played an excellent tactical RPG series (Shining Force) ten years ago. In fact, I was disappointed to see that you have very little tactical decision when it comes to NPC levels and death:
1. In Fire Emblem, when your characters are dead, they're gone. In Shining Force you have the choice of bringing your characters back, for a fee. This is sometimes a better choice than replaying the battle, and infinitely less tedious. Alternately, if you don't care for the character, you have the option of doing the same thing Fire Emblem forces you to: leave them dead. The penalties go up as levels go up, so you are forced to learn good tactics, but there's lots of leeway in the system.
With Fire Emblem character deaths, I just ended up replaying the battle because I don't want to lose characters who MIGHT be important later in the game. This forces me to either cheat with an online guide to see who makes the best cannon fodder, or play a perfect game. That kind of pressure and tedium gets boring fast.
2. In Fire Emblem, you can only fight battles once, and if you retreat you lose all your progress. In Shining Force, you can retreat and maintain your progress, and fight again. In fact, this makes it easier to level-up characters you previously neglected, but suddenly find you need to win a battle. You can either win the battle without the character or put in the time to beef them up, but at least you have the choice.
On the contrary, when you combine Fire Emblem's lost progress with the permadeath, it's like you're caught between a rock and a hard place managing player skills. Pumping tons of time and experience into a character just to watch them die permanantly is - depressing.
3. There's no freedom of movement. While I won't tout Shining Force 1 as the poster child for freedom of movemenet, it does allow you to interact on your own with every town you visit (prior to the next battle). Shining Force 2 gives you full freedom of movement throughout the world, which is an impressive improvement. With Fire Emblem, I see towns, but it's always "look but don't touch." Don't these people EVER get a break from battle? Don't these people EVER leave the barracks?
I think we sometimes forget that the $600 on the RETAIL is not what SONY really get per UNIT sold (retailer margin, Handling cost, etc.). And in EU, they pay more tax than in US. So, I am pretty sure that the loss per console sold is not that much different between region.
Yup. VAT in most regions is part of the advertised price ($599 Euros = 788 USD, as of today). The EU price is high because it has to take in account the HIGHEST VAT in the whole region (%25, you can thank Denmark and Sweden for that). If they charge $788 USD for it in Denmark with VAT, the pre-VAT price is $630 USD, VERY close to the US pre-tax price. The rest is just added to take into account currency fluctuations.
That's how it is with all highly-successful consoles. By the time the next-generation comes out, you have leftover momentum. Not using that momentum to your advantage would be foolish.
I remember buying Castlevania III for NES even with the Genesis already out, and the SNES (with Castlevania IV no less) just over the horizon. I even remember buying NES games after the SNES release, although none of them were AAA releases. Nintendo released a heavily revises NES at a low price to meet this demand.
Sega did the same thing with the Genesis. They had a couple of major revisions that shrunk the system down to a square half the size of the original. Sega also had several major releaes in 1993-94 (Eternal Champions, Gunstar Heroes, Phantasy Star IV, Sonic 3, Sonic & Knuckles), the 5th and 6th year of the console's life.
On the converse, with only 20-25 million Xboxes sold, I'd expect nothing less than to see MS drop the console cold. The only time I've seen a company support a dead-end product like that was Sega porting Sonic to the Master System. But it was more of a question of convenience - if you're already going to port it to the brand-new Game Gear, why not take 5 extra minutes to make it work on the Master System?
It apparently didn't work for the season finale, which IMO sucked monkey balls. It felt unfinished, like they forgot the punchline, and was only funny in a "eh" sort of way.
What, are you so unmotivated that you can't make one intuitive leap for yourself? Are you so retarded that you cannot grasp a simple punchline without having it stuffed in your face?
-spoilers-
Stan was obviously in a movie - the twist was he was in the WRONG movie. The losing team in a sports movie is always a shallow group of assholes that get about 5 seconds of screen time off the ice. The winning team is always portrayed as a slightly less-shallow group that you feel sorry for because they're really great guys, but they all suck, and someone is dying of [insert disease here]. The roles were reversed.
You RARELY see a sports movie from the loser's perspective, because it's just too depressing.
Moreover though, saying they should prove actual value lost is a nonsense and precisely that reason why a figure was legislated in the first place. You don't know how many people downloaded the song. You don't know how many of those people passed it on to others. You do know that the person put the music up for download onto a network where it was effectively available, without control, to millions of anonymous strangers.
Ever been to a library?
They let people TAKE HOME books, CDs and DVDs!
You don't know how many people copied the book / song / movie. You don't know how many of those people passed the copy on to others. The only difference is, at least one link in the chain has to have a library card, but it's not so different: on Kazaa, at least one link in the chain has to buy the CD.
This flat $750 fine per song is bullshit. If you can be prosecuted for assumed damages, you might as well prosecute libraries and their patrons. It's about time somebody challenged this damn law. I'm all for paying for what I enjoy, and I think these people who got caught should pay for their music...but the price is due for renegotiation.