The portable video device market isn't a "Well sometime in the future" deal. It's here NOW. Don't believe me? Sony was awfully surprised at how popular UMD movies suddenly became. Why? Let's look at what the average Joe has for portable video options:
1. Portable DVD player. It's bulky, it chews power, it's expensive to get one with a screen, the media is widely available but also bulky, if you travel internationally you have to worry about region headaches, etc.
2. Hard drive based video player like Archos etc. These are nice, often run a long time, store a ton of video. But getting video onto them is a chore. Disregarding DMCA issues, transcoding video is a pain in the ass. It takes forever, there's a million different options, and it's very user unfriendly in most cases. I've been encoding mpeg4s since before The Matrix hit theaters, and believe me the situation has gotten worse in terms of complexity, not better.
3. PSP. This is where the money is going. Why? It's got a bright high-res screen. It's not too bulky to fit in a carryon bag, and a whole flight's worth of video fits in a pocket (the batteries to run it are another matter). The UMD is plug and play average person useable type stuff. You buy one, you put it in, the movie plays. Sure it's expensive and lacks alot of DVD features, but IT WORKS! That's what people want. Developers are jumping ship on the PSP due to its flaws, but Sony doesn't care because suddenly UMDs are turning into a cash cow. They've already got two titles with over 100,000 units sold in 2 months. It took their first DVDs 9 months to get there. They're now estimating movies will make up the majority of UMD sales pretty soon.
So, where does this leave Apple? Pretty screwed unless they get something out in a 6-12 month window. Sony is eating up the market and rapidly becoming the standard with a proprietary physical and logical format. That's BAD! If Apple wants to have a chance they need a video iPod now. Even if it's just to make the market pause and look at them. Ideally they need several key features:
1. Content content content. They need to get video over itunes, or strike a deal for release on SD card or similar. They can't afford to rely on ripping like they do with CDs, it's a headache of epic proportions, and it's not happening.
2. Some form of portable storage to get new movies on, as well as a HD to store some on. This will need some DRM finesse, but if anyone can manage it, it's Apple. The aforementioned SD card would be a good start. There's few movies you can't get at least TV quality on in a 256 or 512MB space with modern mpeg4 and h.264 codecs. 256MB SD cards are cheap as dirt these days, distributing movies on them is a no brainer and will probably have comparable media cost to the UMD, which is a caddy-bound propritary disk. SD will probably drop in cost quicker due to volume and density increase.
3. TV-friendliness. At least TV output. Preferably also TV-input. To solve the problem of content, you need a way for the average guy to get content on there. TiVo is immensely popular so this should be a no brainer. Make your HD based iPod Video a PVR. You can get away without the fancy scheduling software (maybe have it as an option when hooked to a PC). If you just have a VCR-style "Push button to reccord" interface, it will work for alot of people. Sure reccording video at 1x speed is slow, but it's a world easier than transcoding and you can do it from virtually any source. There might be issues once everything goes digital but that's to worry about later. Apple can even bow down to Macrovision, enough DVD and VCRs have workarounds for it anyway, like regions.
I'm not sure who's asleep at the wheel over there, maybe it's Jobs, but they better wake up or Sony's going to give them Walkman Revenge up the ass for catching them asleep in the digital music player market a few years ago. Once we're locked into something like UMD, you can kiss fair use goodbye. Sony's not going to stand up to MPAA pricing like Apple does to the RIAA, they're on the other side of the fence!
I did preview it, accidentally leaving it as "HTML Formatted", one of the most retarded default settings in the world. Who the hell writes
tags by default instead of just skipping a line.
I was going to reset it to plain text to get the spacing back, but on proofreading I found it worked better this way, more of a stream of consciousness thing.
Don't assume just because you make stupid mistakes others make the same ones. Oh and you forgot to log in.
I think we've lost the plot somewhere along the way. People have no clue what a government is supposed to do and not do.
The government is not here to babysit your children and make sure they don't hear naughty words or see a boob.
The government is not here to enforce your religious views on everyone who doesn't subscribe to your religion, whether you're a majority or not.
The government is not here to guarantee a right to profit for corporations.
The government is not here to keep track of what everyone does day and night in order to prevent any possible crime or terrorism.
In general, IT IS NOT THE GOVERNMENT'S JOB TO CATCH CRIMINALS! That is a SECONDARY function to what governments should be doing.
A government is supposed to protect the freedom of the people. i.e. life, liberty, pursuit of happiness.
A government is supposed to protect the people from external threats, i.e. terorrism, war, but not at the expense of the freedom of its people.
A government is supposed to maintain order, and prevent chaos from threatening people, i.e. catch criminals, prevent theft, prevent murder, etc.
A government is supposed to ensure basic quality of living and services, i.e. infrastructure: roads, electricity, water, sewage.
All this nonsense about giving up civil liberties to "prevent terrorism" is counterintuitive. You are there to ensure liberty, not remove it. Removing liberty all the time to prevent terrorism some of the time is not a positive net change for the people. Ensuring liberty is your primary function, if you are working against that, there is a problem with the plan or the execution. There needs to be another solution. Restricting freedoms of people who are not themselves the threat IS NOT AN OPTION.
Restricting freedom of people to live, marry, immigrate, visit in hospitals, raise families, and be happy because your religious beliefs do not agree with it is not a valid action. There needs to be a threat to others in order for freedoms to be taken away. Claiming it is a threat because "think of the children" is a fallacy. You are responsible for your children. The government is not here to impose your moral values on others. It is not the government's job to instill morality in your children, that is your job as a parent. Try living up to your end of the bargain.
If I had one wish in all the world, it would be for an empty habitable space to found my own society, based on reason instead of stupidity, with a design towards reducing corruption. The mult-branch thing was a good idea but didn't cut it. Plurality vote gravitates towards fewer parties. Lifetime politicians who have more interest in their private finances and companies than their jobs is a problem.
We need to fix this but the systems around have sunk in and fossilized. There's no way to actively remove them, not even by force anymore. Revolution is virtually impossible in the age of modern weaponry. There is nowehre else to go. Any initiative that makes serious progress will be sabotaged by some existing faction in power, either political, religious, or corporate.
I've really lost hope, so I guess do whatever you want.
I'm having a hard time believing the following can be true:
1. This doesn't get you high, even if taken at higher doses, like cough medicine.
2. It does't get you high if you combine it with other legal or prescription substances.
3. It's not addictive.
One of the above is probably false. And that's bad. I give it two weeks before the first college kid goes on a 3 day binge the weekend before midterms, and pops 5x the reccomended dosage at 6am Monday morning, with a BAC still over the legal limit where it's been since Thursday.
Granted these could be very useful and I would probably want to use them myself, but people are idiots, and this is going to harm or kill them, I guarantee it. I'm not anti-drug, I believe what you do with your own body is your own business and what I do with mine is mine (if only a single government on the planet agreed). But in the world we live in, this isn't going to fly. There'll be lawsuits all over the place.
It's just a matter of the right design. For example:
1. A dynamic array of all active objects in the game.
2. An object processor class that gets object pointers from the list in a critical section (i.e. locked) state, then unlocks and processes.
Using this system, what happens is you spawn multiple threads that run object processors. Each processor locks access to the queue when it goes to get a pointer, grabs the next object on the list, and updates the index to the position in queue, then releases the lock and goes and does its thing. This way, no two threads grab the same object at the same time for processing.
Now, how about multiple kinds of processes i.e. Input, AI, Physics, Rendering?
Not a problem as long as you maintain the objects in an "always valid" state. I.e. you do collision detection after move but before returning the object. They don't even have to run at the same frequency, your AI can run at 10hz while your renderer runs at 60hz.
So, you spawn an input processor that delivers inputs to the objects. Your player and master AI both feed this process with targeted commands that are delivered to the objects in the same way. Your move processor handles movement and collision detection or flags objects for another processor. Your physics does the same, your renderer also does the same, etc. As long as you're doing a full traverse of the game world, and nothing is left in an invalid state, everything should be fine. You only need to perform locking for processors of the same type, ex: if a move processor has a locked object, the renderer doesn't care, it can read the current state and draw it before the copy is modified on return. As long as the previous state when the move thread took the pointer was valid, it should be fine. A bit complicated, but doable.
Here's the fun part: It's asynchronous. Spawn 2 threads on the CPU and 1 on the GPU. Asymmetric multiprocessing. Think it'll screw things up? Order of return doesn't matter here, it's first come first serve, so if the GPU thread is 4x faster, it can just grab objects 4x as often. As long as your algorithms return identical results it doesn't matter who does the job.
Of course in a game, not all objects update all the time. If you were using a physics driven system, what you'd want to do is have a vector queue instead. Then grab the kinetic vectors and apply them to the respective objects, check for collision and if so update the existing vectors and then spawn new vector objects into the queue as motion transfers. Cull out vectors below a certain magnitude as "residual noise". Put in friction and the system will keep itself manageable by dropping vectors as they become less significant.
I can't wait to build a physics engine running on this design, but I'm still developing my skills for a project of that magnitude.
I didn't mean the entire ocean, just a little bit. As for the cancer thing, there was a study done where they used very small amounts of antimatter to cause very localized irradiation and kill tumors without causing massive system damage like traditional chemo or radiation therapy. Problem is, it costs a fortune to make the stuff, so nobody can really afford the treatement. That problem goes away when you have say, a billion times the electricity available.
The orbit to ground thing is a problem, yeah. It'd probably have to be reflected or converted to microwave. Fortunately if your satellites are in geosync orbit they're over the same spot and less likely to have microwave beams wandering around cooking people, but accidents happen (see Sim City 2000).
Yes there's meteor strikes, so there'll be repairs. If you can build it, you can fix it. As for hitting the laser, that'd probably need to be rebuilt too (and it would also probably have been locally built) and I was thinking of quite a few along the meridian of the moon that's locked parallel to the earth's surface (i.e. the earth's always on the horizon), so single point failure wouldn't be a problem.
As for what to do with self replicating solar and laser and robot building robots during downtime... hey, free battlebots online via remote control anyone?
We're getting bogged down by energy requirements. Oil is going through the roof, batteries are barely crawling along in improvements, fuel cells still cost a fortune, and everyone is still afraid of the nuclear boogeyman.
What does that leave? Geothermal? Fat chance of seeing that go wide spread. So that leaves solar.
Why the hell is the moon not coated with solar cells? I mean, seriously. Ok let's say we don't want to change how it looks. The bitch is tidelocked! Just put them on the back! Oh but we'll have to go up there and it'll take forever to build! No it won't. Robots, people! I remember reading in Discover around 1992-93 or so about a new all-electrical process someone had developed for extracting materials from sand. He had a bunch of little robots running around the desert building solar cells out of the raw silicon. The moon's got that in spades, and aluminum for the connections. Yeah the efficiency won't be great but who cares when you have an entire MOON (or even half a moon) of them?
How do we get the robots there? Send some. But it won't be enough! Self-replicating. Is this really such a hard challenge? We're seeing basic steps towards it today. Tell me it would cost more than a major space program like a Mars trip to get it working and on the backside of the moon.
How does the power get back? "Laser". But won't it cook the earth? Not if you lock the depression angle so it can only hit geosynchronous orbit and not cross the earth.
But won't people abuse it and fight over it? Declare the moon array itself public domain. Make all the receiving sattelites privatized to create competition and prevent government death rays. Make all the ground stations government owned to prevent slum-shopping for placements by over-greedy immoral corporations. There, you have a case for competition and a nice construction project for all those 3rd world equator countries with the best views of orbit.
What would that get you for your hundred billion or so invesment? UNLIMITED POWER! We wouldn't NEED oil, or fusion, or anything else with that running. Want to use it to go into space? Point the lasers the other way and use them with sails or to power ion drive systems. We'd be mining the asteroid belt with Mark 2 replicating robots in no time. Then we have unlimited energy AND unlimited resources.
Then the real fun starts. Want to end world hunger? Desalinate the ocean and irrigate the entire sahara desert. It'd be cheap. Want to end pollution? Electrochemical reclamation. With virtually free power, post-problem pollution fixes are cheap enough to work. Want to educate everyone? What kind of network can you run when you don't need to worry about electrical losses? Want to cure cancer? There's some promising work with antimatter. Build accelerators to produce it, more efficient ones than the general-purpose kind we have now. Don't want them on earth? Put them on the moon too, make a bigass one around the equator, ship the people there on vacation. Want to get rid of that threatening asteroid headed for earth? Zap it with a petawatt or two before it passes Mars and watch the vapor pressure push it away. Maybe into a nice orbit where we can strip mine it.
All that aside, biotech is going to be the next kick ass field. Read Wired in the last couple years? We can just about cure f'ing BLINDNESS! Eat that you boomer fossils! We're going to see fixes for spinal injuries, better transplants, a doubling of life span, improved prosthetics or maybe even regrown parts. Think some religious-based policies will stop that? Maybe in the US, that's just going to open the door for someone else to take the lead. We're going to be 130 and bitching our great grand kids want tails and wings for xmas and how immoral it is and back in our days we just hijacked cars on playstation and hacked virtual sex in, and that was fine for us!
Well it would probably be best done as a set CPU mode that doesn't expose the cores to the system individually. Kind of a reverse of how hyperthreading makes one core look like two. I'm sure it wouldn't be nearly as good for multithreaded apps, but for those few hard unparallelizable problems, it might be a way to continue performance improvements even though clockspeed seems to be hitting the wall.
But is it feasible? I don't have the tech background to really know. There would likely have to be some kind of internal bus to allow each core to copy the internal state of each pipeline stage of another core so they could resync.
But I think there are several ways to take advantage of multiple cores that are being overlooked.
1. Modern OSes have so many threads running in the background that not having to constantly context switch would probably be a big speedup for single threaded apps like games.
2. The big thing in games is physics. It can be parallelized quite nicely. Dedicated accelerators are still in the vaporware "It's coming any time now!" stage.
3. I'm no CPU designer, but I was thinking. If you had a way to sync state between cores really rapidly, like say in less clocks than a pipeline flush, couldn't you use multiple cores for speculative execution? Let's think about this for a second. The big killer in single thread speed is branching. We spend alot of transistor budget on things like cache and branch prediction to try and reduce it. But what if instead when we hit a branch we forked the execution, so now one core is computing one result, and another the other case. With four cores you could do this two branches deep, with 8 three branches etc. What happens is when the branch actually resolves, the CPU could select the core with the correct prediction, resync the other cores from that CPU's values, and continue execution, without ever having stalled for a branch or flushed the pipeline for a mispredict. As long as your core state resync is less clocks than a flush, it would make sense to do this. Obviously not viable on the dual core P4s now, but with a new design coming, why not? This would let single thread performance scale with core count.
We've got a nice IRC channel for people who don't believe SA has internet. Also our own server, irc.ac.za (though your ping to it will probably suck), and for the clientless, http://www.ircd.co.za/ for a java client to it. Or if you just like forums, http://mybroadband.co.za/vb/
I would have submitted this sooner, but it seems that any time I'm at work, if I try to post a comment on slashdot, it gives me "You can't post to this page." Yet it works fine at home on another ISP. I've tried asking what's going on, if it's some kind of domain ban, but nobody ever replied.
Actually we saw them at E3, filmed some of them to take back and produce our E3 video on the magazine's DVD for July.
Killzone? Yawn. Have you ever actually played the first one? It's not exactly a showstopper. Yay it looks pretty. So what? The 360 games all look on par so far. Where's Devil May Cry 4? Where's a jaw-dropping platformer like Ratchet and Clank or Jak 3? Where's an even better and more absurd Burnout? Why is Soul Calibur 3 a crappy PS2 title that looks if anything worse than the previous one?
Where's ANY ACTUAL RUNNING GAMEPLAY FOOTAGE? All we've seen are "realtime" cinematics. Whoopee. I'd like to see some actual games please.
DSL is commonly oversubscribed at 20:1 and 50:1 ratios to the backbone bandwidth. It's true, look it up. When my cable ISP says "You have 10mbps download." I pull 1200KB/sec off an FTP. The line drops me on an average of less than once a YEAR. Beat that!
Cable bandwidth delivers, it's just the upstream that sucks.
I had no problems with my cablemodem during a power failure.
You remember that big east coast blackout? I live on Long Island. I got home about 7pm, after 2-3 hours crawling through insane traffic because all the lights were down. I unplugged my auto-shutdown desktop, and plugged my laptop into the 1500VA CyberPower UPS that the cablemodem and router share.
I was still on IRC most of the night. People kept asking me "Wait, aren't you in New York?" To which I replied "Yes, and my ISP obviously has a backup generator." My service didn't cut out until around 1am, when the governor got on the radio and told everyone in the state to turn off everything they could so the power grid could be rebalanced and restarted. Then my cable went out and I went to bed.
If your VOIP and cable are dying in a power failure, either you need a UPS, or your ISP does. If you ISP doesn't have one, find one that does, I'm sure someone out there has a reliable business class VOIP and internet solution that won't go down when the lights flicker.
This is exactly what killed it. Sure there were no major gaming studios looking to port the latest PS2 titles to it, but there were tons of indie developers interested in putting popcap-style games and last-generation style 3D games on this thing. But you couldn't thanks to the wonderful PalmOS and the signed code system.
If this had been an open platform based on PocketPC, there would have been alot more games for it. Sad to see it go, my husband and his best friend both own one and got alot of fun out of them, but the promise of this machine never fully materialized.
The hardware's superior, where's the software? Looks like Nintendo scared too many developers away with their previous N64 policies, and now they don't have the install base to attract enough anymore.
* Nintendo DS too bulky compared to PSP
This is a load of crap. My husband owns both, he's a professional game reviewer for a major magazine (print, not web). Guess which one fits in his jeans pocket? I'll give you a hint, it's not the one that attracts dust from 20 meters, is so expensive you're constantly nervous about scratching or dropping it, has constant hardware problems (his has a broken UMD latch but no dead pixels fortunately), has almost no titles, has no easy way to find other people to play wireless games with, has been delayed for made up reasons in most of the world, and has "support" in the form of patches that break anything interesting you can do with it.
DSlinux.org and Gamemaker port for the win!
* People are waiting for PS3 (Cell processor + Linux!), not GC3
I think people are waiting for a console they can actually afford that has some games they want. I've seen a good dozen HD trailers of complete or nearly so Xbox 360 games, where's the PS2 titles? Any killer apps yet? I haven't seen any. Maybe developers are having problems dealing with a CPU that has terrible integer performance and a wacky memory system with too little local memory to do a few algorithms like say... collision detection on the SPEs?
Unless the drives support an asynchronous write system, which they do. NCQ will reorder the writes anyway. Latency is primarly a read-side issue, random writes are not as common as random reads.
You can always improve your seek time by adding more redundant mirrors. If we apply the formula the formula seen here where x is the number of redundant mirrors, we can calculate the value of p which will give us our rotational latency for the mean seek time (hence the 0.5 because we want the 50% point for seek times).
Using this you can get 7200rpm drives to easily outseek a 15000rpm drive by using 4 or more redundant sources, and it's still cheaper for the same capacity, AND more failure tolerant.
This is why RAID always wins. Quantity has a quality all its own. SCSI used RAID to defeat the SLED concept in mainfraimes, commodity drives are doing the same to SCSI, by playing with the same rules.
I got out of a major ticket this way when I was 18. The officer had shoddy handwriting so it looked like his odometer in his log contradicted his car's test reccords (which was important because he said I was caught speeding by pacing, not radar). The judge, knowing odometers don't normally go backwards (and also thinking this guy's 4s were 9s), let me off. Which is good because in many states, one moving violation under the age of 21 and you have no license until you're 21, which basically means you're unemployable.
Optimum Online is already testing a 100mbps system to see if they're going to roll it out. Right now they're only running it around 50mbps. I believe the pilot project is in Smithtown, out on Long Island.
Try South Africa's Telkom. Here you can't get DSL without a 3GB per month cap. Of course local traffic and your uploads count on your cap. No you can't get another ISP because all ISPs just resell their lines. You can get DSL that cuts off 3 days into the month if you play onlne games, or you can get 56K and have 1800ms pings to everywhere. Yay.
It depends. If you're cute I wouldn't really mind at all, you're welcome to come over and share him. It wouldn't be the first time.
The funny thing about the internet is you can never assume what kind of people you're talking to.
Wow, you mean they want to own what they pay for?
on
Britons Frustrated by DRM
·
· Score: 4, Insightful
What a novel concept. Those out there saying "well they should have read the fine print" don't seem to get it. It's not that they expected one thing and got another, it's that even people who know what the deal is don't have a legal option to OWN unrestricted files. It's not presented, at any price. That's where the real problem is.
I figured once DRM got widespread enough to start causing problems with mainstream devices the average Joe (or whatever the name in the UK is) would start taking notice. I've been hearing "But WHY can't I tape my DVD like I do my other tapes?" for awhile now, so I figured it was only a matter of time. The broadcast flag will likely have the same effect. A couple months of nothing major and then suddenly rising complaints of not being able to do the things that were always just fine.
You could be here in South Africa, where people are forced to pay TV licenses (even though the retailers claim they suffer no penalties if they don't comply) to fund complete crap that you can barely recieve anyway. Oh and if you don't like the one station the money goes to? Tough. If you just watch DVDs and play video games and don't have an antenna? Tough. What a crock.
Slashdot Mirror
The portable video device market isn't a "Well sometime in the future" deal. It's here NOW. Don't believe me? Sony was awfully surprised at how popular UMD movies suddenly became. Why? Let's look at what the average Joe has for portable video options:
1. Portable DVD player. It's bulky, it chews power, it's expensive to get one with a screen, the media is widely available but also bulky, if you travel internationally you have to worry about region headaches, etc.
2. Hard drive based video player like Archos etc. These are nice, often run a long time, store a ton of video. But getting video onto them is a chore. Disregarding DMCA issues, transcoding video is a pain in the ass. It takes forever, there's a million different options, and it's very user unfriendly in most cases. I've been encoding mpeg4s since before The Matrix hit theaters, and believe me the situation has gotten worse in terms of complexity, not better.
3. PSP. This is where the money is going. Why? It's got a bright high-res screen. It's not too bulky to fit in a carryon bag, and a whole flight's worth of video fits in a pocket (the batteries to run it are another matter). The UMD is plug and play average person useable type stuff. You buy one, you put it in, the movie plays. Sure it's expensive and lacks alot of DVD features, but IT WORKS! That's what people want. Developers are jumping ship on the PSP due to its flaws, but Sony doesn't care because suddenly UMDs are turning into a cash cow. They've already got two titles with over 100,000 units sold in 2 months. It took their first DVDs 9 months to get there. They're now estimating movies will make up the majority of UMD sales pretty soon.
So, where does this leave Apple? Pretty screwed unless they get something out in a 6-12 month window. Sony is eating up the market and rapidly becoming the standard with a proprietary physical and logical format. That's BAD! If Apple wants to have a chance they need a video iPod now. Even if it's just to make the market pause and look at them. Ideally they need several key features:
1. Content content content. They need to get video over itunes, or strike a deal for release on SD card or similar. They can't afford to rely on ripping like they do with CDs, it's a headache of epic proportions, and it's not happening.
2. Some form of portable storage to get new movies on, as well as a HD to store some on. This will need some DRM finesse, but if anyone can manage it, it's Apple. The aforementioned SD card would be a good start. There's few movies you can't get at least TV quality on in a 256 or 512MB space with modern mpeg4 and h.264 codecs. 256MB SD cards are cheap as dirt these days, distributing movies on them is a no brainer and will probably have comparable media cost to the UMD, which is a caddy-bound propritary disk. SD will probably drop in cost quicker due to volume and density increase.
3. TV-friendliness. At least TV output. Preferably also TV-input. To solve the problem of content, you need a way for the average guy to get content on there. TiVo is immensely popular so this should be a no brainer. Make your HD based iPod Video a PVR. You can get away without the fancy scheduling software (maybe have it as an option when hooked to a PC). If you just have a VCR-style "Push button to reccord" interface, it will work for alot of people. Sure reccording video at 1x speed is slow, but it's a world easier than transcoding and you can do it from virtually any source. There might be issues once everything goes digital but that's to worry about later. Apple can even bow down to Macrovision, enough DVD and VCRs have workarounds for it anyway, like regions.
I'm not sure who's asleep at the wheel over there, maybe it's Jobs, but they better wake up or Sony's going to give them Walkman Revenge up the ass for catching them asleep in the digital music player market a few years ago. Once we're locked into something like UMD, you can kiss fair use goodbye. Sony's not going to stand up to MPAA pricing like Apple does to the RIAA, they're on the other side of the fence!
Check the sig I've been using the last few years for more info.
I did preview it, accidentally leaving it as "HTML Formatted", one of the most retarded default settings in the world. Who the hell writes
tags by default instead of just skipping a line.
I was going to reset it to plain text to get the spacing back, but on proofreading I found it worked better this way, more of a stream of consciousness thing.
Don't assume just because you make stupid mistakes others make the same ones. Oh and you forgot to log in.
I think we've lost the plot somewhere along the way. People have no clue what a government is supposed to do and not do. The government is not here to babysit your children and make sure they don't hear naughty words or see a boob. The government is not here to enforce your religious views on everyone who doesn't subscribe to your religion, whether you're a majority or not. The government is not here to guarantee a right to profit for corporations. The government is not here to keep track of what everyone does day and night in order to prevent any possible crime or terrorism. In general, IT IS NOT THE GOVERNMENT'S JOB TO CATCH CRIMINALS! That is a SECONDARY function to what governments should be doing. A government is supposed to protect the freedom of the people. i.e. life, liberty, pursuit of happiness. A government is supposed to protect the people from external threats, i.e. terorrism, war, but not at the expense of the freedom of its people. A government is supposed to maintain order, and prevent chaos from threatening people, i.e. catch criminals, prevent theft, prevent murder, etc. A government is supposed to ensure basic quality of living and services, i.e. infrastructure: roads, electricity, water, sewage. All this nonsense about giving up civil liberties to "prevent terrorism" is counterintuitive. You are there to ensure liberty, not remove it. Removing liberty all the time to prevent terrorism some of the time is not a positive net change for the people. Ensuring liberty is your primary function, if you are working against that, there is a problem with the plan or the execution. There needs to be another solution. Restricting freedoms of people who are not themselves the threat IS NOT AN OPTION. Restricting freedom of people to live, marry, immigrate, visit in hospitals, raise families, and be happy because your religious beliefs do not agree with it is not a valid action. There needs to be a threat to others in order for freedoms to be taken away. Claiming it is a threat because "think of the children" is a fallacy. You are responsible for your children. The government is not here to impose your moral values on others. It is not the government's job to instill morality in your children, that is your job as a parent. Try living up to your end of the bargain. If I had one wish in all the world, it would be for an empty habitable space to found my own society, based on reason instead of stupidity, with a design towards reducing corruption. The mult-branch thing was a good idea but didn't cut it. Plurality vote gravitates towards fewer parties. Lifetime politicians who have more interest in their private finances and companies than their jobs is a problem. We need to fix this but the systems around have sunk in and fossilized. There's no way to actively remove them, not even by force anymore. Revolution is virtually impossible in the age of modern weaponry. There is nowehre else to go. Any initiative that makes serious progress will be sabotaged by some existing faction in power, either political, religious, or corporate. I've really lost hope, so I guess do whatever you want.
I'm having a hard time believing the following can be true:
1. This doesn't get you high, even if taken at higher doses, like cough medicine.
2. It does't get you high if you combine it with other legal or prescription substances.
3. It's not addictive.
One of the above is probably false. And that's bad. I give it two weeks before the first college kid goes on a 3 day binge the weekend before midterms, and pops 5x the reccomended dosage at 6am Monday morning, with a BAC still over the legal limit where it's been since Thursday.
Granted these could be very useful and I would probably want to use them myself, but people are idiots, and this is going to harm or kill them, I guarantee it. I'm not anti-drug, I believe what you do with your own body is your own business and what I do with mine is mine (if only a single government on the planet agreed). But in the world we live in, this isn't going to fly. There'll be lawsuits all over the place.
Given that there's no such thing as static IP in this country, I wouldn't doubt it, but I have no idea what the exact issue is or how to adress it.
http://games.slashdot.org/article.pl?sid=05/08/11/ 2019242&from=rss
Have a look there, lots of nice motion blur.
It's just a matter of the right design. For example:
1. A dynamic array of all active objects in the game.
2. An object processor class that gets object pointers from the list in a critical section (i.e. locked) state, then unlocks and processes.
Using this system, what happens is you spawn multiple threads that run object processors. Each processor locks access to the queue when it goes to get a pointer, grabs the next object on the list, and updates the index to the position in queue, then releases the lock and goes and does its thing. This way, no two threads grab the same object at the same time for processing.
Now, how about multiple kinds of processes i.e. Input, AI, Physics, Rendering?
Not a problem as long as you maintain the objects in an "always valid" state. I.e. you do collision detection after move but before returning the object. They don't even have to run at the same frequency, your AI can run at 10hz while your renderer runs at 60hz.
So, you spawn an input processor that delivers inputs to the objects. Your player and master AI both feed this process with targeted commands that are delivered to the objects in the same way. Your move processor handles movement and collision detection or flags objects for another processor. Your physics does the same, your renderer also does the same, etc. As long as you're doing a full traverse of the game world, and nothing is left in an invalid state, everything should be fine. You only need to perform locking for processors of the same type, ex: if a move processor has a locked object, the renderer doesn't care, it can read the current state and draw it before the copy is modified on return. As long as the previous state when the move thread took the pointer was valid, it should be fine. A bit complicated, but doable.
Here's the fun part: It's asynchronous. Spawn 2 threads on the CPU and 1 on the GPU. Asymmetric multiprocessing. Think it'll screw things up? Order of return doesn't matter here, it's first come first serve, so if the GPU thread is 4x faster, it can just grab objects 4x as often. As long as your algorithms return identical results it doesn't matter who does the job.
Of course in a game, not all objects update all the time. If you were using a physics driven system, what you'd want to do is have a vector queue instead. Then grab the kinetic vectors and apply them to the respective objects, check for collision and if so update the existing vectors and then spawn new vector objects into the queue as motion transfers. Cull out vectors below a certain magnitude as "residual noise". Put in friction and the system will keep itself manageable by dropping vectors as they become less significant.
I can't wait to build a physics engine running on this design, but I'm still developing my skills for a project of that magnitude.
I didn't mean the entire ocean, just a little bit. As for the cancer thing, there was a study done where they used very small amounts of antimatter to cause very localized irradiation and kill tumors without causing massive system damage like traditional chemo or radiation therapy. Problem is, it costs a fortune to make the stuff, so nobody can really afford the treatement. That problem goes away when you have say, a billion times the electricity available.
The orbit to ground thing is a problem, yeah. It'd probably have to be reflected or converted to microwave. Fortunately if your satellites are in geosync orbit they're over the same spot and less likely to have microwave beams wandering around cooking people, but accidents happen (see Sim City 2000).
Yes there's meteor strikes, so there'll be repairs. If you can build it, you can fix it. As for hitting the laser, that'd probably need to be rebuilt too (and it would also probably have been locally built) and I was thinking of quite a few along the meridian of the moon that's locked parallel to the earth's surface (i.e. the earth's always on the horizon), so single point failure wouldn't be a problem.
As for what to do with self replicating solar and laser and robot building robots during downtime... hey, free battlebots online via remote control anyone?
We're getting bogged down by energy requirements. Oil is going through the roof, batteries are barely crawling along in improvements, fuel cells still cost a fortune, and everyone is still afraid of the nuclear boogeyman.
What does that leave? Geothermal? Fat chance of seeing that go wide spread. So that leaves solar.
Why the hell is the moon not coated with solar cells? I mean, seriously. Ok let's say we don't want to change how it looks. The bitch is tidelocked! Just put them on the back! Oh but we'll have to go up there and it'll take forever to build! No it won't. Robots, people! I remember reading in Discover around 1992-93 or so about a new all-electrical process someone had developed for extracting materials from sand. He had a bunch of little robots running around the desert building solar cells out of the raw silicon. The moon's got that in spades, and aluminum for the connections. Yeah the efficiency won't be great but who cares when you have an entire MOON (or even half a moon) of them?
How do we get the robots there? Send some. But it won't be enough! Self-replicating. Is this really such a hard challenge? We're seeing basic steps towards it today. Tell me it would cost more than a major space program like a Mars trip to get it working and on the backside of the moon.
How does the power get back? "Laser". But won't it cook the earth? Not if you lock the depression angle so it can only hit geosynchronous orbit and not cross the earth.
But won't people abuse it and fight over it? Declare the moon array itself public domain. Make all the receiving sattelites privatized to create competition and prevent government death rays. Make all the ground stations government owned to prevent slum-shopping for placements by over-greedy immoral corporations. There, you have a case for competition and a nice construction project for all those 3rd world equator countries with the best views of orbit.
What would that get you for your hundred billion or so invesment? UNLIMITED POWER! We wouldn't NEED oil, or fusion, or anything else with that running. Want to use it to go into space? Point the lasers the other way and use them with sails or to power ion drive systems. We'd be mining the asteroid belt with Mark 2 replicating robots in no time. Then we have unlimited energy AND unlimited resources.
Then the real fun starts. Want to end world hunger? Desalinate the ocean and irrigate the entire sahara desert. It'd be cheap. Want to end pollution? Electrochemical reclamation. With virtually free power, post-problem pollution fixes are cheap enough to work. Want to educate everyone? What kind of network can you run when you don't need to worry about electrical losses? Want to cure cancer? There's some promising work with antimatter. Build accelerators to produce it, more efficient ones than the general-purpose kind we have now. Don't want them on earth? Put them on the moon too, make a bigass one around the equator, ship the people there on vacation. Want to get rid of that threatening asteroid headed for earth? Zap it with a petawatt or two before it passes Mars and watch the vapor pressure push it away. Maybe into a nice orbit where we can strip mine it.
All that aside, biotech is going to be the next kick ass field. Read Wired in the last couple years? We can just about cure f'ing BLINDNESS! Eat that you boomer fossils! We're going to see fixes for spinal injuries, better transplants, a doubling of life span, improved prosthetics or maybe even regrown parts. Think some religious-based policies will stop that? Maybe in the US, that's just going to open the door for someone else to take the lead. We're going to be 130 and bitching our great grand kids want tails and wings for xmas and how immoral it is and back in our days we just hijacked cars on playstation and hacked virtual sex in, and that was fine for us!
Well it would probably be best done as a set CPU mode that doesn't expose the cores to the system individually. Kind of a reverse of how hyperthreading makes one core look like two. I'm sure it wouldn't be nearly as good for multithreaded apps, but for those few hard unparallelizable problems, it might be a way to continue performance improvements even though clockspeed seems to be hitting the wall.
But is it feasible? I don't have the tech background to really know. There would likely have to be some kind of internal bus to allow each core to copy the internal state of each pipeline stage of another core so they could resync.
Disclaimer: I am not an electrical engineer
But I think there are several ways to take advantage of multiple cores that are being overlooked.
1. Modern OSes have so many threads running in the background that not having to constantly context switch would probably be a big speedup for single threaded apps like games.
2. The big thing in games is physics. It can be parallelized quite nicely. Dedicated accelerators are still in the vaporware "It's coming any time now!" stage.
3. I'm no CPU designer, but I was thinking. If you had a way to sync state between cores really rapidly, like say in less clocks than a pipeline flush, couldn't you use multiple cores for speculative execution? Let's think about this for a second. The big killer in single thread speed is branching. We spend alot of transistor budget on things like cache and branch prediction to try and reduce it. But what if instead when we hit a branch we forked the execution, so now one core is computing one result, and another the other case. With four cores you could do this two branches deep, with 8 three branches etc. What happens is when the branch actually resolves, the CPU could select the core with the correct prediction, resync the other cores from that CPU's values, and continue execution, without ever having stalled for a branch or flushed the pipeline for a mispredict. As long as your core state resync is less clocks than a flush, it would make sense to do this. Obviously not viable on the dual core P4s now, but with a new design coming, why not? This would let single thread performance scale with core count.
We've got a nice IRC channel for people who don't believe SA has internet. Also our own server, irc.ac.za (though your ping to it will probably suck), and for the clientless, http://www.ircd.co.za/ for a java client to it. Or if you just like forums, http://mybroadband.co.za/vb/
I would have submitted this sooner, but it seems that any time I'm at work, if I try to post a comment on slashdot, it gives me "You can't post to this page." Yet it works fine at home on another ISP. I've tried asking what's going on, if it's some kind of domain ban, but nobody ever replied.
Actually we saw them at E3, filmed some of them to take back and produce our E3 video on the magazine's DVD for July.
Killzone? Yawn. Have you ever actually played the first one? It's not exactly a showstopper. Yay it looks pretty. So what? The 360 games all look on par so far. Where's Devil May Cry 4? Where's a jaw-dropping platformer like Ratchet and Clank or Jak 3? Where's an even better and more absurd Burnout? Why is Soul Calibur 3 a crappy PS2 title that looks if anything worse than the previous one?
Where's ANY ACTUAL RUNNING GAMEPLAY FOOTAGE? All we've seen are "realtime" cinematics. Whoopee. I'd like to see some actual games please.
DSL is commonly oversubscribed at 20:1 and 50:1 ratios to the backbone bandwidth. It's true, look it up. When my cable ISP says "You have 10mbps download." I pull 1200KB/sec off an FTP. The line drops me on an average of less than once a YEAR. Beat that!
Cable bandwidth delivers, it's just the upstream that sucks.
I had no problems with my cablemodem during a power failure.
You remember that big east coast blackout? I live on Long Island. I got home about 7pm, after 2-3 hours crawling through insane traffic because all the lights were down. I unplugged my auto-shutdown desktop, and plugged my laptop into the 1500VA CyberPower UPS that the cablemodem and router share.
I was still on IRC most of the night. People kept asking me "Wait, aren't you in New York?" To which I replied "Yes, and my ISP obviously has a backup generator." My service didn't cut out until around 1am, when the governor got on the radio and told everyone in the state to turn off everything they could so the power grid could be rebalanced and restarted. Then my cable went out and I went to bed.
If your VOIP and cable are dying in a power failure, either you need a UPS, or your ISP does. If you ISP doesn't have one, find one that does, I'm sure someone out there has a reliable business class VOIP and internet solution that won't go down when the lights flicker.
This is exactly what killed it. Sure there were no major gaming studios looking to port the latest PS2 titles to it, but there were tons of indie developers interested in putting popcap-style games and last-generation style 3D games on this thing. But you couldn't thanks to the wonderful PalmOS and the signed code system.
If this had been an open platform based on PocketPC, there would have been alot more games for it. Sad to see it go, my husband and his best friend both own one and got alot of fun out of them, but the promise of this machine never fully materialized.
* Gamecube not competitive enough with PS2
The hardware's superior, where's the software? Looks like Nintendo scared too many developers away with their previous N64 policies, and now they don't have the install base to attract enough anymore.
* Nintendo DS too bulky compared to PSP
This is a load of crap. My husband owns both, he's a professional game reviewer for a major magazine (print, not web). Guess which one fits in his jeans pocket? I'll give you a hint, it's not the one that attracts dust from 20 meters, is so expensive you're constantly nervous about scratching or dropping it, has constant hardware problems (his has a broken UMD latch but no dead pixels fortunately), has almost no titles, has no easy way to find other people to play wireless games with, has been delayed for made up reasons in most of the world, and has "support" in the form of patches that break anything interesting you can do with it.
DSlinux.org and Gamemaker port for the win!
* People are waiting for PS3 (Cell processor + Linux!), not GC3
I think people are waiting for a console they can actually afford that has some games they want. I've seen a good dozen HD trailers of complete or nearly so Xbox 360 games, where's the PS2 titles? Any killer apps yet? I haven't seen any. Maybe developers are having problems dealing with a CPU that has terrible integer performance and a wacky memory system with too little local memory to do a few algorithms like say... collision detection on the SPEs?
Unless the drives support an asynchronous write system, which they do. NCQ will reorder the writes anyway. Latency is primarly a read-side issue, random writes are not as common as random reads.
You can always improve your seek time by adding more redundant mirrors. If we apply the formula the formula seen here where x is the number of redundant mirrors, we can calculate the value of p which will give us our rotational latency for the mean seek time (hence the 0.5 because we want the 50% point for seek times).
Using this you can get 7200rpm drives to easily outseek a 15000rpm drive by using 4 or more redundant sources, and it's still cheaper for the same capacity, AND more failure tolerant.
This is why RAID always wins. Quantity has a quality all its own. SCSI used RAID to defeat the SLED concept in mainfraimes, commodity drives are doing the same to SCSI, by playing with the same rules.
I got out of a major ticket this way when I was 18. The officer had shoddy handwriting so it looked like his odometer in his log contradicted his car's test reccords (which was important because he said I was caught speeding by pacing, not radar). The judge, knowing odometers don't normally go backwards (and also thinking this guy's 4s were 9s), let me off. Which is good because in many states, one moving violation under the age of 21 and you have no license until you're 21, which basically means you're unemployable.
Optimum Online is already testing a 100mbps system to see if they're going to roll it out. Right now they're only running it around 50mbps. I believe the pilot project is in Smithtown, out on Long Island.
Try South Africa's Telkom. Here you can't get DSL without a 3GB per month cap. Of course local traffic and your uploads count on your cap. No you can't get another ISP because all ISPs just resell their lines. You can get DSL that cuts off 3 days into the month if you play onlne games, or you can get 56K and have 1800ms pings to everywhere. Yay.
It depends. If you're cute I wouldn't really mind at all, you're welcome to come over and share him. It wouldn't be the first time.
The funny thing about the internet is you can never assume what kind of people you're talking to.
What a novel concept. Those out there saying "well they should have read the fine print" don't seem to get it. It's not that they expected one thing and got another, it's that even people who know what the deal is don't have a legal option to OWN unrestricted files. It's not presented, at any price. That's where the real problem is.
I figured once DRM got widespread enough to start causing problems with mainstream devices the average Joe (or whatever the name in the UK is) would start taking notice. I've been hearing "But WHY can't I tape my DVD like I do my other tapes?" for awhile now, so I figured it was only a matter of time. The broadcast flag will likely have the same effect. A couple months of nothing major and then suddenly rising complaints of not being able to do the things that were always just fine.
You could be here in South Africa, where people are forced to pay TV licenses (even though the retailers claim they suffer no penalties if they don't comply) to fund complete crap that you can barely recieve anyway. Oh and if you don't like the one station the money goes to? Tough. If you just watch DVDs and play video games and don't have an antenna? Tough. What a crock.