It doesn't make sense to store the data in a complex format to reduce disk space requirements. It just doean't scale. You don't need "oodles" of disk space either; in fact holding every zoom level (assuming power-of-two zoom levels, which is what everybody uses including Google) only requires 1/3 more space than holding the highest zoom level by itself (not 2 or 5 times more space as was speculated earlier). OTOH producing the images on the fly and encoding them to JPEGs on every request would require a beefy server, or server farm if you're talking about decent amounts of traffic. And it would still crumble under unexpected heavy loads. Better to buy 1/3 more hard drives and not worry about on-the-fly image processing.
If you think the majority of people would sit through three minutes of ads at their computer while the "seek" bar of their video player was just sitting there unused, you are mistaken. If many videos started having ads, the next generation of open-source video players would come with 30-second-skip buttons, guaranteed. People might even go so far as to distribute files matched to each episode that mark where the commercials are so your player can skip them for you automatically.
Also, Internet TV doesn't have the same "plop on the couch and veg out for a few hours" vibe. People won't be zoning out in front of their computers, content to watch whatever drivel the networks shove at them just because they're too lazy to change the channel. Instead they will be preferentially downloading the content they have heard about and want to watch and not downloading anything else.
Basically, Internet TV means you need better content with less skippable ads, and you have to beat the pirates on price and convenience. Right now the audience is pretty small too, compared to regular TV. TV execs are already petrified about TiVo; it's no wonder they haven't decided to go for Internet TV. I don't see them embracing Internet TV until the pirates really ram it down their throats, and even then grudgingly; same as the music industry.
GIMPShop doesn't make the GIMP into an MDI application with tool panels and images in one window, which is the usual request. It simply renames some tools and rearranges the menus to be the same as in Photoshop. The rearranged menus are probably quite useful to new photoshop converts. However, if this gets popular, I'll bet Adobe will try some legal action to stop it. It is after all a blatant copy of much of the text in their user interface.
If it is true that the music store transfers the song with no encryption applied, you could make a better application that would still allow you to use iTunes to browse the music store and purchase the music. It would simply capture the song as it is being transferred normally between iTunes and the music store. Apple would be powerless to detect this.
but all this does is shift the pollution elsewhere
This is the stupidest argument ever, and every time I hear it repeated my faith in human intelligence decreases. Well anyway, here's the standard demolition of that argument:
Centralized power production is desirable because...
Pollution generation can be moved, for example to keep it away from fragile ecosystems or large population centers. (This is probably the least compelling reason.)
Large power plants can use very large, unweildy, and expensive but highly effective means of filtering pollution from their output, decreasing the amount of pollution added to the environment even from the same energy sources.
Large power plants can be very much more efficient than car engines, reducing the amount of energy wasted overall.
Power plants can use forms of energy that would be impractical to use in cars such as wind, hydroelectric, and geothermal power.
Power generation can be easily switched to newer, more efficient, cleaner forms of energy production by simply building a new plant, instead of replacing hundreds of millions of cars.
Thus "shifting the pollution elsewhere" would actually reduce pollution immediately *and* encourage the use of alternative energy sources now and in the future.
More like.5 to 5 FPS at just 1024x768 on reasonably sized scenes with no antialiasing, soft shadows, subsurface scattering, caustics, or global illumination. The soft shadow example they gave looked like it was using about 5 rays per pixel, way too few for a good-looking soft shadow. As for subsurface scattering, caustics, and global illumination, they didn't demonstrate any of that as far as I saw (though I didn't download everything, the site was too slow). Certainly you can't do all those effects with a measley 2-5 rays per pixel as in most of their demonstration shots.
Huh? If the reflections in Doom 3 failed to impress, it's not the fault of environment mapping as a technique. Have you seen Project Gotham Racing 2? There's some compelling realtime reflections in that game.
Why do you assume that I have not considered the "long view"? The previous post *is* my consideration of the "long view". My argument is not that it's too slow right now; my argument is that it will *always* be slower than polygon rasterization. Perhaps in the future more advanced techniques will be able to be implemented in realtime raytracers, but by that time rasterizers will be implementing even more advanced techniques, staying a generation ahead.
Bah. Raytracing is not required for good graphics. Pixar's Photorealistic RenderMan didn't even have raytracing until version 11, which came out *after* Monsters, Inc.
Raytracers can easily do hard shadows, reflection, refraction, and order-independent transparency. Today's rasterizers can do almost all that too: hard shadows (stencil shadows), and "good enough" reflections and refractions (using environment maps and shaders). Order-independent transparency is a tough one; it can be kludged using shaders, but it is often better simply to work around it.
Realtime raytracing is a dead end, because all of the techniques that make offline raytraced images good (soft shadows, subsurface scattering, caustics, global illumination, etc) are too slow to implement in a real time raytracer. Rasterizing renderers require hacks to simulate many things that raytracing does more naturally, but those hacks run tens if not thousands of times faster than their more physically accurate raytraced equivalents. What those hacks lose in accuracy they gain back in speed, essentially producing more image quality per unit time. And in real-time graphics, time is the most important thing.
I don't know how Google decides which ads to serve to which people; however I still get the ad. I know other people do too, since it was posted on a friend's blog and several other people commented on it.
Iraq War
Huge selection, great deals on
Iraq War. Aff
eBay.com
What really annoys me more than the link is the fact that this ad was obviously generated by a computer and never proofread by a human. It contains no useful information about what is being advertised. If the same ad said "Iraq War Merchanise; Auctions for pins, stickers, shirts, photos, videos, and more" then it would be fine. I might even be interested in it if I knew I wanted some pins and stickers related to the war, or something.
Google needs standards for their ads: they should not be computer-generated, they should contain relevant information about the products being advertised, they should link directly to the advertised products, and they should definitely not contain the search keywords more than once.
How about Google AdSense stop sucking? When it first came out, the ads were great because they were relevant, timely, and interesting, with links that almost always pointed straight to the page you were interested in. Now, many of the ads are obviously generated by bots picking keywords at random, and the links often point only to a site's homepage instead of the relevant page you want to see.
Searching for anything will just get you EBay affiliate ads claiming to sell it, regardless of whether EBay does or not, or whether what you're searching for is even a physical commodity. Case in point. Some of the links you get when you click "more sponsored links" are much more interesting and relevant than EBay, but somehow EBay is always at the top of the results. EBay's not the only offender either; they're just the most obnoxious.
Remember those stupid search engine ads that would just say "Search for [your search terms] on [shopping service] now?", no matter what you typed in? That's what AdWords is turning into. Google really needs to enforce some kind of quality and relevance standard on ads, because I've just started ignoring them, the same as every other kind of crappy irrelevant web ad.
I have also noticed that Google fails to autocorrect some common misspellings of words since the change. Answers.com has pages for some misspelled words (did you mean...), and Google links to those instead of indicating that the word is misspelled. Caught me off-guard when I was using Google as a spell-checker...
I'm not sure what causes it to ask, it doesn't ask all the time (perhaps you have to download a certain amount first). But when it does, the dialog is big, flashy, and annoying, with a delay before you can dismiss it like those annoying shareware nag screens. It just happened to me when I started Azureus to compare with the new official client, and it's happened before.
if bandwidth is at a premium, the routers can drop, or throttle, the bittorrent data to make room for high priority data.
No. In order for users to voluntarily mark their packets as "bulk data", there has to be a benefit for them. That benefit is supposed to be higher overall transfer rate. The tradeoff is higher latency. So a router that receives a BitTorrent packet and a VOIP packet at the same time would send the VOIP packet first to reduce latency, and queue the BitTorrent packet for afterwards. But if the queue is full it would *not* preferentially drop the BitTorrent packet because that would reduce throughput. In fact, if the queue has many VOIP packets, the router should preferentially drop incoming VOIP packets, because it would not be able to send them with low latency anyway. This limits VOIP throughput, which is fine. In fact that's the result we want: VOIP = low latency low throughput, BitTorrent = high latency high throughput.
At least, I hope this is how ISPs implement routing for packets marked as bulk data, because otherwise it will never be adopted.
After trying the new official BitTorrent client, I'm going to dump Azureus. Azureus is good, but it's big and clunky and always asking me things (do you want to download the update? do you want to restart now? do you want to donate, or would you prefer to wait until we decide you can click no?).
The official client's new interface is streamlined and simple. It installs in a second (literally). It has all of the options that I need (I'm a simple man, I just want to download some files). And it never asks me annoying questions while I'm just trying to download something.
In short, the new interface is great; I wholeheartedly recommend it (unless you have a fetish for tabs and arcane configuration dialogs, in which case by all means continue using Azureus).
Just because it was patented doesn't mean it's a good idea of course; I believe the amount of power that can be generated by this method is too small to be of much use, even at theoretical 100% efficient conversion. Instead of allocating extra space, weight, and manufacturing cost for this complex charger system, simply including a bigger battery will improve battery life more. There is some interesting discussion along these lines in the original Slashdot story about the patent (also from '99).
This is a good thing. If there are no victims, then there was no crime. This requirement stops companies from going after (for example) parody sites, claiming that consumers might be confused. This new law could quite easily make parody sites illegal, or it could even have loopholes that enable companies to shut down large classes of sites that simply use their trademarked name or some other BS. If law enforcement just can't figure out who the victims of the real scams are, perhaps we need better, more tech savvy law enforcement instead of more laws "outlawing" things that are already illegal.
More laws == more complex legal system == more loopholes == you need the advice of a lawyer to do anything at all in this country anymore. Let's not make it worse.
The problem with your approach is that phase 1 will never end, and nobody will want to work on phase 2. Andrew Morton's not going to stop writing patches, and neither is anyone else. You've gotta cut the updates off sometime if you want to release within our lifetimes. This means that some good code won't make it in, but it isn't the end of the world if the release cycle is short enough; you can just add it next cycle and it will be in a stable released kernel in a month or two.
If you look at Linus's proposal, it's just your idea with short release cycles. He gives an odd version number to the end of your "Phase 1" and an even one to the end of your "Phase 2". It sounds like an excellent proposal to me.
Nothing in the article says they're opening up to alternative AIM clients. They are allowing companies to "partner" with them, probably involving large licensing fees, to add AIM "presence" (on/offline information) into their products and produce approved plugins for the advertising-riddled official client. That doesn't help GAIM or Adium or Jabber or any other open-source project. It probably doesn't help Trillian either.
It is non-volatile by nature. But it is not likely to be fast enough to replace RAM. Instead it could replace Flash memory or even (depending on cost) hard drives. The real question is, how long until it's practical to manufacture and use in mass-produced products? The answer seems to be (according to the article) 2007-2010 timeframe.
lets think about, what SSL is supposed to do: Encrypt and secure the traffic between a browser and the server! Point! It is not supposed to give you the impression, that a website is trustworthy or even say anything about its identity.
Um, no. Let's *really* think about what SSL is supposed to do: it is supposed to allow you to communicate to (for example) PayPal through routers controlled by (for example) an evil mastermind, but ensure that the evil mastermind can't read or modify your data, even though he is the one delivering it back and forth. The evil mastermind ends up with a complete transcript of your conversation (including the part where you send encryption keys back and forth), but he can't read any of it. Pretty cool actually. It is rather surprising (to me at least) that this is even possible.
However, without verification of identity, the whole scheme falls apart. Sure, you can still establish an encrypted connection, to somebody. But you don't know who! You could have just established an encrypted connection with the evil mastermind, who has programmed his routers to deliver your packets to him instead of PayPal. The encryption doesn't help so much when you send your enemies the key! If the evil mastermind is really clever, he will then establish a second encrypted connection to PayPal, and send your data through it. Now he can read your entire conversation with PayPal, just as if you weren't using any encryption at all, and you are no better off.
This is your basic "man-in-the-middle" attack, which everybody who has ever learned anything about security should know. I feel stupid even explaining it here (you knew this right? good), but apparently these guys haven't heard of it.
That's why SSL's identity verification is critical, not useless, and why VeriSign gets paid the money they do. Sure you can get your certs signed by Joe Blow for free. But if anybody can get a cert signed by Joe Blow (perhaps after a little bribe, say $1,000,000; chump change compared to how much organized crime could steal with a key to a large commerce site) then your cert is basically worthless, even for "only" encrypting traffic.
You're completely right; like I said, I certainly don't recommend buying one of these right now for gaming. It's obviously stupid (look, I can run doom 3 without texture compression!!!1). But again, like I said, if you do 3D modeling, visualization, offline rendering, vector processing, or work on experimental accelerated GUIs, then it's entirely possible for you to find a use for that much video RAM today.
What is likely happening is: the card is hardly used at all; it is sitting there idle most of the time. The bottleneck is elsewhere. Specifically, since you don't have enough video RAM to store all the textures, the rest are stored in system RAM and copied over the AGP or PCI Express bus every frame (like el cheapo video cards used to do). This is the reason video RAM exists: using main memory over an AGP or PCI Express connection is extraordinarily slow in comparison.
Slashdot Mirror
It doesn't make sense to store the data in a complex format to reduce disk space requirements. It just doean't scale. You don't need "oodles" of disk space either; in fact holding every zoom level (assuming power-of-two zoom levels, which is what everybody uses including Google) only requires 1/3 more space than holding the highest zoom level by itself (not 2 or 5 times more space as was speculated earlier). OTOH producing the images on the fly and encoding them to JPEGs on every request would require a beefy server, or server farm if you're talking about decent amounts of traffic. And it would still crumble under unexpected heavy loads. Better to buy 1/3 more hard drives and not worry about on-the-fly image processing.
Also, Internet TV doesn't have the same "plop on the couch and veg out for a few hours" vibe. People won't be zoning out in front of their computers, content to watch whatever drivel the networks shove at them just because they're too lazy to change the channel. Instead they will be preferentially downloading the content they have heard about and want to watch and not downloading anything else.
Basically, Internet TV means you need better content with less skippable ads, and you have to beat the pirates on price and convenience. Right now the audience is pretty small too, compared to regular TV. TV execs are already petrified about TiVo; it's no wonder they haven't decided to go for Internet TV. I don't see them embracing Internet TV until the pirates really ram it down their throats, and even then grudgingly; same as the music industry.
GIMPShop doesn't make the GIMP into an MDI application with tool panels and images in one window, which is the usual request. It simply renames some tools and rearranges the menus to be the same as in Photoshop. The rearranged menus are probably quite useful to new photoshop converts. However, if this gets popular, I'll bet Adobe will try some legal action to stop it. It is after all a blatant copy of much of the text in their user interface.
If it is true that the music store transfers the song with no encryption applied, you could make a better application that would still allow you to use iTunes to browse the music store and purchase the music. It would simply capture the song as it is being transferred normally between iTunes and the music store. Apple would be powerless to detect this.
This is the stupidest argument ever, and every time I hear it repeated my faith in human intelligence decreases. Well anyway, here's the standard demolition of that argument:
Centralized power production is desirable because...
- Pollution generation can be moved, for example to keep it away from fragile ecosystems or large population centers. (This is probably the least compelling reason.)
- Large power plants can use very large, unweildy, and expensive but highly effective means of filtering pollution from their output, decreasing the amount of pollution added to the environment even from the same energy sources.
- Large power plants can be very much more efficient than car engines, reducing the amount of energy wasted overall.
- Power plants can use forms of energy that would be impractical to use in cars such as wind, hydroelectric, and geothermal power.
- Power generation can be easily switched to newer, more efficient, cleaner forms of energy production by simply building a new plant, instead of replacing hundreds of millions of cars.
Thus "shifting the pollution elsewhere" would actually reduce pollution immediately *and* encourage the use of alternative energy sources now and in the future.More like .5 to 5 FPS at just 1024x768 on reasonably sized scenes with no antialiasing, soft shadows, subsurface scattering, caustics, or global illumination. The soft shadow example they gave looked like it was using about 5 rays per pixel, way too few for a good-looking soft shadow. As for subsurface scattering, caustics, and global illumination, they didn't demonstrate any of that as far as I saw (though I didn't download everything, the site was too slow). Certainly you can't do all those effects with a measley 2-5 rays per pixel as in most of their demonstration shots.
Huh? If the reflections in Doom 3 failed to impress, it's not the fault of environment mapping as a technique. Have you seen Project Gotham Racing 2? There's some compelling realtime reflections in that game.
Why do you assume that I have not considered the "long view"? The previous post *is* my consideration of the "long view". My argument is not that it's too slow right now; my argument is that it will *always* be slower than polygon rasterization. Perhaps in the future more advanced techniques will be able to be implemented in realtime raytracers, but by that time rasterizers will be implementing even more advanced techniques, staying a generation ahead.
Raytracers can easily do hard shadows, reflection, refraction, and order-independent transparency. Today's rasterizers can do almost all that too: hard shadows (stencil shadows), and "good enough" reflections and refractions (using environment maps and shaders). Order-independent transparency is a tough one; it can be kludged using shaders, but it is often better simply to work around it.
Realtime raytracing is a dead end, because all of the techniques that make offline raytraced images good (soft shadows, subsurface scattering, caustics, global illumination, etc) are too slow to implement in a real time raytracer. Rasterizing renderers require hacks to simulate many things that raytracing does more naturally, but those hacks run tens if not thousands of times faster than their more physically accurate raytraced equivalents. What those hacks lose in accuracy they gain back in speed, essentially producing more image quality per unit time. And in real-time graphics, time is the most important thing.
Iraq War
Huge selection, great deals on
Iraq War. Aff
eBay.com
What really annoys me more than the link is the fact that this ad was obviously generated by a computer and never proofread by a human. It contains no useful information about what is being advertised. If the same ad said "Iraq War Merchanise; Auctions for pins, stickers, shirts, photos, videos, and more" then it would be fine. I might even be interested in it if I knew I wanted some pins and stickers related to the war, or something.
Google needs standards for their ads: they should not be computer-generated, they should contain relevant information about the products being advertised, they should link directly to the advertised products, and they should definitely not contain the search keywords more than once.
Searching for anything will just get you EBay affiliate ads claiming to sell it, regardless of whether EBay does or not, or whether what you're searching for is even a physical commodity. Case in point. Some of the links you get when you click "more sponsored links" are much more interesting and relevant than EBay, but somehow EBay is always at the top of the results. EBay's not the only offender either; they're just the most obnoxious.
Remember those stupid search engine ads that would just say "Search for [your search terms] on [shopping service] now?", no matter what you typed in? That's what AdWords is turning into. Google really needs to enforce some kind of quality and relevance standard on ads, because I've just started ignoring them, the same as every other kind of crappy irrelevant web ad.
I have also noticed that Google fails to autocorrect some common misspellings of words since the change. Answers.com has pages for some misspelled words (did you mean...), and Google links to those instead of indicating that the word is misspelled. Caught me off-guard when I was using Google as a spell-checker...
I'm not sure what causes it to ask, it doesn't ask all the time (perhaps you have to download a certain amount first). But when it does, the dialog is big, flashy, and annoying, with a delay before you can dismiss it like those annoying shareware nag screens. It just happened to me when I started Azureus to compare with the new official client, and it's happened before.
No. In order for users to voluntarily mark their packets as "bulk data", there has to be a benefit for them. That benefit is supposed to be higher overall transfer rate. The tradeoff is higher latency. So a router that receives a BitTorrent packet and a VOIP packet at the same time would send the VOIP packet first to reduce latency, and queue the BitTorrent packet for afterwards. But if the queue is full it would *not* preferentially drop the BitTorrent packet because that would reduce throughput. In fact, if the queue has many VOIP packets, the router should preferentially drop incoming VOIP packets, because it would not be able to send them with low latency anyway. This limits VOIP throughput, which is fine. In fact that's the result we want: VOIP = low latency low throughput, BitTorrent = high latency high throughput.
At least, I hope this is how ISPs implement routing for packets marked as bulk data, because otherwise it will never be adopted.
The official client's new interface is streamlined and simple. It installs in a second (literally). It has all of the options that I need (I'm a simple man, I just want to download some files). And it never asks me annoying questions while I'm just trying to download something.
In short, the new interface is great; I wholeheartedly recommend it (unless you have a fetish for tabs and arcane configuration dialogs, in which case by all means continue using Azureus).
Just because it was patented doesn't mean it's a good idea of course; I believe the amount of power that can be generated by this method is too small to be of much use, even at theoretical 100% efficient conversion. Instead of allocating extra space, weight, and manufacturing cost for this complex charger system, simply including a bigger battery will improve battery life more. There is some interesting discussion along these lines in the original Slashdot story about the patent (also from '99).
It was patented in 1999 (the keyboard idea). here.
More laws == more complex legal system == more loopholes == you need the advice of a lawyer to do anything at all in this country anymore. Let's not make it worse.
And get some better colors too, as long as we're dreaming that the Slashdot editors actually care what our opinions are.
If you look at Linus's proposal, it's just your idea with short release cycles. He gives an odd version number to the end of your "Phase 1" and an even one to the end of your "Phase 2". It sounds like an excellent proposal to me.
Nothing in the article says they're opening up to alternative AIM clients. They are allowing companies to "partner" with them, probably involving large licensing fees, to add AIM "presence" (on/offline information) into their products and produce approved plugins for the advertising-riddled official client. That doesn't help GAIM or Adium or Jabber or any other open-source project. It probably doesn't help Trillian either.
It is non-volatile by nature. But it is not likely to be fast enough to replace RAM. Instead it could replace Flash memory or even (depending on cost) hard drives. The real question is, how long until it's practical to manufacture and use in mass-produced products? The answer seems to be (according to the article) 2007-2010 timeframe.
Um, no. Let's *really* think about what SSL is supposed to do: it is supposed to allow you to communicate to (for example) PayPal through routers controlled by (for example) an evil mastermind, but ensure that the evil mastermind can't read or modify your data, even though he is the one delivering it back and forth. The evil mastermind ends up with a complete transcript of your conversation (including the part where you send encryption keys back and forth), but he can't read any of it. Pretty cool actually. It is rather surprising (to me at least) that this is even possible.
However, without verification of identity, the whole scheme falls apart. Sure, you can still establish an encrypted connection, to somebody. But you don't know who! You could have just established an encrypted connection with the evil mastermind, who has programmed his routers to deliver your packets to him instead of PayPal. The encryption doesn't help so much when you send your enemies the key! If the evil mastermind is really clever, he will then establish a second encrypted connection to PayPal, and send your data through it. Now he can read your entire conversation with PayPal, just as if you weren't using any encryption at all, and you are no better off.
This is your basic "man-in-the-middle" attack, which everybody who has ever learned anything about security should know. I feel stupid even explaining it here (you knew this right? good), but apparently these guys haven't heard of it.
That's why SSL's identity verification is critical, not useless, and why VeriSign gets paid the money they do. Sure you can get your certs signed by Joe Blow for free. But if anybody can get a cert signed by Joe Blow (perhaps after a little bribe, say $1,000,000; chump change compared to how much organized crime could steal with a key to a large commerce site) then your cert is basically worthless, even for "only" encrypting traffic.
You're completely right; like I said, I certainly don't recommend buying one of these right now for gaming. It's obviously stupid (look, I can run doom 3 without texture compression!!!1). But again, like I said, if you do 3D modeling, visualization, offline rendering, vector processing, or work on experimental accelerated GUIs, then it's entirely possible for you to find a use for that much video RAM today.
What is likely happening is: the card is hardly used at all; it is sitting there idle most of the time. The bottleneck is elsewhere. Specifically, since you don't have enough video RAM to store all the textures, the rest are stored in system RAM and copied over the AGP or PCI Express bus every frame (like el cheapo video cards used to do). This is the reason video RAM exists: using main memory over an AGP or PCI Express connection is extraordinarily slow in comparison.