The term "release version" is somewhat nebulous these days. In an age where products seem to be stuck in eternal "beta", the distinction is just relative levels of stability.
In fact, the beta channel of Chrome has the same type of download page (http://www.google.com/landing/chrome/beta/) as the final versions. Only the dev channel doesn't have that nicety. So it'd be more apt to say that Chrome 4.0 *is* a release version, and it supports extensions.
Of course, beta is still not as stable as final, and I can certainly understand that some people would like to wait.
The same is true (that it can be exploited with no user intervention) of an unpatched Linux box vulnerable to various buffer overflow or kernel exploits. The reason why an unpatched Windows box will get rooted so fast is simple market share.
I don't want to get into the debate on who is more proactive on patching than who (we all know that Microsoft doesn't have a stellar reputation in that regard), but please don't act as if they're the only one with issues and that everybody else has a sterling reputation; a lot of your concerns apply to most operating systems, not just Windows.
While you may have had a good experience, Linode was met with silence when they requested that their netblocks be removed.
They weren't even listed for spam, they were listed because they were supposedly dynamic! That's right, a SERVER HOST specializing in providing servers with STATIC IP addresses.
The chrome extension API specifically includes the exact functionality needed for ad blocking via the filter APIs... and yet here we have conspiracy theorists breaking out their tin foil hats and claiming that Chrome is Google's plot to get rid of ad blockers. *facepalm*
The adblock extension I linked above isn't the only one, although it's the only one that I've tried. It's a bit buggy and the UI isn't all there yet, but it does subscribe to the real ABP's easylist, and it *does* block the ads in the list.
Or how about listing entire netblocks because the RDNS of an IP "looks" dynamic?
I'm serious, they've blocked huge swathes of Linode (a virtual server provider) because Linode's default RDNS format (li12-345.members.linode.com) looks dynamic as if such a thing exists.
Linode's attempts to get the netblocks delisted was met with silence; SORBS simply ignores anybody who tries to get delisted.
How can they advertise a 250GB cap when they make it incredibly difficult to reach it?
Their standard tier is 12Mbps. That means that downloading a 1GB file can cause you to become throttled, because it would take longer than 15 minutes at 70% of your capacity.
This means that pretty much any large download (buy a video from iTunes, download a large update to WoW, buy a game on Steam, etc) will throttle you.
I disagree, and I'm not sure what everybody is getting so upset about; Roger McBride Allen's "Second Robot" series does more or less the same thing (adds new books to the Robots/Empire/Foundation universe), and I quite enjoyed it. It also explored variations of the laws in an interesting manner.
I agree... It won't be long before SSDs offer two or three orders of magnitude more performance, and only cost two or there times more than traditional disks.
At some point, the cost difference will simply become irrelevant because it will be small enough. SSDs don't need to cost less to take over the market.
They weren't considering AAC, they were considering AAC+.
Vorbis sucks at low bitrates. AAC+ sucks at high bitrates. Same-codec comparisons would be meaningless; the two codecs are targeted at different use cases.
It's not just bitrate, at this point. They're comaparing 160kbit Vorbis to 48kbit AAC+.
AAC+ utilizes two tricks to make low-bitrate audio sound better: parametric stereo, and spectral band replication.
The first, parametric stereo, stores the audio as monaural with an extremely low-bitrate sideband (2-3kbit/s) to store stereo information.
The second, spectral band replication, stores half the frequency explicitly (low and midrange). The upper frequencies are then recreated from shaped noise, which works quite well.
These two techniques are psychoacoustics taken to the extreme. They're incredible at low bitrates, but useless at higher bitrates.
The quebec system is not in phase/sync with the rest of eastern north america. But why? Is it because of political, geographic or technical reasons?
I was under the impression that other systems were scared of the potential instability and thus the quebec system isn't very strongly tied to the easter north america one.
It's not tied in because we have our own interconnect. Why does Texas have their own? Why does Alaska/BC have their own? Why does the eastern and western US/Canada have separate ones too?
HydroQuebec adds ties to neighbours where the market demands it; where they need to import/export power. For example, HydroQuebec provides the majority of Vermont's electricity.
I'm not sure the exact reason as to why it evolved this way, but I suspect it's simply because of the insular nature of Quebec. This is due in part to the linguistic and cultural differences, but also because Hydro-Quebec began before what we call the "Quiet Revolution" (a period of immense reform and secularization in the 60s).
Personally, I'm happy that we're on our own grid. The US power grid has so many problems that it's nice to be insulated from them. While you cite reliability concerns, on the grid-level, it's proven more reliable than the US grid. Apart from the 1989 solar issue, the only major outage was the ice storm. Considering as that physically destroyed over 3000KM of power lines in the middle of the winter, it's not the kind of thing that any above-ground power grid could survive.
And considering that the ice storm was perhaps the largest/worst natural disaster in our history that caused millions of Quebeckers to put their lives on hold for over a week (some people had no power for 33 days, 90% lost power for 7+) and even abandon their homes (as my family was forced to, admittedly due to the lack of power), it kind of transcends traditional power outages.
and hydroquebec's 735kV lines also get affected by solar storms leading to grid instability and are the reason no other grids are connected to it except by DC.
No, Quebec's power grid is a separate interconnect. THAT is why it is (and must be) connected via DC. The same is true of *ALL* connections between the major interconnects.
(a story about a 20-year old incident)
You're referring to a singular problem TWENTY YEARS AGO, a problem that has not re-occurred since. I'll note that during the big blackout of the northeast a few years back, Quebec was unaffected.
Besides, that 9-hour outage pales in comparison to the multiple-week outage suffered during the 1998 ice storm. HydroQuebec was forced to essentially rebuild the vast majority of the electrical grid from scratch.
Natural disasters are always a concern. You do your best to prepare for them, but every once in a while, you get caught by surprise. It's just the way things are.
Are you seriously trying to argue that obscene quantities of money should be spent on short-distance superconductive power lines just because of solar storms?
So, their response came from the marketing department. That shouldn't have any bearing on the validity of the optimizations itself.
I don't think anybody is denying that offloading to optimally utilize available resources is the problem. But if Intel has not yet had success detecting the need to do so on the fly, is it really such a big deal that they have to profile the applications in advance?
HydroQuebec's 735kV grid has over 11 thousand kilometres of lines, and suffers 4.5 to 8 percent loss depending on environmental and operating conditions.
The power loss over 8.5 miles should be inconsequential...
It adds MLPPP support (bond multiple PPPoE-based broadband connections which you probably don't care about unless your DSL or Cable ISP both uses PPPoE and supports MLPPP), and 1:1 NAT support (which probably has a slightly broader audience). There are other related enhancements, but most of them pertain to MLPPP.
It makes sense; you'd only want to perform these optimizations in games where you're significantly GPU bound. CPU-heavy games, such as Supreme Commander, are probably better off spending the CPU time on the game itself.
I'd see this as more laziness than anything else; it's easier to just hard-code in a list of GPU-bottlenecked games than it would be to actually have your driver auto-detect if there is idle CPU time that could be better spent on offloading.
I don't really see much of an issue with what Intel is doing, though. This behaviour represents their drivers real-world performance and optimizations. How is it cheating to make the best use of available resources?
Heck, their future products will blur the line so much between CPU and GPU that they might well be offloading all sorts of things back and forth. And maybe they'll still be lazy about deciding what games should offload what. Is it still cheating when there's not much difference between your CPU and GPU?
Slashdot Mirror
The term "release version" is somewhat nebulous these days. In an age where products seem to be stuck in eternal "beta", the distinction is just relative levels of stability.
In fact, the beta channel of Chrome has the same type of download page (http://www.google.com/landing/chrome/beta/) as the final versions. Only the dev channel doesn't have that nicety. So it'd be more apt to say that Chrome 4.0 *is* a release version, and it supports extensions.
Of course, beta is still not as stable as final, and I can certainly understand that some people would like to wait.
The same is true (that it can be exploited with no user intervention) of an unpatched Linux box vulnerable to various buffer overflow or kernel exploits. The reason why an unpatched Windows box will get rooted so fast is simple market share.
I don't want to get into the debate on who is more proactive on patching than who (we all know that Microsoft doesn't have a stellar reputation in that regard), but please don't act as if they're the only one with issues and that everybody else has a sterling reputation; a lot of your concerns apply to most operating systems, not just Windows.
They definitely aren't yet in the stable build, but they are in the beta builds.
The Content Script part of the API can be reviewed here:
http://code.google.com/chrome/extensions/content_scripts.html
I don't see anything there that would indicate that the code would not execute before dependencies such as ads would be fetched.
While you may have had a good experience, Linode was met with silence when they requested that their netblocks be removed.
They weren't even listed for spam, they were listed because they were supposedly dynamic! That's right, a SERVER HOST specializing in providing servers with STATIC IP addresses.
The lack of extension support is a myth. As is the supposed lack of adblocking extensions.
The chrome extension API specifically includes the exact functionality needed for ad blocking via the filter APIs... and yet here we have conspiracy theorists breaking out their tin foil hats and claiming that Chrome is Google's plot to get rid of ad blockers. *facepalm*
The adblock extension I linked above isn't the only one, although it's the only one that I've tried. It's a bit buggy and the UI isn't all there yet, but it does subscribe to the real ABP's easylist, and it *does* block the ads in the list.
Or how about listing entire netblocks because the RDNS of an IP "looks" dynamic?
I'm serious, they've blocked huge swathes of Linode (a virtual server provider) because Linode's default RDNS format (li12-345.members.linode.com) looks dynamic as if such a thing exists.
Linode's attempts to get the netblocks delisted was met with silence; SORBS simply ignores anybody who tries to get delisted.
provided you're not stupid enough to run an executable from an untrusted source.
So, in other words, your claims of virus immunity have nothing to do with the operating system itself.
What's so hard to believe about it? RTM was made available well before launch through certain channels, and there's also the very popular free RC.
How can they advertise a 250GB cap when they make it incredibly difficult to reach it?
Their standard tier is 12Mbps. That means that downloading a 1GB file can cause you to become throttled, because it would take longer than 15 minutes at 70% of your capacity.
This means that pretty much any large download (buy a video from iTunes, download a large update to WoW, buy a game on Steam, etc) will throttle you.
I disagree, and I'm not sure what everybody is getting so upset about; Roger McBride Allen's "Second Robot" series does more or less the same thing (adds new books to the Robots/Empire/Foundation universe), and I quite enjoyed it. It also explored variations of the laws in an interesting manner.
I agree... It won't be long before SSDs offer two or three orders of magnitude more performance, and only cost two or there times more than traditional disks.
At some point, the cost difference will simply become irrelevant because it will be small enough. SSDs don't need to cost less to take over the market.
They weren't considering AAC, they were considering AAC+.
Vorbis sucks at low bitrates. AAC+ sucks at high bitrates. Same-codec comparisons would be meaningless; the two codecs are targeted at different use cases.
I don't agree. AAC+ is designed for low-bitrate use, and uses a bunch of techniques targeted at that. Vorbis doesn't.
There is no one codec that is good at all bitrates, unless it incorporates a variety of techniques.
Using the best tool for the job is a perfectly valid comparison, IMO.
It's not just bitrate, at this point. They're comaparing 160kbit Vorbis to 48kbit AAC+.
AAC+ utilizes two tricks to make low-bitrate audio sound better: parametric stereo, and spectral band replication.
The first, parametric stereo, stores the audio as monaural with an extremely low-bitrate sideband (2-3kbit/s) to store stereo information.
The second, spectral band replication, stores half the frequency explicitly (low and midrange). The upper frequencies are then recreated from shaped noise, which works quite well.
These two techniques are psychoacoustics taken to the extreme. They're incredible at low bitrates, but useless at higher bitrates.
Assuming it works by the old "spinning display surface" trick, no. There are only two dimensions at any given point in time.
The quebec system is not in phase/sync with the rest of eastern north america. But why? Is it because of political, geographic or technical reasons?
I was under the impression that other systems were scared of the potential instability and thus the quebec system isn't very strongly tied to the easter north america one.
It's not tied in because we have our own interconnect. Why does Texas have their own? Why does Alaska/BC have their own? Why does the eastern and western US/Canada have separate ones too?
HydroQuebec adds ties to neighbours where the market demands it; where they need to import/export power. For example, HydroQuebec provides the majority of Vermont's electricity.
I'm not sure the exact reason as to why it evolved this way, but I suspect it's simply because of the insular nature of Quebec. This is due in part to the linguistic and cultural differences, but also because Hydro-Quebec began before what we call the "Quiet Revolution" (a period of immense reform and secularization in the 60s).
Personally, I'm happy that we're on our own grid. The US power grid has so many problems that it's nice to be insulated from them. While you cite reliability concerns, on the grid-level, it's proven more reliable than the US grid. Apart from the 1989 solar issue, the only major outage was the ice storm. Considering as that physically destroyed over 3000KM of power lines in the middle of the winter, it's not the kind of thing that any above-ground power grid could survive.
And considering that the ice storm was perhaps the largest/worst natural disaster in our history that caused millions of Quebeckers to put their lives on hold for over a week (some people had no power for 33 days, 90% lost power for 7+) and even abandon their homes (as my family was forced to, admittedly due to the lack of power), it kind of transcends traditional power outages.
and hydroquebec's 735kV lines also get affected by solar storms leading to grid instability and are the reason no other grids are connected to it except by DC.
No, Quebec's power grid is a separate interconnect. THAT is why it is (and must be) connected via DC. The same is true of *ALL* connections between the major interconnects.
(a story about a 20-year old incident)
You're referring to a singular problem TWENTY YEARS AGO, a problem that has not re-occurred since. I'll note that during the big blackout of the northeast a few years back, Quebec was unaffected.
Besides, that 9-hour outage pales in comparison to the multiple-week outage suffered during the 1998 ice storm. HydroQuebec was forced to essentially rebuild the vast majority of the electrical grid from scratch.
Natural disasters are always a concern. You do your best to prepare for them, but every once in a while, you get caught by surprise. It's just the way things are.
Are you seriously trying to argue that obscene quantities of money should be spent on short-distance superconductive power lines just because of solar storms?
So, their response came from the marketing department. That shouldn't have any bearing on the validity of the optimizations itself.
I don't think anybody is denying that offloading to optimally utilize available resources is the problem. But if Intel has not yet had success detecting the need to do so on the fly, is it really such a big deal that they have to profile the applications in advance?
HydroQuebec's 735kV grid has over 11 thousand kilometres of lines, and suffers 4.5 to 8 percent loss depending on environmental and operating conditions.
The power loss over 8.5 miles should be inconsequential...
A better question is, wouldn't it be cheaper to just use normal wires?
HydroQuebec's 735kV lines can carry 2GW each at a distance of up to 1000KM (power consumed in Montreal is produced ~1000KM north).
Would building three 735kV lines really cost more than building a superconductive conduit? I really doubt it...
Tomato/MLPPP.
It adds MLPPP support (bond multiple PPPoE-based broadband connections which you probably don't care about unless your DSL or Cable ISP both uses PPPoE and supports MLPPP), and 1:1 NAT support (which probably has a slightly broader audience). There are other related enhancements, but most of them pertain to MLPPP.
BitBoys' Glaze3D.
And so, if your GPU isn't powerful enough, you would offload as much as you can, no?
It makes sense; you'd only want to perform these optimizations in games where you're significantly GPU bound. CPU-heavy games, such as Supreme Commander, are probably better off spending the CPU time on the game itself.
I'd see this as more laziness than anything else; it's easier to just hard-code in a list of GPU-bottlenecked games than it would be to actually have your driver auto-detect if there is idle CPU time that could be better spent on offloading.
I don't really see much of an issue with what Intel is doing, though. This behaviour represents their drivers real-world performance and optimizations. How is it cheating to make the best use of available resources?
Heck, their future products will blur the line so much between CPU and GPU that they might well be offloading all sorts of things back and forth. And maybe they'll still be lazy about deciding what games should offload what. Is it still cheating when there's not much difference between your CPU and GPU?