Creative makes great hardware They make popular, passable hardware which everyone QA's with because, oh, they're popular. This probably insulates you when they violate the PCI spec and fit things together with spit and duct tape.
All browsers still need to cope with "the animation must be smooth", so even after reaching 100/100 you're still not necessarily passing. I think this mostly revolves around garbage collection; I wouldn't hold my breath if it turns out to need a concurrent collector or so.
Flash media (NAND-gate type) is fundamentally slower than hard drives for sustained serial write behaviour And, er, random writes. You're lucky to get 20/sec because every little write ends up in a read-modify-write of a 4MB block or so.
The upcoming solution to this seems to be to turn random writes into serial ones; presumably buffering up writes in battery backed up memory.
You're right they are rarely useful, but they are ubiquitous - why reproduce one in software? It can tell the difference between multiple drives, and reading or writing. It can even give a good idea of the level of activity.
I get 55MB/s using samba with my GigE network, and that's mostly limited by the STR of my now long-in-the-tooth 400G drives.
That's with a quad GigE card on a PCI-X interface one end (which is basically just 4 £15 Intel 1000/Pro's on a card), and an old motherboard chipset nFarce GigE PHY (PCIe) on the other. I'm pretty sure I got similar performance out of my Intel Mac Mini, but it's been a while since I tested that.
Are you still on an old 32bit PCI system? My last desktop was an AMD K7, and it performed not unlike yours; it'd sync at 1000Mbps but it simply didn't have the bus bandwidth to even make a dent.
The latest batch of consumer drives will pretty much saturate it; Seagate 7200.11's push around 105MB/s, never mind having more than one. Even WD GreenPowers can push 75M/s at 5400RPM, which is more than most GigE chipsets will push comfortably.
I'm certainly looking forward to 10GigE; I remember being annoyed at Fast Ethernet because a single one of my HD's was 4x faster than it; I'm probably only a couple of years away from that with GigE, and already well past it if you take into account the aggregate bandwidth of all the drives in my filestore.
And work? I have a cluster of machines with 25 Savvio 15k's *each*, I'd certainly appreciate being able to sync them at anywhere near the bandwidth they can push out.
The thing about Safari is that it is effectively insulated from SPI changes because it comes out with OS releases *Looks at his WebKit nightly builds*
No, not it doesn't. Well, unless they pick seperate code paths for every OS revision and fall back to public APIs if I try using an old build on a new OS. Ugh. I'm wondering why the fuck the public APIs are so terrible that it's worth all that effort on a large codebase like WebKit; surely if they're experimental you want to be using tools that can easily isolate and measure them. You don't make a better API by saying "hey, it feels smoother today!".
since it's safe for them to do so, where's the harm? There's no monopoly involved, certainly. If it's safe for Safari, it means it's documented sufficiently that they can conditionally use them on the right OS revisions, and it's safe to let anyone else use them with the same caveats.
Really, if you can do one, why is the other so much more trouble that you would ignore it? Here's why. Some appear to have been resolved, but many haven't.
Feel free to hire a developer to get things moving.
Using Broadcom ServerWorks motherboard chipset? Some pretty serious DMA bugs in the HT1000 were worked around pretty recently, not sure if they made it to RC1. There were also reports of problems with the Marvell SATA chipset used on that card, though mine works fine for what little use I have of it.
Quite a few FreeBSD SoC projects make it into the system or ports, or at least had some of their work help with it; a quick glance at the SoC wiki pages I see enhancements to libalias and ipfw (I think some of this eventually made it; we now have kernel NAT with ipfw), bsnmpd bridge monitoring, FUSE port, gvinum enhancements, GEOM storage virtualisation, Apple hardware support enhancements, and what became the name service caching daemon.
Other things may not have made it in, but were good research projects both for the project and for the students; FreeBSD now has a very functional port of OpenBSD's hardware sensors suite, though it wasn't accepted into base because of architectural concerns. gjournal started life as a SoC project, and while rejected it did help spur development of a new more functional one, and the student went on to produce gvirstor, the aforementioned GEOM storage virtualisation layer which *did* make it. The Linux KVM port got far enough to boot FreeBSD 7 as a guest and will hopefully continue development. I'm sure I've left lots out.
Just because a SoC project doesn't make it into a "product", doesn't mean that project wasn't a success. Even if it never produces something deployable, it's given a student some experience in development, it's given the project some interesting if not necessarily immediately useful code and it's helped lay groundwork for future development, even if it only does so by providing those concerned some experience.
expense of running pumps to keep the silo dry (underground facilities like this tend to collect a lot of water, and would rapidly fill up without substantial pumping facilities) A bunker with such severe flooding issues probably won't sell well. The last bunker I saw, the flooding issues were mostly solved by strategic placement of a bucket.
Bunkers are handy against EMP's too. You also generally get good overall physical security with a bunker, and data centers built in them tend to follow through with more practical aspects of security, like escorts instead of just letting you find your own way, as is common in most.
Bunkers also make for a relatively inexpensive readymade secure location, generally with good immortal power and HVAC. People don't put data centers in bunkers because of zomg sekure, they do it because it's often more practical than building your own from scratch.
I remember some oldish sci-fi book where the protagonist made himself very rich developing something like this. Supposedly it worked by making the surface vibrate slightly so that water, dirt, etc simply wouldn't adhere to it.
I'm trying to think what it was.. something by Arthur C. Clarke maybe? This review of The Ghost from the Grand Banks mentions "a really satisfactory windshield wiper". Ah yes, Chapter 3, "A Better Mousetrap", "[the Mark V Wave Wiper] doesn't merely keep off water -- it shakes off any dirt that's already there".
The blindingly obvious realisation that makes him truly filthy rich is that not only can cars make use of it, but that is has huge potential applications for buildings, skyscrapers especially.
ntl:Telewest Business costs £3 more than Virgin's 20Mbit package for 10Mbit, but doesn't have throttling, has call centers in the UK (no crossing your fingers hoping you get Ireland and not India), optional static IP, and supposedly has some sort of SLA for support (6 hours iirc). 20Mbit's supposed to be available soonish too.
Nobody's saying to delay the release of a fix, but to delay release of the details of the issue and the code showing how to exploit it. Pls be improvink ur reading comprehension k tnx.
True, suitably smrt and inclined people can probably look at the fix and come up with the exploit on their own, but there are very few such people compared to the masses of kiddies who just want a readymade exploit they can hack on and use.
I don't see how expressing dipleasure at something on a blog is an overreaction. "Screeching" is stretching it pretty fucking far, since it's basically saying what happened. Where in the blog entry is there screeching, perhaps the bold on "responsible", or maybe the ":("? Wouldn't it be better to link to the blog entry directly and not some dumb opinionated elreg article? Really, did you even read the original source before deciding "the developer needs a chill pill"?
At the end of the day, Mozilla would have acted better by keeping the exploits closed for a few more days, as they would hope anyone else would do for them. By not doing so, they upset people, and others expressing that upset is perfectly understandable. There's no mass outcry at Opera, no press release or open letter saying the Mozilla team are dicks, there's a few words saying what happened and a couple of emoticons on a developer blog entry.
At present hardware will crash if a few bits get in the wrong places or if they're stored incorrectly No, generally it's software which crashes or wildly misbehaves when that happens, unless something critical gets put in an incorrect state or error detection kicks in and *forces* a crash.
There are plenty of algorithms which can cope with flipped bits; for example, one of my favourites is the Bloom Filter; bloom filters let you store whether or not you've seen a given item before, in very little space. They're probablistic; they have a chance of thinking they've seen something before when they really haven't, which goes up as they get full.
Hardware faults resulting in bits getting flipped simply makes them less reliable; they might give a false negitive, or more false positives, but provided the error probabilities are roughly within a known bound they can still be very useful.
You could certainly imagine a data structure for a neural network which is similarly fail-soft; a flipped bit changes the connectivity of a single neuron, or the weight it gives to a certain input, but it doesn't invalidate the entire network.
A lot of apps which depend on such things probably only check CPUID and capabilities of a single CPU, and assume it goes for all available ones; this would be Bad if, say, one of your CPUs had SSE3 and one didn't - an app which can make use of it would only work if it happened to only be scheduled on the more capable CPU (so SSE3's always available), or happened to be scheduled on the less capable one when it did its capabilities check (so it never uses it).
k8temp is a good example; if you have one really old Opteron without a thermal sensor, and one with it, you'd have a 1 in NUM_CORE chance of it refusing to give a reading every time you ran it, since I use the cpuid instruction to detect those capabilities, and that is local to just one CPU.
Core count is unlikely to be an issue in this way; it's easier and more reliable to just ask the OS if that matters to you; it's one thing to use cpuid to check the SSE3 capability bit, it's quite another to use it to find the core count on all the CPUs you might run on now and in the future.
Video encoding's been multithreaded in at least some encoders for a while now, some decoders too. Compression can be done multithreaded (e.g. pbzip2, WinRAR), as can generation of par files (I use concurrent par2cmdline for backups). My audio player's supported running NUM_CPU conversion/ReplayGain threads for years.
And yes, even when apps don't use it themselves, SMP's nice; I was doing it 9 years ago when I got my first BP6 and it was great, despite relatively little other than the OS actually making use of multiple cores; that system lasted me years more than a single socket would have. Increasingly though the CPU heavy things that actually make some people plop down £150+ on one are using multiple cores, and even if some do tend to max out around 2 cores, at least quad core lets you run 2 of them at full pelt; that was an advantage a decade ago enjoyed by people willing to spend a bit more, and it's an advantage now.
Of course what actually tends to be an issue more than available CPU is available IO. Being able to run 4 compression/encoding threads at once isn't so hot when each thread's waiting hundreds or even thousands of ms for a HD to get around to servicing them. I expect as core counts go up we'll probably see more consumer level hardware plonking down the extra £40 for RAID-1, or more.
"Are "divine attributes" well defined in Islam?" Yes. Any references? What makes something a deity?
You do not see consumerism as a polytheism, probably Not really, because I don't see people attributing godlike attributes to things to do with it (those crazy American megachurches aside). However, your idea of what a god is probably doesn't match mine, hence my questions.
You do not see putting human law above God's law as a polytheism No, because both of them are "human law" to me; gods are human concepts, and their laws are ultimately derived from people. That's an illuminating perspective, though, thanks.
Atheism IS Shirk, polytheism, paganism, praying to a Golden Calf. The lack of one overarching deity for a nontheist results in other things taking its place, and thus taking on godlike attributes, then? Like.. for example, I'm awestruck by the concept of evolution and the creative power it reveals, and by not attributing that to one Real Ultimate Power, I'm being polytheistic? That doesn't really match up with what I think of as godlike, but I can see where you're coming from.
This doesn't get me closer to understanding where images fit in, though; it's words which explain and convince. How does abhorring images of animals help?
Slashdot Mirror
All browsers still need to cope with "the animation must be smooth", so even after reaching 100/100 you're still not necessarily passing. I think this mostly revolves around garbage collection; I wouldn't hold my breath if it turns out to need a concurrent collector or so.
The upcoming solution to this seems to be to turn random writes into serial ones; presumably buffering up writes in battery backed up memory.
CFS is a CPU scheduler, and rather similar to FreeBSD's ULE; CFQ and Deadline are IO schedulers, and completely different beasts.
Yes; Jeff Roberson, the author of FreeBSD's new ULE scheduler, wrote a bit about it on his blog a couple of months ago.
I get 55MB/s using samba with my GigE network, and that's mostly limited by the STR of my now long-in-the-tooth 400G drives.
That's with a quad GigE card on a PCI-X interface one end (which is basically just 4 £15 Intel 1000/Pro's on a card), and an old motherboard chipset nFarce GigE PHY (PCIe) on the other. I'm pretty sure I got similar performance out of my Intel Mac Mini, but it's been a while since I tested that.
Are you still on an old 32bit PCI system? My last desktop was an AMD K7, and it performed not unlike yours; it'd sync at 1000Mbps but it simply didn't have the bus bandwidth to even make a dent.
The latest batch of consumer drives will pretty much saturate it; Seagate 7200.11's push around 105MB/s, never mind having more than one. Even WD GreenPowers can push 75M/s at 5400RPM, which is more than most GigE chipsets will push comfortably.
I'm certainly looking forward to 10GigE; I remember being annoyed at Fast Ethernet because a single one of my HD's was 4x faster than it; I'm probably only a couple of years away from that with GigE, and already well past it if you take into account the aggregate bandwidth of all the drives in my filestore.
And work? I have a cluster of machines with 25 Savvio 15k's *each*, I'd certainly appreciate being able to sync them at anywhere near the bandwidth they can push out.
No, not it doesn't. Well, unless they pick seperate code paths for every OS revision and fall back to public APIs if I try using an old build on a new OS. Ugh. I'm wondering why the fuck the public APIs are so terrible that it's worth all that effort on a large codebase like WebKit; surely if they're experimental you want to be using tools that can easily isolate and measure them. You don't make a better API by saying "hey, it feels smoother today!". since it's safe for them to do so, where's the harm? There's no monopoly involved, certainly. If it's safe for Safari, it means it's documented sufficiently that they can conditionally use them on the right OS revisions, and it's safe to let anyone else use them with the same caveats.
Feel free to hire a developer to get things moving.
Using Broadcom ServerWorks motherboard chipset? Some pretty serious DMA bugs in the HT1000 were worked around pretty recently, not sure if they made it to RC1. There were also reports of problems with the Marvell SATA chipset used on that card, though mine works fine for what little use I have of it.
Quite a few FreeBSD SoC projects make it into the system or ports, or at least had some of their work help with it; a quick glance at the SoC wiki pages I see enhancements to libalias and ipfw (I think some of this eventually made it; we now have kernel NAT with ipfw), bsnmpd bridge monitoring, FUSE port, gvinum enhancements, GEOM storage virtualisation, Apple hardware support enhancements, and what became the name service caching daemon.
Other things may not have made it in, but were good research projects both for the project and for the students; FreeBSD now has a very functional port of OpenBSD's hardware sensors suite, though it wasn't accepted into base because of architectural concerns. gjournal started life as a SoC project, and while rejected it did help spur development of a new more functional one, and the student went on to produce gvirstor, the aforementioned GEOM storage virtualisation layer which *did* make it. The Linux KVM port got far enough to boot FreeBSD 7 as a guest and will hopefully continue development. I'm sure I've left lots out.
Just because a SoC project doesn't make it into a "product", doesn't mean that project wasn't a success. Even if it never produces something deployable, it's given a student some experience in development, it's given the project some interesting if not necessarily immediately useful code and it's helped lay groundwork for future development, even if it only does so by providing those concerned some experience.
Bunkers are handy against EMP's too. You also generally get good overall physical security with a bunker, and data centers built in them tend to follow through with more practical aspects of security, like escorts instead of just letting you find your own way, as is common in most.
Bunkers also make for a relatively inexpensive readymade secure location, generally with good immortal power and HVAC. People don't put data centers in bunkers because of zomg sekure, they do it because it's often more practical than building your own from scratch.
I remember some oldish sci-fi book where the protagonist made himself very rich developing something like this. Supposedly it worked by making the surface vibrate slightly so that water, dirt, etc simply wouldn't adhere to it.
I'm trying to think what it was.. something by Arthur C. Clarke maybe? This review of The Ghost from the Grand Banks mentions "a really satisfactory windshield wiper". Ah yes, Chapter 3, "A Better Mousetrap", "[the Mark V Wave Wiper] doesn't merely keep off water -- it shakes off any dirt that's already there".
The blindingly obvious realisation that makes him truly filthy rich is that not only can cars make use of it, but that is has huge potential applications for buildings, skyscrapers especially.
It's supposed to be 3GB, then you get throttled for 5 hours.
ntl:Telewest Business costs £3 more than Virgin's 20Mbit package for 10Mbit, but doesn't have throttling, has call centers in the UK (no crossing your fingers hoping you get Ireland and not India), optional static IP, and supposedly has some sort of SLA for support (6 hours iirc). 20Mbit's supposed to be available soonish too.
Why read-only? Maybe you want to share a crypted volume between machines, not exactly an exotic use-case.
Does Linux's own block device crypto support plausable deniability?
Nobody's saying to delay the release of a fix, but to delay release of the details of the issue and the code showing how to exploit it. Pls be improvink ur reading comprehension k tnx.
True, suitably smrt and inclined people can probably look at the fix and come up with the exploit on their own, but there are very few such people compared to the masses of kiddies who just want a readymade exploit they can hack on and use.
I don't see how expressing dipleasure at something on a blog is an overreaction. "Screeching" is stretching it pretty fucking far, since it's basically saying what happened. Where in the blog entry is there screeching, perhaps the bold on "responsible", or maybe the ":("? Wouldn't it be better to link to the blog entry directly and not some dumb opinionated elreg article? Really, did you even read the original source before deciding "the developer needs a chill pill"?
At the end of the day, Mozilla would have acted better by keeping the exploits closed for a few more days, as they would hope anyone else would do for them. By not doing so, they upset people, and others expressing that upset is perfectly understandable. There's no mass outcry at Opera, no press release or open letter saying the Mozilla team are dicks, there's a few words saying what happened and a couple of emoticons on a developer blog entry.
7200.11/ES.2's seem to have very good NCQ firmware, helps keep bandwidth up in face of heavy seeking compared with most other drives.
I went for the WD GreenPower drives, though; let me drop a fan without my HD's hitting 60+c, even if they are significantly slower in most respects.
There are plenty of algorithms which can cope with flipped bits; for example, one of my favourites is the Bloom Filter; bloom filters let you store whether or not you've seen a given item before, in very little space. They're probablistic; they have a chance of thinking they've seen something before when they really haven't, which goes up as they get full.
Hardware faults resulting in bits getting flipped simply makes them less reliable; they might give a false negitive, or more false positives, but provided the error probabilities are roughly within a known bound they can still be very useful.
You could certainly imagine a data structure for a neural network which is similarly fail-soft; a flipped bit changes the connectivity of a single neuron, or the weight it gives to a certain input, but it doesn't invalidate the entire network.
A lot of apps which depend on such things probably only check CPUID and capabilities of a single CPU, and assume it goes for all available ones; this would be Bad if, say, one of your CPUs had SSE3 and one didn't - an app which can make use of it would only work if it happened to only be scheduled on the more capable CPU (so SSE3's always available), or happened to be scheduled on the less capable one when it did its capabilities check (so it never uses it).
k8temp is a good example; if you have one really old Opteron without a thermal sensor, and one with it, you'd have a 1 in NUM_CORE chance of it refusing to give a reading every time you ran it, since I use the cpuid instruction to detect those capabilities, and that is local to just one CPU.
Core count is unlikely to be an issue in this way; it's easier and more reliable to just ask the OS if that matters to you; it's one thing to use cpuid to check the SSE3 capability bit, it's quite another to use it to find the core count on all the CPUs you might run on now and in the future.
Video encoding's been multithreaded in at least some encoders for a while now, some decoders too. Compression can be done multithreaded (e.g. pbzip2, WinRAR), as can generation of par files (I use concurrent par2cmdline for backups). My audio player's supported running NUM_CPU conversion/ReplayGain threads for years.
And yes, even when apps don't use it themselves, SMP's nice; I was doing it 9 years ago when I got my first BP6 and it was great, despite relatively little other than the OS actually making use of multiple cores; that system lasted me years more than a single socket would have. Increasingly though the CPU heavy things that actually make some people plop down £150+ on one are using multiple cores, and even if some do tend to max out around 2 cores, at least quad core lets you run 2 of them at full pelt; that was an advantage a decade ago enjoyed by people willing to spend a bit more, and it's an advantage now.
Of course what actually tends to be an issue more than available CPU is available IO. Being able to run 4 compression/encoding threads at once isn't so hot when each thread's waiting hundreds or even thousands of ms for a HD to get around to servicing them. I expect as core counts go up we'll probably see more consumer level hardware plonking down the extra £40 for RAID-1, or more.
No, it's nothing to do with virtualization, it's about interacting with pipes. vmsplice() lets you put data into a pipe without having to copy it.
I can't find a configuration option for it; it's probably a bit fine-grained even for Linux.
This doesn't get me closer to understanding where images fit in, though; it's words which explain and convince. How does abhorring images of animals help?