Not all XEONs have hardware virtualization. Only some of the most expensive chips have it and even then, it can be spotty.
Not true. Every Xeon since 2006 has shipped with VT-x support. Look at the Xeon 5030 for example. Absolute bottom of the line ($150 at launch) from 2006, and it supports VT-x.
You're probably thinking of Intel's desktop line where to do artificially hobble large swaths of their CPUs with respect to VT-x.
Yes, that is the disgustingly awful part about C-11, but you missed the upside:
The goverment eventually arrived a trade-off that most Canadians would make: a tougher provision to target sites that facilitate infringement (the law already allows rights holders to do this) in return for a full cap on liability for non-commercial infringement. This applies not only to individuals (likely bringing to an end the prospect of file sharing lawsuits in Canada) but to any non-commercial entity including educational institutions and libraries (who may adopt more aggressive interpretations of the law with less risk of liability).
Drivers, installed base, drivers, familiar windows interface, drivers, most users can barely power their machine on much less install linux, drivers, forget installing linux software...see comment before the last comment, drivers, lack of vendor support, and drivers.
Oh did I mention drivers?
You play weird video games. Personally, I like playing the "my computer works already, I didn't have to hunt down twenty drivers from twenty different sites and make sure I kept them all up to date individually" game, that's why I already use Linux (and have for nearly a decade).
Just wondering: Is there a point (or is this close to it) where in using HDDs and certain RAID configurations, you can match or beat speed while maintaining better redundancy with larger capacity, cheaper drives? What is the main application these excel at? I assume power would be one, and cached content on webservers? Help me understand:-)
You'd need several dozen hard drives to even approach the IOPS of a single consumer level SSD. The SSD wins so many times over it's not funny.
Now, if you're talking about sequential read/write speeds, that's a whole different matter. You'd need roughly 3-4 hard drives (in RAID 0 (no redundancy)... double that figure for RAID 10) to match the typical sequential read/write speeds of an SSD. At that point, the raw cost of the hard drives far exceeds that of the SSD, and that's ignoring the need for the extra SATA ports, cooling, physical space and the extra drive failures you need to deal with. So, the SSD wins again, hands down.
Now, say you needed to store more than roughly 200 gigabytes of data and performance didn't matter at all, in that case, hard drive(s) will be more cost effective than SSDs.
Basically, hard drives excel at bulk storage of stuff where performance doesn't matter. SSDs excel at everything else.
A poorly designed merge section from one highway to another is what convinced me I needed a quick car.
It isn't safe merging into 60+ MPH traffic at 30 MPH. Top speed typically isn't a problem but acceleration on cars with wimpy engines is.
If only more people understood this. 99% of the time when I'm behind someone while merging onto the highway, they'll be going way under the speed limit as they merge. It is infuriating.
Re:Still no TRIM on software RAID (md)
on
Linux 3.6 Released
·
· Score: 3, Informative
Support for TRIM on RAID linear/0/1/10 md devices was quite recently added. The patch series is here: https://lkml.org/lkml/2012/3/11/261. I can't find the actual merge now, but I believe it'll be in 3.7.
Now this seriously bums me out, as I believe in competition and have been building AMD exclusively for the past few years, ever since the OEM bribery and compiler scandals came out. But since MSFT has already said they aren't gonna backport the scheduler fix (they have only released a patch, since withdrawn I believe) and thanks to this boneheaded design you'll have a boat anchor tied on your system you really only have 4 choices, 1.-Disable half the cores, 1 per module, so you are basically only getting half of what you paid for but each core then has a full FP unit,2.- OC the living hell out of it to use speed to make up for the penalty, 3.-Stick with the AM3 Phenom II units, this has been what I've been doing as the Phenom II quads and hexacores are dirt cheap now and still have decent speed, or 4.-Don't buy AMD.
You're missing the elephant in the room... don't use Windows.
A CS degree is a requisite but not sufficient property to make a good developer. They also need a genuine interest in the field, which most often manifests as being self-taught before getting a degree, and continuing to self-teach after getting said degree.
Purely self-taught developers will miss learning a lot of important topics, not because they're difficult, but because they don't realize what they don't know. In particular, data structures (anything beyond arrays), databases (and normal forms) and algorithmic complexity. I don't care how good you are with $language, if you don't understand the above topics like the back of your hand, you're going to make a mess.
On the flip side, purely academic developers are typically going to have knowledge gaps in more practical topics like input validation, version control systems and bug trackers. Again, you can get by without these, but you're going to make a mess (or someone else is going to make a mess of it for you when they exploit it).
I'm in favor, but there is a problem. There are a great many makes and models that CAN NOT maintain 85mph. And many many more than that can't do it safely.
So? Driving is a priviledge, not a right. Should we also give up on driving exams and just let everybody drive?
As an added bonus, you'll stimulate the economy (through car manufacturers).
People refuse to obey the speed limits because the speed limits are retarded. Take any major highway in North America and you'll find massive stretches of more or less completely straight road where there's no reason you couldn't drive all day at your car's top speed, except, the posted speed limit is a third (or less) of said top speed.
This will never change because the government strongly prefers to keep everyone a criminal, they're much easier to control that way. If speed limits were strictly enforced (and not increased to sane values), there would be riots in the streets.
Don't be silly, insurance rates will rise across the board. After all, all these fancy new electronics make the cars more expensive and therefore more expensive to repair in a crash, therefore, everybody's insurance rates need to go up. Furthermore, insurance rates will rise even more for those of us who refuse to enable the tracking data, as obviously only filthy terrorists would value their privacy.
Especially since I can't rightly say I have a better plan and neither does Mr. deRaadt.
The better plan is to sue Microsoft for abuse of their monopoly.
Perfect, then we can wait a decade for the case to go anywhere, only to have it thrown out in the end and all computers made within the past decade remain unusable.
TFA is talking about in reverse proxies (of which Varnish is one of many), which are very commonplace. In fact, you're seeing this page through (at least) one, as Slashdot uses Varnish.
The importance of this question is not to be underestimated. Anything that adds availability also adds complexity, and if not done correctly will actually worsen your availability. More than once I've seen new "high availability" setups that were less reliable than the original "non-high availability" setup.
If you want to protect against hardware failure, get a server with redundant power supplies and hard drives (can be had for under $2k), and monitor them. Now, redundant power supplies are a no-brainer, but keep in mind that for hardware RAID (both fakeraid and real hardware RAID), when your controller dies all your data is gone unless you have an identical replacement controller. You can go further and add redundant network cards plus redundant network paths (including switches), but again there's a non-trivial complexity cost associated.
If you want availability during updates, then your life just got a whole lot more complicated.
How often do you ever accelerate flat out from 0 to 60?
Way more often than I go over 200 km in one trip. They need to drop the 0-60 time back into the low 4s range like the roadster, and halve the range if need be.
FWIW, the GTX 680 supports up to four displays per (single GPU) card. I haven't tried it yet, but I would expect it to work much better than the 4 monitors across 2 cards support that one was limited to in the past.
Multimonitor across multiple cards has never been stellar in any OS, probably because it's a very niche use case, but multimonitor on a single card is a very common use case.
ALSA's default config since the early 2000s has allowed multiple simultaneous sounds to be played without any specific application support, even on hardware that doesn't have hardware mixing.
Unfortunately, somebody got really drunk awhile ago and thought PulseAudio was a good idea, and now most distros ship with broken sound configs again. I estimate it'll be a couple years until we're completely past the damage that PA has done, and we'll only be back where we started at best.
To this day I have no idea what problem PulseAudio was trying to solve, but I'm pretty sure it was a net loss.
Monitor options have steadily gotten worse over the last 5 years or so. We were making great progress and then somebody decided that shortscreen (16:10) was the new way to go. Then, not long after that, somebody decided that shorterscreen (16:9) was the new way to go.
Luckily, I bought two 1600x1200 20" monitors right around the time progress stopped and they've remained top of the line ever since. I'd love higher resolution (note - not larger, 2x 20" is perfect, neck would get sore looking at 2x 27" or 2x 30") displays, but they don't exist for under $1000.
Oops, totally misread your summary, I thought you had several terabytes of storage needed rather than 35G. Anyways, the rest still applies, but you can buy a cheaper server with less drives.
Slashdot Mirror
Not all XEONs have hardware virtualization. Only some of the most expensive chips have it and even then, it can be spotty.
Not true. Every Xeon since 2006 has shipped with VT-x support. Look at the Xeon 5030 for example. Absolute bottom of the line ($150 at launch) from 2006, and it supports VT-x.
You're probably thinking of Intel's desktop line where to do artificially hobble large swaths of their CPUs with respect to VT-x.
Yes, that is the disgustingly awful part about C-11, but you missed the upside:
The goverment eventually arrived a trade-off that most Canadians would make: a tougher provision to target sites that facilitate infringement (the law already allows rights holders to do this) in return for a full cap on liability for non-commercial infringement. This applies not only to individuals (likely bringing to an end the prospect of file sharing lawsuits in Canada) but to any non-commercial entity including educational institutions and libraries (who may adopt more aggressive interpretations of the law with less risk of liability).
Emphasis mine, see http://www.michaelgeist.ca/content/view/6544/125/
Drivers, installed base, drivers, familiar windows interface, drivers, most users can barely power their machine on much less install linux, drivers, forget installing linux software...see comment before the last comment, drivers, lack of vendor support, and drivers.
Oh did I mention drivers?
You play weird video games. Personally, I like playing the "my computer works already, I didn't have to hunt down twenty drivers from twenty different sites and make sure I kept them all up to date individually" game, that's why I already use Linux (and have for nearly a decade).
Just wondering: Is there a point (or is this close to it) where in using HDDs and certain RAID configurations, you can match or beat speed while maintaining better redundancy with larger capacity, cheaper drives? What is the main application these excel at? I assume power would be one, and cached content on webservers? Help me understand :-)
You'd need several dozen hard drives to even approach the IOPS of a single consumer level SSD. The SSD wins so many times over it's not funny.
Now, if you're talking about sequential read/write speeds, that's a whole different matter. You'd need roughly 3-4 hard drives (in RAID 0 (no redundancy)... double that figure for RAID 10) to match the typical sequential read/write speeds of an SSD. At that point, the raw cost of the hard drives far exceeds that of the SSD, and that's ignoring the need for the extra SATA ports, cooling, physical space and the extra drive failures you need to deal with. So, the SSD wins again, hands down.
Now, say you needed to store more than roughly 200 gigabytes of data and performance didn't matter at all, in that case, hard drive(s) will be more cost effective than SSDs.
Basically, hard drives excel at bulk storage of stuff where performance doesn't matter. SSDs excel at everything else.
A poorly designed merge section from one highway to another is what convinced me I needed a quick car. It isn't safe merging into 60+ MPH traffic at 30 MPH. Top speed typically isn't a problem but acceleration on cars with wimpy engines is.
If only more people understood this. 99% of the time when I'm behind someone while merging onto the highway, they'll be going way under the speed limit as they merge. It is infuriating.
Support for TRIM on RAID linear/0/1/10 md devices was quite recently added. The patch series is here: https://lkml.org/lkml/2012/3/11/261. I can't find the actual merge now, but I believe it'll be in 3.7.
You're missing the elephant in the room... don't use Windows.
A CS degree is a requisite but not sufficient property to make a good developer. They also need a genuine interest in the field, which most often manifests as being self-taught before getting a degree, and continuing to self-teach after getting said degree.
Purely self-taught developers will miss learning a lot of important topics, not because they're difficult, but because they don't realize what they don't know. In particular, data structures (anything beyond arrays), databases (and normal forms) and algorithmic complexity. I don't care how good you are with $language, if you don't understand the above topics like the back of your hand, you're going to make a mess.
On the flip side, purely academic developers are typically going to have knowledge gaps in more practical topics like input validation, version control systems and bug trackers. Again, you can get by without these, but you're going to make a mess (or someone else is going to make a mess of it for you when they exploit it).
Yup, anybody who is driving an SUV loses the right to ever complain about gas prices again.
Fixed that for you.
When you speed, you put other lives at risk, not just your own.
When you drive the speed limit, you put other lives at risk, not just your own.
I'm in favor, but there is a problem. There are a great many makes and models that CAN NOT maintain 85mph. And many many more than that can't do it safely.
So? Driving is a priviledge, not a right. Should we also give up on driving exams and just let everybody drive?
As an added bonus, you'll stimulate the economy (through car manufacturers).
People refuse to obey the speed limits because the speed limits are retarded. Take any major highway in North America and you'll find massive stretches of more or less completely straight road where there's no reason you couldn't drive all day at your car's top speed, except, the posted speed limit is a third (or less) of said top speed.
This will never change because the government strongly prefers to keep everyone a criminal, they're much easier to control that way. If speed limits were strictly enforced (and not increased to sane values), there would be riots in the streets.
Don't be silly, insurance rates will rise across the board. After all, all these fancy new electronics make the cars more expensive and therefore more expensive to repair in a crash, therefore, everybody's insurance rates need to go up. Furthermore, insurance rates will rise even more for those of us who refuse to enable the tracking data, as obviously only filthy terrorists would value their privacy.
I normally keep the AC at 78-80, kids/wife crank it down to 72 or lower because they felt "hot" for some reason.
Well no wonder, 72F is still too hot.
Especially since I can't rightly say I have a better plan and neither does Mr. deRaadt.
The better plan is to sue Microsoft for abuse of their monopoly.
Perfect, then we can wait a decade for the case to go anywhere, only to have it thrown out in the end and all computers made within the past decade remain unusable.
TFA is talking about in reverse proxies (of which Varnish is one of many), which are very commonplace. In fact, you're seeing this page through (at least) one, as Slashdot uses Varnish.
The importance of this question is not to be underestimated. Anything that adds availability also adds complexity, and if not done correctly will actually worsen your availability. More than once I've seen new "high availability" setups that were less reliable than the original "non-high availability" setup.
If you want to protect against hardware failure, get a server with redundant power supplies and hard drives (can be had for under $2k), and monitor them. Now, redundant power supplies are a no-brainer, but keep in mind that for hardware RAID (both fakeraid and real hardware RAID), when your controller dies all your data is gone unless you have an identical replacement controller. You can go further and add redundant network cards plus redundant network paths (including switches), but again there's a non-trivial complexity cost associated.
If you want availability during updates, then your life just got a whole lot more complicated.
How often do you ever accelerate flat out from 0 to 60?
Way more often than I go over 200 km in one trip. They need to drop the 0-60 time back into the low 4s range like the roadster, and halve the range if need be.
FWIW, the GTX 680 supports up to four displays per (single GPU) card. I haven't tried it yet, but I would expect it to work much better than the 4 monitors across 2 cards support that one was limited to in the past.
Multimonitor across multiple cards has never been stellar in any OS, probably because it's a very niche use case, but multimonitor on a single card is a very common use case.
ALSA's default config since the early 2000s has allowed multiple simultaneous sounds to be played without any specific application support, even on hardware that doesn't have hardware mixing.
Unfortunately, somebody got really drunk awhile ago and thought PulseAudio was a good idea, and now most distros ship with broken sound configs again. I estimate it'll be a couple years until we're completely past the damage that PA has done, and we'll only be back where we started at best.
To this day I have no idea what problem PulseAudio was trying to solve, but I'm pretty sure it was a net loss.
Especially any sort of OpenGL, which doesn't work over the network.
Wrong! OpenGL works over the network, in fact it was designed to.
OT: The register page in your sig is broken:
Taskbars are always set to autohide, so I don't see how this helps.
Monitor options have steadily gotten worse over the last 5 years or so. We were making great progress and then somebody decided that shortscreen (16:10) was the new way to go. Then, not long after that, somebody decided that shorterscreen (16:9) was the new way to go.
Luckily, I bought two 1600x1200 20" monitors right around the time progress stopped and they've remained top of the line ever since. I'd love higher resolution (note - not larger, 2x 20" is perfect, neck would get sore looking at 2x 27" or 2x 30") displays, but they don't exist for under $1000.
Oops, totally misread your summary, I thought you had several terabytes of storage needed rather than 35G. Anyways, the rest still applies, but you can buy a cheaper server with less drives.