Also, while the problem of flash wearing out has been vastly exaggerated, imagine how quickly a contended lock would wear out the 100.000 write cycles. You could easily do that many in a second, and no wear levelling can cope with that.
HP's 1U servers tend to have a couple SATA slots left over, especially if you forego the optical drive (and with PXE or iLO, you don't need it). The actual hard drives tend to run from a SAS RAID controller which often takes up a valuable PCI-E-slot.
Blade servers. They usually have 2 HDD slots at best. The challenge is that they tend to be low on SAS sockets, so you'd need a very small SAS multiplexer as well.
I have a wake up light, and I have supplemented it with one of those bright anti-depression lights on a timer. So when the wake up light hits maximum brightness, the 40W fluorescent comes on -- and THAT works.
Running a complete ISP with peerings/transit and BGP routers and billing and lawful interception is rather complicated. In contrast, running a switched network which can tunnel traffic from a particular port to a particular ISP is relatively cheap and easy. It makes a lot of sense to separate the digging and getting permissions and stringing cable part from the routing and billing part, especially when providers can provide TV and VoIP services on the same network.
I must admit to not RTFA, but usually Coverity provides the results for free to free software projects. I haven't heard of them holding anyone for ransom like that before, so I'm a bit sceptic of your claim.
It isn't astroturfing. No one is pretending to be from the "community" or "grass roots" or anything. It's plain marketing.
Coverity provides free code checks to many free software projects, in exchange for being able to make press releases like this one. The mainline Linux kernel has been through it at least a few times, but Coverity seems a bit confused or unhappy about the fact that Linus won't discuss bugs in secret. Many other large free software projects have a group of people who are willing to sign NDA's when dealing with security bugs, so bugs can be patched before being announced.
Diesel engines are heavy, expensive and already good at handling partial loads. That makes them rather unsuitable for hybrid vehicles. The only reason why locomotives are diesel-electric is that it is reasonably difficult to design a clutch which will survive getting 40 loaded wagons rolling from a stop.
1MW is less than 1 wind turbine. Adding a dedicated 1MW feed really isn't a problem, as long as you don't want to do it to homes.
Yes, there will be the occasional accident. It is impossible to get 100% safety when dealing with the kind of power that cars need. It should be possible to make it no less unsafe than current fuel stations, at least.
You can't economically bring 1MW to each home at this point, and when the car is at home you rarely care whether it charges in 6 minutes or 6 hours. You are probably enjoying that it can charge at home at all, because most people don't have a petrol station in their garage. At home you slow-charge, at the "petrol" station you fast charge. There will be a limited number of fast charge stations, just like there is a limited number of pumps today.
It's unfair that you count Windows 3 times, and the whole point of the Oracle lawsuit is that Android doesn't run standard Java. Besides, does Java really exist for SunOS 4.x? Irix is dead, and you can't get a recent Java for it. So you're at 7, really.
With Java you're tied to a particular JVM, unless you go out of your way to make your code portable. With C you're tied to a particular machine model and operating system unless you go out of your way to make your code portable. The former is slightly better than the other, unless you're on an architecture/OS not supported by the particular JVM required by the application. In that case you need to run the JVM in an emulator, which really really sucks.
Even OpenJDK and Oracle Java are nowhere near compatible, despite the fact that they're pretty much the same code.
If an SSD runs out of write cycles, it goes read only. Hardly the end of the world. It could die in other ways of course, but mechanical hard drives go bang all the time anyway.
SSDs may find a home in specialized situations where the pros outweigh the cons, like laptops, but I doubt they will ever displace magnetic hard drives in most applications.
There is no particular problem with adding multiple fields or a variable length field.
If we're going to roll out updates to every NAT device and host, we might as well roll out IPv6, its already partially rolled-out. Job done!
Except the job is not done. My proposal would be trivial to add to NAT devices, unlike IPv6, which only Apple cares to add to their devices.
I like IPv6, it is a great protocol and the autoconfiguration features are wonderful. However, it is a really large job for an ISP to add IPv6 to an existing network, especially if they provide a bridged IPv4 network with DHCP today. For enterprises, maintaining and securing two separate Internet-connected networks is no fun at all. Approximately 0% of IT staff are trained in IPv6.
All this can be fixed, but can it be fixed in 2 years?
I'm not saying that my solution would have been better. I'm just saying that it could have been done, and that the result would be backwards compatible. So far there is 1 manufacturer of consumer IPv6 CPE's, and hardly any ISP's provide IPv6. I sure hope that changes, but to me it looks like we'll be stuck with Carrier Grade NAT for the next 5 years at least. Possibly a lot longer.
Slashdot Mirror
If you use Linux, you'll be using it soon.
We'll see about that.
If the code quality follows the same curve as Pulseaudio, it will be a long while before I switch to systemd. Even if it means dumping Fedora.
If you're trying to put a lot of local storage into a blade server, You're Doing It Wrong.
Why? What is the advantage of using a larger and usually more power consuming form factor?
For some applications there are no advantages to non-local storage, and the disadvantages can be significant.
If you have a molex connector you can easily split it into as many as you want. Flash drives use negligible power.
Also, while the problem of flash wearing out has been vastly exaggerated, imagine how quickly a contended lock would wear out the 100.000 write cycles. You could easily do that many in a second, and no wear levelling can cope with that.
HP's 1U servers tend to have a couple SATA slots left over, especially if you forego the optical drive (and with PXE or iLO, you don't need it). The actual hard drives tend to run from a SAS RAID controller which often takes up a valuable PCI-E-slot.
Blade servers. They usually have 2 HDD slots at best. The challenge is that they tend to be low on SAS sockets, so you'd need a very small SAS multiplexer as well.
How about HP Solaris?
One of the problems of the lethal injection is that doctors aren't involved. Incompatible with the Hippocratic Oath.
Your ping is going to be horrible, and you'll never be picked as the host.
I have a wake up light, and I have supplemented it with one of those bright anti-depression lights on a timer. So when the wake up light hits maximum brightness, the 40W fluorescent comes on -- and THAT works.
Running a complete ISP with peerings/transit and BGP routers and billing and lawful interception is rather complicated. In contrast, running a switched network which can tunnel traffic from a particular port to a particular ISP is relatively cheap and easy. It makes a lot of sense to separate the digging and getting permissions and stringing cable part from the routing and billing part, especially when providers can provide TV and VoIP services on the same network.
I must admit to not RTFA, but usually Coverity provides the results for free to free software projects. I haven't heard of them holding anyone for ransom like that before, so I'm a bit sceptic of your claim.
Why not? They're selling lots and lots of iPods, why wouldn't they eventually include phone functionality with lower-end iPods?
It isn't astroturfing. No one is pretending to be from the "community" or "grass roots" or anything. It's plain marketing.
Coverity provides free code checks to many free software projects, in exchange for being able to make press releases like this one. The mainline Linux kernel has been through it at least a few times, but Coverity seems a bit confused or unhappy about the fact that Linus won't discuss bugs in secret. Many other large free software projects have a group of people who are willing to sign NDA's when dealing with security bugs, so bugs can be patched before being announced.
Diesel engines are heavy, expensive and already good at handling partial loads. That makes them rather unsuitable for hybrid vehicles. The only reason why locomotives are diesel-electric is that it is reasonably difficult to design a clutch which will survive getting 40 loaded wagons rolling from a stop.
1MW is less than 1 wind turbine. Adding a dedicated 1MW feed really isn't a problem, as long as you don't want to do it to homes.
Yes, there will be the occasional accident. It is impossible to get 100% safety when dealing with the kind of power that cars need. It should be possible to make it no less unsafe than current fuel stations, at least.
Please hand in your geek card. Or at least learn the difference between power (in W) and energy (in J or Wh).
A wind turbine is 2MW+ these days. We hook them up to the grid just fine.
You can't economically bring 1MW to each home at this point, and when the car is at home you rarely care whether it charges in 6 minutes or 6 hours. You are probably enjoying that it can charge at home at all, because most people don't have a petrol station in their garage. At home you slow-charge, at the "petrol" station you fast charge. There will be a limited number of fast charge stations, just like there is a limited number of pumps today.
It's unfair that you count Windows 3 times, and the whole point of the Oracle lawsuit is that Android doesn't run standard Java. Besides, does Java really exist for SunOS 4.x? Irix is dead, and you can't get a recent Java for it. So you're at 7, really.
With Java you're tied to a particular JVM, unless you go out of your way to make your code portable. With C you're tied to a particular machine model and operating system unless you go out of your way to make your code portable. The former is slightly better than the other, unless you're on an architecture/OS not supported by the particular JVM required by the application. In that case you need to run the JVM in an emulator, which really really sucks.
Even OpenJDK and Oracle Java are nowhere near compatible, despite the fact that they're pretty much the same code.
If an SSD runs out of write cycles, it goes read only. Hardly the end of the world. It could die in other ways of course, but mechanical hard drives go bang all the time anyway.
SSDs may find a home in specialized situations where the pros outweigh the cons, like laptops, but I doubt they will ever displace magnetic hard drives in most applications.
Laptops aren't specialized situations. Desktops are.
Copyright maximalism is pretty much the antithesis of free market enthusiasm.
There is no particular problem with adding multiple fields or a variable length field.
If we're going to roll out updates to every NAT device and host, we might as well roll out IPv6, its already partially rolled-out. Job done!
Except the job is not done. My proposal would be trivial to add to NAT devices, unlike IPv6, which only Apple cares to add to their devices.
I like IPv6, it is a great protocol and the autoconfiguration features are wonderful. However, it is a really large job for an ISP to add IPv6 to an existing network, especially if they provide a bridged IPv4 network with DHCP today. For enterprises, maintaining and securing two separate Internet-connected networks is no fun at all. Approximately 0% of IT staff are trained in IPv6.
All this can be fixed, but can it be fixed in 2 years?
I'm not saying that my solution would have been better. I'm just saying that it could have been done, and that the result would be backwards compatible. So far there is 1 manufacturer of consumer IPv6 CPE's, and hardly any ISP's provide IPv6. I sure hope that changes, but to me it looks like we'll be stuck with Carrier Grade NAT for the next 5 years at least. Possibly a lot longer.