But how many Yucca Mountains will there be when that time finally comes?
If corporations won't pony up for this research in the present (and god knows most corporations plan only as far as the next shareholder presentation), then someone should. And if you don't think this impetus should come from the public sector, well, perhaps you should remember where the Internet came from - just one example of a multitude of technologies that have been borne out of public sector research.
Oh, believe me, of course I understand what science is, and I'm all for the independant and stringent validation of assumptions and even common knowledge. The point is, it's not really news, especially not on slashdot, is it?
You can definitely hear the difference when a pigeon takes off due to being startled, and when it launches itself normally. I always considered the nosie to be from coming from its wings, and as a signal conveying danger. Somehow this news seems like it isn't.
Yeah, I get a decent amount of bandwidth for a very reasonable price, but there's no way in hell I could actually *use* that bandwidth due to the lack of speed. I think the networks are just way oversubscribed, so the contention ratio is insane. Making a new outbound connection sometimes takes *minutes*, and once the connection is made I get maybe 20kb/s if I'm lucky.
Hilariously, 3G/HSDPA is way quicker and more usable in South Africa, a country infamous for its otherwise completely crap Internet infrastructure.
If Hutchinson/Three can get away with calling *this* connection "Mobile Broadband", you can get away with fucking anything. I speak as an ex-dial-up user of many years, this is FAR worse than dial-up.
What a load of crap, anyway. What next, are they gonna call the abacus a microcomputer? These terms should *never* regress.
Yeah, the thing with EA (speaking as an ex-EA employee) is they penalise any projects not conforming to their restrictive and short-sighted specifications (eg, game behind schedule, not properly incorporating their input, etc) by withdrawing spending on advertising. So either the studio / developers suffer, or the game quality suffers. However you look at it, the gamer suffers.
And this is why asserts should *never* go into production builds of any project. It's fine to have asserts in your debug build, but ALWAYS deal with the unexpected case immediately after your assert (which should be compiled out in release mode).
If you have no way of throwing an error and handling it gracefully back up your call stack (no, you don't always need exceptions for this), then you've done a shit job!
Compared to the current Chrome 2, Firefox 3.5 with JIT enabled gets 1/2 the speed here, 7/8th the speed here, but about 2x the speed here. That's a much better result than ff3.1!
It's something I've done on company time, but my director has expressed interest in opensourcing some of this stuff (he likes the 'many eyes' idea), so in the near future we'll make it available somehow, when I'm sure it's properly threadsafe (lockless!) and works in a few other environments. Don't think that helps you right now though, sorry!
That said, the real magic comes from the packet mmap API with Johann Baudy's TX Ring patch. The stuff I've been doing is encapsulating this down-to-the-metal interface with something approximating really simple sockets, with all the other support stuff (ARP, device and OS IP route enumeration, etc) that you need in order to use it as a general IP-over-Ethernet stack.
Because our situation involves high pps I've spent some time keeping the per-packet stuff simple and therefore quick - eg, on 1x2.5Ghz core: 180m ARP lookups/sec, 200m route lookups/sec.
I'm sure standard BSD sockets would perform in the same ballpark if they didn't have such broad application!
No, I'm only talking about UDP - specifically, moving and routing large numbers of media streams. I really hope you don't think that I have no idea what a SOCK_DGRAM is:-)
The can of worms that is TCP is not something I have the energy or need (thankfully) to re-implement!
Well it wasn't exactly an overnight hack, heh! I would really be a lot more comfortable if something similar, widespread, and well tested was in existence. It's kinda scary ripping the networking code out of my other more mature projects which are in full production in order to test this.
Slashdot Mirror
But how many Yucca Mountains will there be when that time finally comes?
If corporations won't pony up for this research in the present (and god knows most corporations plan only as far as the next shareholder presentation), then someone should. And if you don't think this impetus should come from the public sector, well, perhaps you should remember where the Internet came from - just one example of a multitude of technologies that have been borne out of public sector research.
Seconded.
ATTENTION WORLD GOVERNMENTS:
Fund. Fucking. Thorium. Fuel. Cycle. Research.
PLEASE.
So he's saying he doesn't like programmers who enjoy what they do? Interesting.
I've never known a *good* programmer who doesn't write code as a hobby.
Is there a version edited for rednecks?
Oh, believe me, of course I understand what science is, and I'm all for the independant and stringent validation of assumptions and even common knowledge. The point is, it's not really news, especially not on slashdot, is it?
You can definitely hear the difference when a pigeon takes off due to being startled, and when it launches itself normally. I always considered the nosie to be from coming from its wings, and as a signal conveying danger. Somehow this news seems like it isn't.
Yeah, I get a decent amount of bandwidth for a very reasonable price, but there's no way in hell I could actually *use* that bandwidth due to the lack of speed. I think the networks are just way oversubscribed, so the contention ratio is insane. Making a new outbound connection sometimes takes *minutes*, and once the connection is made I get maybe 20kb/s if I'm lucky.
Hilariously, 3G/HSDPA is way quicker and more usable in South Africa, a country infamous for its otherwise completely crap Internet infrastructure.
If Hutchinson/Three can get away with calling *this* connection "Mobile Broadband", you can get away with fucking anything. I speak as an ex-dial-up user of many years, this is FAR worse than dial-up.
What a load of crap, anyway. What next, are they gonna call the abacus a microcomputer? These terms should *never* regress.
Yeah, the thing with EA (speaking as an ex-EA employee) is they penalise any projects not conforming to their restrictive and short-sighted specifications (eg, game behind schedule, not properly incorporating their input, etc) by withdrawing spending on advertising. So either the studio / developers suffer, or the game quality suffers. However you look at it, the gamer suffers.
Can anyone make a quick summary for me?
I did mean charge! Trying to post comments while in the middle of a conversation at a party - probably not the best idea :)
A Coulomb is a heck of a lot of current, and a bolt of a lightning happens in a heck of a short time. The number's about right for a *large* bolt.
Down with this sort of thing!
Up with this we will not put!
And this is why asserts should *never* go into production builds of any project. It's fine to have asserts in your debug build, but ALWAYS deal with the unexpected case immediately after your assert (which should be compiled out in release mode).
If you have no way of throwing an error and handling it gracefully back up your call stack (no, you don't always need exceptions for this), then you've done a shit job!
You ASCII stupid question, you get a stupid ANSI.
Yeah, I meant them 3.1 beta release jobbies.
Compared to the current Chrome 2, Firefox 3.5 with JIT enabled gets 1/2 the speed here, 7/8th the speed here, but about 2x the speed here. That's a much better result than ff3.1!
Well done, guys.
Just a heads up, you've truncated pi instead of rounding it off — the next digit is 6. No, I didn't just look that up :(
Whoops, you weren't talking to me, sorry!
It's something I've done on company time, but my director has expressed interest in opensourcing some of this stuff (he likes the 'many eyes' idea), so in the near future we'll make it available somehow, when I'm sure it's properly threadsafe (lockless!) and works in a few other environments. Don't think that helps you right now though, sorry!
That said, the real magic comes from the packet mmap API with Johann Baudy's TX Ring patch. The stuff I've been doing is encapsulating this down-to-the-metal interface with something approximating really simple sockets, with all the other support stuff (ARP, device and OS IP route enumeration, etc) that you need in order to use it as a general IP-over-Ethernet stack.
Because our situation involves high pps I've spent some time keeping the per-packet stuff simple and therefore quick - eg, on 1x2.5Ghz core: 180m ARP lookups/sec, 200m route lookups/sec.
I'm sure standard BSD sockets would perform in the same ballpark if they didn't have such broad application!
No, I'm only talking about UDP - specifically, moving and routing large numbers of media streams. I really hope you don't think that I have no idea what a SOCK_DGRAM is :-)
The can of worms that is TCP is not something I have the energy or need (thankfully) to re-implement!
Well it wasn't exactly an overnight hack, heh! I would really be a lot more comfortable if something similar, widespread, and well tested was in existence. It's kinda scary ripping the networking code out of my other more mature projects which are in full production in order to test this.
Stupid thing posted me anonymously despite being logged in!
and two for the goo.
Finnish is also very similar in structure and sound to English
Hahahahaha! HAHAHAHAHA!
Consider that a correction, vittu :)