And where. exactly, does a script kiddie get a "single 1G connection"? Google Fiber? (and do you really think Google wouldn't notice a flooded connection?) Plus, firing off a DoS from a single location makes it down right trivial to kill -- there's only one machine that has to be unplugged.
Do the phrases "designed by committee" and "political motivation" mean anything to you? The "other distros" are jumping on systemd as a "me too" reaction more than any real technical merit. (generally, the people making the call are not technical people.) GNOME and udev are two of the prime pushes to go with systemd -- as systemd has eaten udev, and GNOME requires logind. systemd does have a few pluses, but it also hauls in a truck load of unnecessary, and complex, bloat.
I don't give a shit what the systemd or upstart camps claim about shell scripts; writing, testing, and debugging init scripts is trivial. (making a single script for *every* distro can be trying, because they all do things a little different. if you think distros aren't going to do the same thing to systemd, think again.)
Instead of actually getting together to make an alternative development stack to systemd
We don't need to. sysvinit WORKS; it's stupid-simple, small, and trivial for anyone with a brain to actually manage. I don't get the whole push for databases, binary logs, 100k+ lines of C code... to do what a few hundred lines of shell scripts and config files have done for decades -- with a shell that's going to be on the system anyway.
(I'm not saying the BSD way of *one* shell script (rc.conf) is the way to go, but systemd is the super-complex bazaaro version of rc.conf.)
Seriously, how often does one reboot linux systems such that the startup is the thing that requires optimization?!? And saying systemd is "a little more complex" is like say the Sahara is a "little dessert".
I thought *NIX purests insisted the entire system be so simple it could be maintained with nothing more than cat and vi. systemd (and SMF) are as far from there as you can get and still be in the solar system.
By definition, it has to be both. A hardware only solution won't do any good when it's no longer in your hands. There has to be some bit of software that will trigger the thermite, as it were, to brick the thing. Erasing the bootloader is a very effective way to do it -- but can still be fixed by skilled hackers. (ever heard of Odin?)
All that's needed is to black list the stolen phones so they can never be used again. Really cuts down the value of a stolen phone there. Even AT&T's GSM phones are of little value outside their US network. Sprint and Verizon phones are useless outside their respective provider's network.
No, it's not "really hard". It simply requires people with a clue to actually take the time to do it correctly. Plugging everything into a stack of unmanaged switches is quick and doesn't require any education at all. Granting you "access to everything" was the quick, expedient solution to a problem that was never later resolved.
Yes, you would have to imagine it. I don't know of any bank that does more than writing to a tape (maybe encrypted) and handing that to Iron Mountain to lose, mangle, and now burn. (yes, I've had those fools lose many of my tapes. About half come back unreadable, and many of those permanently unusable.)
I'd say 90+% is the simple fact that it happens so infrequently no one living there has any clue what so ever on how to drive in snow (and ice.) I (in Raleigh NC) drove home Tuesday night in 1-2" of 19F sand-like snow. I had no problems at all with it -- the few other drivers out were apparently flat-out terrified, even when they were on dry pavement.
(It was nice to have a 4 lane interstate all to myself.)
4min is still an eternity. (plus the time it takes to get it back to the pit, diagnose, get it behind the wall to swap... blow an engine or gearbox and you're race is over.)
By that marketing research, all cars are zero emission on demand (there's an off position)
Yes, I read that... so, run one lap on battery, pull in the pit for an hour while it recharges, then head back out for a lap. Regenerative braking, no matter how aggressive, is only going to give a small fraction of the power back -- and next to none in race conditions where you drive full-on to the last second before applying the maximum brake force for the minimum amount of time. (I could look at our own data, but for the most part, brake application should be under 1 second -- not that we're nearly that perfect. Yes, there will be some recovery, but way more than 90% is going into a brake pad.)
That's my point... NASCAR stock cars aren't stable at any speed. While there's very little of a turn, you *do* have to move the wheel to go around a NASCAR oval. People don't realize just how little input it takes; doesn't matter if you're going 80kph (which is about as slow as you can go) or 300+kph.
Go drive one and tell me how easy it is to control at 180mph+. (actually, they won't let you (us) go that fast, but 140-160mph will certainly change your perspective)
You'll get no disagreement from me. It's boring as shit to watch. That's why I prefer to be the driver. *grin* 24hrs of LeMons wouldn't let us on the oval at Charlotte -- wise choice as some of those crapcans wouldn't be able to stay on it. Chumpcar runs the whole "roval" (infield plus oval.) Plus, there's the lemons moto: any idiot can drive around in a circle. The first race is at Barber this weekend; and that is most definitely not a circle.
F1, you don't have to be smart, but you do have to be able to think and react *very* fast.
NASRCAR is just boring. Those cars can be a handful at the speeds they're going -- and there's no computers doing any driving in there. If there were computers in there, the human really wouldn't be necessary. Executing a pass is the only real task, and that's 99% because of the driver in front doing everything possible to keep you behind him. (blocking is punished in many sanctioned racing organizations because it promotes actively punting people into walls)
Ah... it IS marketing BS... "The automaker claims its ZEOD RC will be the first car to complete an entire race lap of the 8.5-mile Le Mans circuit on nothing but electric power." [link] In other words, the "zero emissions" part is what's "on demand", most of it's time will be spent spewing emissions!
It sounds like that's what it will be doing at race speeds. I've seen what happens to the "hybrid synergy drive" from Toyota when it's used for racing -- the traction belt(s) inside it comes apart. (and then looks like it's been packed with steal wool)
EXACTLY! Takes some serious marketing to call a car with a GASOLINE ENGINE in it, "zero emmision". My hybrid has a flashy sticker saying "ULTRA LOW EMISSION", not ZERO EMISSION.
As en enduro racer, *grin* no, it doesn't have to last 24hrs. It does if you want to FINISH. (also, there's nothing in the rules that prevent an engine replacement during the race. It takes a fair amount of time to swap an engine.) You don't see F1 teams doing it because there's no point; they'd never recover the dozen lost laps. NASCAR has been known to, but they're getting back out to maintain season points. We do it because we wanna race; we're going to be 50+ laps down, but we don't care at that point. (hell, we replaced the transmission at an HPDE once -- my first HPDE, actually. Replaced a head gasket at another.)
Actually, it's not... push new config, test services availability and functionality, revert to previous config if anything isn't correct. It's the exact same thing engineers have been doing for decades: reload in 15min; make changes; if your changes foobar things, get a cup of joe and wait for the reload. (for some carrier grade systems (nortel), it's automatic; if you don't commit your changes, they automatically revert after some time, I forget the period.)
We aren't talking redneck installs here. We're talking proper, to code, in-building installation by (presumed) professionals. If your fiber is in a conduit with power, you've already failed. If your fiber is rubbing against *high voltage lines* you've failed big time -- or something else has failed as well (think: downed power line.) The tracer isn't a big wire, and it doesn't have to be continuous -- no signals cross it, it's only there for a metal detector to detect. (a BB every few inches would do the same thing.) Any tracer / armor doesn't go all the way into the ONT; it's not bonded to the ONT. For the NYC cases, the in-building cable doesn't have tracers or armor, so there's nothing at all to conduct any "dangerous nail induced high voltage." The installation of an ONT doesn't add any additional risk to the pre-existing coax and/or phone wiring.
Slashdot Mirror
And where. exactly, does a script kiddie get a "single 1G connection"? Google Fiber? (and do you really think Google wouldn't notice a flooded connection?) Plus, firing off a DoS from a single location makes it down right trivial to kill -- there's only one machine that has to be unplugged.
Binary log files are Evil(tm). A single stray byte out of place and the log is destroyed. (never used checkpoint firewall, or innoDB I guess)
Do the phrases "designed by committee" and "political motivation" mean anything to you? The "other distros" are jumping on systemd as a "me too" reaction more than any real technical merit. (generally, the people making the call are not technical people.) GNOME and udev are two of the prime pushes to go with systemd -- as systemd has eaten udev, and GNOME requires logind. systemd does have a few pluses, but it also hauls in a truck load of unnecessary, and complex, bloat.
I don't give a shit what the systemd or upstart camps claim about shell scripts; writing, testing, and debugging init scripts is trivial. (making a single script for *every* distro can be trying, because they all do things a little different. if you think distros aren't going to do the same thing to systemd, think again.)
We don't need to. sysvinit WORKS; it's stupid-simple, small, and trivial for anyone with a brain to actually manage. I don't get the whole push for databases, binary logs, 100k+ lines of C code... to do what a few hundred lines of shell scripts and config files have done for decades -- with a shell that's going to be on the system anyway.
(I'm not saying the BSD way of *one* shell script (rc.conf) is the way to go, but systemd is the super-complex bazaaro version of rc.conf.)
No, it's a crap-ton of binaries that depend on another crap-ton of bullshit libraries. (anything that requires D-Bus, is, by definition, a mistake)
What the fuck is wrong with shell scripts and shell fragment configuration files? It's worked for many decades.
Seriously, how often does one reboot linux systems such that the startup is the thing that requires optimization?!? And saying systemd is "a little more complex" is like say the Sahara is a "little dessert".
I thought *NIX purests insisted the entire system be so simple it could be maintained with nothing more than cat and vi. systemd (and SMF) are as far from there as you can get and still be in the solar system.
By definition, it has to be both. A hardware only solution won't do any good when it's no longer in your hands. There has to be some bit of software that will trigger the thermite, as it were, to brick the thing. Erasing the bootloader is a very effective way to do it -- but can still be fixed by skilled hackers. (ever heard of Odin?)
All that's needed is to black list the stolen phones so they can never be used again. Really cuts down the value of a stolen phone there. Even AT&T's GSM phones are of little value outside their US network. Sprint and Verizon phones are useless outside their respective provider's network.
No, it's not "really hard". It simply requires people with a clue to actually take the time to do it correctly. Plugging everything into a stack of unmanaged switches is quick and doesn't require any education at all. Granting you "access to everything" was the quick, expedient solution to a problem that was never later resolved.
Yes, you would have to imagine it. I don't know of any bank that does more than writing to a tape (maybe encrypted) and handing that to Iron Mountain to lose, mangle, and now burn. (yes, I've had those fools lose many of my tapes. About half come back unreadable, and many of those permanently unusable.)
I'd say 90+% is the simple fact that it happens so infrequently no one living there has any clue what so ever on how to drive in snow (and ice.) I (in Raleigh NC) drove home Tuesday night in 1-2" of 19F sand-like snow. I had no problems at all with it -- the few other drivers out were apparently flat-out terrified, even when they were on dry pavement.
(It was nice to have a 4 lane interstate all to myself.)
4min is still an eternity. (plus the time it takes to get it back to the pit, diagnose, get it behind the wall to swap... blow an engine or gearbox and you're race is over.)
By that marketing research, all cars are zero emission on demand (there's an off position)
Yes, I read that... so, run one lap on battery, pull in the pit for an hour while it recharges, then head back out for a lap. Regenerative braking, no matter how aggressive, is only going to give a small fraction of the power back -- and next to none in race conditions where you drive full-on to the last second before applying the maximum brake force for the minimum amount of time. (I could look at our own data, but for the most part, brake application should be under 1 second -- not that we're nearly that perfect. Yes, there will be some recovery, but way more than 90% is going into a brake pad.)
That's my point... NASCAR stock cars aren't stable at any speed. While there's very little of a turn, you *do* have to move the wheel to go around a NASCAR oval. People don't realize just how little input it takes; doesn't matter if you're going 80kph (which is about as slow as you can go) or 300+kph.
Go drive one and tell me how easy it is to control at 180mph+. (actually, they won't let you (us) go that fast, but 140-160mph will certainly change your perspective)
You'll get no disagreement from me. It's boring as shit to watch. That's why I prefer to be the driver. *grin* 24hrs of LeMons wouldn't let us on the oval at Charlotte -- wise choice as some of those crapcans wouldn't be able to stay on it. Chumpcar runs the whole "roval" (infield plus oval.) Plus, there's the lemons moto: any idiot can drive around in a circle. The first race is at Barber this weekend; and that is most definitely not a circle.
F1, you don't have to be smart, but you do have to be able to think and react *very* fast.
NASRCAR is just boring. Those cars can be a handful at the speeds they're going -- and there's no computers doing any driving in there. If there were computers in there, the human really wouldn't be necessary. Executing a pass is the only real task, and that's 99% because of the driver in front doing everything possible to keep you behind him. (blocking is punished in many sanctioned racing organizations because it promotes actively punting people into walls)
Ah... it IS marketing BS... "The automaker claims its ZEOD RC will be the first car to complete an entire race lap of the 8.5-mile Le Mans circuit on nothing but electric power." [link] In other words, the "zero emissions" part is what's "on demand", most of it's time will be spent spewing emissions!
That's what Z and E stand for in the name... ZEOD RC (Zero Emission On Demand Racing Car)
And what's the torque like? Indeally, the engine powers only a generator, so torque matters.
On a side note, someone put a harley engine in a prius for a lemons race a while back. (I'm not sure how well that turned out.)
It sounds like that's what it will be doing at race speeds. I've seen what happens to the "hybrid synergy drive" from Toyota when it's used for racing -- the traction belt(s) inside it comes apart. (and then looks like it's been packed with steal wool)
Maybe, but 400hp is nothing to F1 and NASCAR. And both series have almost molecular requirements for their engines.
EXACTLY! Takes some serious marketing to call a car with a GASOLINE ENGINE in it, "zero emmision". My hybrid has a flashy sticker saying "ULTRA LOW EMISSION", not ZERO EMISSION.
As en enduro racer, *grin* no, it doesn't have to last 24hrs. It does if you want to FINISH. (also, there's nothing in the rules that prevent an engine replacement during the race. It takes a fair amount of time to swap an engine.) You don't see F1 teams doing it because there's no point; they'd never recover the dozen lost laps. NASCAR has been known to, but they're getting back out to maintain season points. We do it because we wanna race; we're going to be 50+ laps down, but we don't care at that point. (hell, we replaced the transmission at an HPDE once -- my first HPDE, actually. Replaced a head gasket at another.)
Actually, it's not... push new config, test services availability and functionality, revert to previous config if anything isn't correct. It's the exact same thing engineers have been doing for decades: reload in 15min; make changes; if your changes foobar things, get a cup of joe and wait for the reload. (for some carrier grade systems (nortel), it's automatic; if you don't commit your changes, they automatically revert after some time, I forget the period.)
Hah. We are the Google Test Environment(TM)
We aren't talking redneck installs here. We're talking proper, to code, in-building installation by (presumed) professionals. If your fiber is in a conduit with power, you've already failed. If your fiber is rubbing against *high voltage lines* you've failed big time -- or something else has failed as well (think: downed power line.) The tracer isn't a big wire, and it doesn't have to be continuous -- no signals cross it, it's only there for a metal detector to detect. (a BB every few inches would do the same thing.) Any tracer / armor doesn't go all the way into the ONT; it's not bonded to the ONT. For the NYC cases, the in-building cable doesn't have tracers or armor, so there's nothing at all to conduct any "dangerous nail induced high voltage." The installation of an ONT doesn't add any additional risk to the pre-existing coax and/or phone wiring.