Re: Oath Hell too please
I'm not sure how they would store your preference if they can't store cookies or include any sort of personal identifier.
889 posts • joined 13 Aug 2007
I'm not sure how they would store your preference if they can't store cookies or include any sort of personal identifier.
The business problem here is companies will pay extra to micro-target ads to, say, male 28-35 year olds with college degrees who make between $80,000 and $120,000/year and own a dog. They believe, often without evidence, that this will result in more effective ads than just scattershot context advertising.
Public domain is best regarded as a myth. Under modern copyright laws it doesn't appear to be possible to actually sign away all your rights. There are licenses that accomplish basically the same thing, but just declaring something "public domain" doesn't necessarily make it so.
I've seen those proposals too, but I'd argue that building a chemical plant that will start up and run without problems, without human intervention, on another planet, is probably even riskier and more complicated than orbital rendezvous. Even when built on earth, chemical processing plants are finicky things.
phuzz covered this well. The basic issue here is that the fuel requirement to launch a payload to orbit scales with the cube of the payload weight. With an out-and-return mission, you have to do this twice -- the fuel for the return mission is part of the launch mission's payload, so you're paying double for your return payload. (This is why the LEM had to be so lightweight.) Mars also has a relatively deep gravity well; not as deep as Earth's, but a lot deeper than the Moon's. The tyranny of the rocket equation strikes again.
This ignores the fuel requirements to get into interplanetary transfer orbits, but IIRC those are actually a fair bit smaller than what it takes to get to orbit to start with. This also doesn't include fuel needed to make a soft landing, but since Mars and Earth both have atmospheres you can use for aerobraking, that's not a large requirement. If you're clever and have a robust enough heat shield, you might even be able to slow down for orbital capture that way. But whatever you're using for your landing on Earth, you've got to lug it both ways, unless you're going to pick your sample up from orbit with another craft.
One way around some of this is to do what Apollo did, and leave some of your fuel (and maybe your Earth re-entry gear) in orbit while a smaller sampling craft descends to the surface. That helps with the fuel requirements, since you're not carrying all that fuel back up to orbit again, but it greatly increases your mission complexity; you'll need to do an automated orbital rendezvous and capture, on limited fuel reserves, and the round-trip communications time is too large for a manual override if it goes wrong. This is not easy.
The basic problem, IMHO, is routers that aren't validating that the IP address they're being configured to forward to is actually an internal IP address. It's the same basic idea as the old FTP "PORT" command trick.
There's also no reason to let anything on the Internet side of the connection configure UPnP settings.
This isn't a UPnP problem, it's a stupid router design problem.
Skype hasn't been peer-to-peer in years. Coincidentally, this change happened right around the time that the US government was leaning on them to make Skype conversations easier to eavesdrop on.
This sounds like my reaction when I found out the Nintendo Switch had been cracked with a stack-smashing buffer overflow attack on its GPU ROMs. Via a memcopy routine that accepted a length parameter from untrusted input, no less. Maybe current firmware coders have just forgotten what the software industry learned 30 years ago.
A lot of the rationale for going with Windows was that it would save the military money, since they'd be using more off-the-shelf stuff instead of custom. It hasn't necessarily worked out that way, but if they dump Windows I suspect they'll go back to their own bespoke OS's instead of moving to OpenBSD or something like that.
I wouldn't see it as a high priority if I were them, either, unless there's some kind of code execution vulnerability involved. If you've got physical access to the machine you have lots of other ways of crashing it.
Did Google give up the data?
"Google did not confirm or deny whether it handed over the requested data to police."
If they refused, they'd be talking about it as a PR move. The most likely possibility is they did hand it over, and the court order included a gag clause preventing them from talking about it.
The order was a doozy, by the way -- everyone within a 17 acre area surrounding a murder scene, including businesses and residences. That's a pretty large dragnet.
While that may be true, Apple isn't an ad broker. Remember, Google's main business is selling ads. All the other stuff just lets them collect data to target the ads better.
The thing that made me stop and think was an incident in North Carolina where the police got a warrant requiring Google to give them a list of everyone who was near the scene of a particular crime. That's getting a little scary. I turned off location history after that, although I remain unsure if that's enough.
I wonder how this plays in places that require consent from all parties for a conversation to be recorded? There are a few such "two-party" states in the US. I may have consented to have myself recorded, in a click-through agreement, but the guy chatting with me hasn't.
I played with the PC version back in the day. It had some neat ideas, but was half-baked at best, and more mainstream OS's pretty quickly overtook it. I think it was the first time I encountered software audio mixing. (I highlighted a bunch of MP3s and clicked 'Play', expecting them to queue, but instead it played them all simultaneously. "Do the thing that's impressive instead of the thing a sane person would want" seemed to be a BeOS motto.)
I love hearing about other people's Dwarf Fortress games, but after looking at the user interface I think I'd rather do something more simple and intuitive -- like program an implementation of Eliza in TECO.
Once upon a time, before we had GPS, I worked briefly for a company developing and deploying a navigation/tracking system. Using not satellites, but land-based beacons, sending out the signals from which to compute a position.
This sounds suspiciously like LORAN.
While technically the US could switch off GPS, it would be so disruptive to our own systems that it's hard for me to imagine a situation where it would happen. Presumably any adversary would simply use devices capable of using other satnav systems. It's not like multi-system setups are hard to come by; my cheap smartphone can use at least three.
As for degrading the civilian signal, they gave up on Selective Availability in 2000, and the latest satellites reportedly no longer have the hardware for it. Again, it's hard to see what could be gained by turning it on at this point.
I think the GPS III block satellites were estimated at $500 million each, not including launch costs. (They're expected to go up on Falcon 9's for something like $90 million each, starting in the next year or two.) The 1990s ones were about 10% of that, so if you're willing to not have the latest and greatest tech you could probably build one cheaper. You'll need about 30 of them, plus spares, so try to negotiate a bulk discount. Mind the replacement costs; estimated lifespan is 7.5 years for current builds, and the record so far is 10, so you'll have to pony up cash for new ones on a regular basis. Also keep in mind you can't just throw them up there and forget about them, you'll need a ground station network to keep tabs on them, do software updates, and make adjustments as necessary.
...your firewall config just expanded from securing ONE IP to an entire subnet on a protocol you aren't familiar with.
That doesn't actually make your firewall config more complex if the default is "block all incoming," which is what you're arguing we should use NAT to do anyway. (This is assuming a bridge-style "transparent" firewall, but those are common even on IPv4 networks at this point.)
...everyone is using NAT and there are no inherent problems with a proven technology that serves a practical purpose.
If you think NAT isn't broken, it's because you're used to the brokenness. I started using it when it was called "IP masquerading" and was an experimental Linux kernel module. It's always been hacky and buggy, people now just think the breakage is normal.
Besides the problem I noted earlier, there are others:
- Double NAT. Right now most traffic only has to traverse one layer of NAT, at the home router. It does usually work OK if you only go through one layer of it. Try to go through two -- say, you're using a mobile hotspot (NAT'd on the phone) on a mobile carrier that's also using NAT -- and things start to break. FTP simply stops working, for example, even in passive mode. As time goes on we'll be seeing more and more levels of NAT applied and more and more protocols will fall apart.
- Peer-to-peer protocols that work across the Internet but fail if you try to use them with someone on your LAN, because the IP addresses don't match up. I've played some online games that I could play with literally everyone in the world except my own family.
- Idle TCP/IP connection timeouts. After 5-10 minutes of silence, a lot of home routers will decide a TCP/IP connection is no longer needed and drop it to free up space in the NAT table. This is why SSH sessions over home routers tend to drop if you step away for a few minutes. This has resulted in a lot of hacky keepalive systems that send useless data every minute or so just to keep the channel open.
- The security disaster that is UPnP, which exists mostly to give NAT'd devices an automated way to request port forwarding.
Anyway, the fact is that IPv6 *does* support a form of NAT. It's called Network Prefix Translation. But it's not commonly implemented because it doesn't actually serve a useful purpose.
You can always give them arbitrary local IPv6 addresses.
fd00::1 is the router, fd00::2 is the printer, fd00::cafe is the coffee machine, etc. ;)
These aren't routable but there's no assumption that a device will only have one address, in IPv6.
That nonsense literally held back adoption, because who the hell wants to go through every switch, router, server, client, phone, printer, etc. and give them all IPv6 addresses and then address them only by that?
Eh? I don't understand what the complaint is here.
Yeah, you plug in the printer and your router will automatically give it a global IPv6 address. Your router will also have a firewall in it, probably on by default, so there's no real security concern here.
If you don't like that global address, that's fine, because IPv6 allows multiple addresses per interface, and the fe80:: block is set aside for local use; you can set up all the private, non-routable IPv6 addresses you want there, and in the fd00:: block. Actually, given modern discovery protocols like Bonjour, your printer will probably do this automatically and broadcast its existence to your computer, so you never actually have to enter the address manually.
The only real difference here is you don't have to go through a broken NAT layer that has to keep track of every single connection, and guess when they're idle in order to clear out that memory table. The limitations of this become readily apparent when you launch BitTorrent on a NAT'd machine and everyone else's SSH connections drop because the NAT table filed up.
reading out and IVP6 address over the phone would leed to errors.
Not necessarily, especially if the addresses are well chosen. Here's a couple OpenDNS server addresses:
It's not obvious to me that the second one is harder to get right than the first. I'll grant that it's less familiar, though.
I *will* admit that it's a lot harder to figure out IPv6 than the old reliable four octet address but maybe I am showing my age.
I dunno, I find smaller CIDR networks kind of a pain to keep track of. I mean, yeah, if you're still doing 1993-style classful addressing IPv4 is easy, but I sure can't rattle off the network and broadcast addresses for an arbitrary /27.
Also, VERY large enterprises may find it simplifies things.
I've heard rumors that Comcast's fairly early switch was because they'd gotten so big, the private IPv4 network they were using to administratively address cable modems had gotten unwieldy.
A lot of the horror stories out there are from people who tried it a few years back, when things were much less well sorted. When your ISP supports it, it's not too bad. At my current house my cable provider turned it on and it took me about a week to notice something had changed.
Supporting dual-stack servers is slightly trickier. A lot depends on how old your software is. Recently maintained software like Apache will happily dual-stack right out of the box. I run a couple MUSHes that have 1990s-era code bases and they were a little trickier. One of them I was able to coax into accepting IPv6 connections over a shared 4/6 socket. The other one I ended up using socat to accept IPv6 connections and relay them via IPv4, which is ugly, but effective.
Who on earth thought that 128-bit addresses was a good idea?
The idea of 128-bit addresses is you can afford to waste some of them in order to simplify routing. IPv6 addressing breaks down on some very specific barriers. There's no "wait, what's the broadcast address for a /27, again?" mental math.
The biggest thing to remember is that the smallest subnet is a /64. The last 64 bits are yours to do what you want with, so you can make them something memorable if you like. So that's a maximum of 16 arbitrary hex digits you might have to remember. But! You also get to omit leading zeroes, and collapse 16-bit blocks that are all 0. So it's not even that bad.
To give a real-world example, Comcast's DNS servers are 2001:558:feed::1 and 2001:558:feed::2. Those don't strike me as harder to memorize than most arbitrary IPv4 addresses. (OK, they're harder than 184.108.40.206, but that's something of a special case.)
IPX was never an Internet protocol, just a LAN one. The actual transition was from ARPANET's NCP to TCP/IP. Since there were relatively few nodes at the time, the Defense Communications Agency simply set a mandatory "flag day" when NCP would stop being routed. Even that wasn't a total success; it was initially supposed to be January 1, 1983, but they had to give a three month extension because not everyone had successfully written TCP/IP stacks by then.
One of the universities I worked for transitioned to IPv6. There were some hiccups at first but once the routing was sorted out, I set up some dual-stack servers without trouble. (I believe the Internet2 academic research network runs IPv6.) In some ways universities have fewer incentives than most, because many of them got huge allocations.
The biggest problem isn't that it's difficult, it's that there's still a lot of old equipment rattling around that either doesn't support it at all, or doesn't perform well enough. I've seen places gradually phase it in as they replaced crusty old routers with new kit.
I've had two cable ISPs that supported IPv6, now. Every home router I've used that had IPv6 had a firewall turned on by default. My cell phone provider also uses IPv6, BTW, with IPv4 support behind carrier NAT.
NAT was never supposed to be a security barrier; that it functions as one is mostly by accident, because it happens to require a firewall.
I've had no problems except for once early on when Comcast turned up IPv6 locally before they were ready to route it. A substantial fraction of my web and video streaming traffic goes over IPv6 now, and I can't say I notice a difference. And that may be part of the problem; for the consumer this isn't really a value-added proposition.
...did you ignore the fact that the US is heavily in debt to China because it has lend it hundreds of millions?
While that's true, it also isn't as much leverage as you might think. These are treasury bonds, not bank loans that they can suddenly call due. About all they could do is have a fire sale on their holdings; while that would hurt the US indirectly -- the drop in price would force a rise in interest rates -- it would also hurt China, because they'd be selling into a falling market. In other words they'd probably lose tens of millions of dollars, and in exchange the US might see an economic contraction in six months to a year.
China's investments in US debt are probably just that, investments used to park some money in a stable currency. As a way to exert pressure on the US they're really not all that effective.
I don't think his grasp of these issues is that sophisticated. If his grasp of economics is tenuous enough that he thinks a trade deficit means we're getting cheated, how's he going to feel about the phrase "weak dollar"? In his mind's eye it'd be a scrawny weakling being bullied by other currencies.
ZTE's phones are actually remarkably cruft-free, from what I've seen. They add a few things but not nearly as much as Sony or Samsung cram in there.
I think when accused of blatantly and obviously violating safety rules, if you're really innocent the correct response would be "have a look at the yellow paint on our floors, see?" At least, if you're actually following the rules. Tesla's prickly response suggests they're guilty and stalling for time. Elon is part of a generation of corporate CEOs that think that laws don't apply to them (see also Uber).
I think the visual merging aspect is mostly accident -- it would be speed dependent after all, and the width is pretty much set by the width of the locomotive -- but you have the basic idea right. "Ditch lights" were added to make trains easier to distinguish when approaching grade crossings. A single headlight gives no good reference for position or distance and can, at a glance, be easily mistaken for a distant house light or streetlight. In the US ditch lights are actually required for going more than 20 mph through grade crossings, and usually go into a flashing mode when the horn is sounded; they replaced the older "Mars lights" as an attention-getting device.
This discussion reminds me of a time I was driving down a section of I-5 in southern Washington, where the BNSF mainline runs down the freeway median. It was the middle of the night and I was complaining loudly about "that idiot who has their high-beams on" when I realized that idiot was driving an oncoming locomotive.
As for the F35. I find it amazing that it has taken a huge chunk of my life to get the damn thing to this point. In the mean time, we've had two complete generations of commercial aircraft designed, developed, built, certified and put in service.
There have been no major innovations in commercial air travel since the Concorde. The mission requirements are all well understood, and not subject to major changes. It's been about building the same airborne buses over and over, with a tweak for more efficiency here or to squeeze in a few more seats there. One of the most popular airliners, the 737, was initially designed in 1967 and is still being made.
The F35, on the other hand, was designed to be "one jet to rule them all" so its mission requirements are complex and ever-shifting. It was trying to do a lot of relatively new things. A lot of its problems have been software-related, which isn't surprising because requirements for it to integrate with other systems are massively complex.
What you're seeing is a combination of baroque government procurement and bidding rules (enacted incrementally over the years to try to prevent Fraud and Abuse™), distributing a project out over as many congressional districts as possible to make it impossible to cancel, and an aircraft that's trying to do an unprecedented number of things at once -- probably too many.
Corporatism is what youre thinking of.
Corporatism and capitalism are synonyms, or at least as alike as to make no difference. Every free market reform we get here in the US leads to more corporate control and higher wealth concentration. The nature of the free market game is that once you start to win, you can rig things so you keep winning.
Launches are exactly the kind of small, capital-intensive market that naturally creates monopolies and duopolies. I expect the current situation of multiple providers won't last long; the winners will buy out the also-rans. Eventually it'll just be a merged Boeing-SpaceX or something like that.
Many of the Shuttle's failings weren't because of NASA requirements, they were because of the Air Force, which asked for some very specific capabilities that drove up weight. One example is the Air Force wanted the Shuttle to be able to take off to put a classified satellite in polar orbit, abort, and land in the US after only a single orbit. This required something like 1000 miles of cross-range capability, which mandated a much larger wing. (Normally we think of the Shuttle's wing as holding it up against gravity, but with a steep bank angle it could also be used to translate it sideways by large amounts.)
I wouldn't say risk aversion as such was the issue. The Shuttle was highly redundant, true, but mostly not to a greater extent than, say, a modern fly-by-wire airliner. Its design of having three flight computers, so that two good ones could outvote one bad one, is still a common technique for critical systems. To quote Akin's Laws, "To design a spacecraft right takes an infinite amount of effort. This is why it's a good idea to design them to operate when some things are wrong."
I do think the Shuttle demonstrated that reusability isn't necessarily a killer idea. At some point making the components reusable adds enough weight and complexity that it's actually cheaper to throw them out and make new ones, as counterintuitive as that seems. The tyranny of the rocket equation means that you pay dearly for every extra pound on your payload.
WHOIS contact details are a relic of the 1990s, when everyone ran fingerd and identd and sysadmins could solve problems by calling each other up for a friendly chat.
Nowadays putting that kind of personal information publicly on the Internet is just asking to be scammed and abused. Last time I published my real phone number in a WHOIS record, I got about a dozen telemarketing and scammer calls a day for weeks. My phone is still largely unusable for incoming calls (because I just ignore it.) At this point it's just a way for registrars to make more money by selling "domain privacy" packages.
The odds of Spectre causing a major security problem for a gaming rig are probably low. A far more likely scenario is an accidental backdoor in one of the games you play, or an intentional backdoor in a sketchy mod you install. If you want to be careful, do your banking on another system.
It's a non-obvious vulnerability that comes about because of fundamental features of how the chips work.
So I'd say it's like suing a car company over carjackings, because they made cars that had to stop at traffic lights.
The phrase "until the sound stack broke again" made me chuckle, because I don't think I've had reliable sound in Linux since they abandoned OSS for ALSA. I recently replaced my Ubuntu media PC setup with Windows 10 because one too many evenings where I wanted to watch a movie ended with me trying to figure out why ALSA's HDMI sound output mysteriously stopped working again.
While it's dead now, the COM executable format in MS-DOS existed to make porting CP/M software easier, and persisted well into the Windows era.
Some vestigial chunks of the IBM XT BIOS persisted into the 90s; I remember accidentally powering up an early 386 with no disks attached, and seeing "NO ROM BASIC" in glorious 40-column text.
Reminds me of the plight of gun sellers in the US. During the Obama administration they sold tons and tons of guns and ammo based on fears that the government was going to make it all illegal. Now that there's a Republican president, no one is buying, leaving them with unsold inventory and excess manufacturing capacity. Remington just filed for bankruptcy.
The helium is in there because it's less dense than air, reducing friction and hence heat and power consumption. Since the whole point is low density, I imagine it's at 1 atm or less, so the worst case would be air mixes in and makes the drive gradually run hotter. For cleanliness reasons drives are normally sealed anyway, so I don't see this being a big problem.
The helium shortage has been wildly overstated: https://www.wired.com/2016/06/dire-helium-shortage-vastly-inflated/
Most helium comes from natural gas wells, and until recently prices were so low that most gas producers didn't bother to separate it out -- even though removing the non-flammable helium improved the heating value of the gas. It just wasn't worth enough to be worth the trouble. The reason for the depressed prices was the US selling off its massive, 1920s-era helium reserve. So when the reserve started to get low, people worried that there was no production capacity in place to replace it -- and then the blockade of Qatar precipitated a minor panic because it temporarily cut off about 30% of the worldwide supply.
Having once found my car's diesel fuel filter clogged with fungus threads, I'm thoroughly impressed by their ability to tolerate extreme living conditions.
Given some of the impressive projects I've seen people build in Minecraft, I suspect it's at least as creative as painting. My generation's version was Lego, but that was more expensive and carried far more of a risk of stepping on something sharp in the middle of the night.
I was slow at math in primary, homework was seemingly hours and hours and hours as the sun set and I wasn't getting released to play...
I was good at math, and actually enjoyed it, but had issues staying focused and motivating myself. I would sit at my desk and do basically everything but the homework in front of me. I remember listening to a local AM station I liked switch to low power as the sun went down, while I sat there with my homework in front of me.
It wouldn't be until adulthood that I'd be diagnosed with inattentive-type ADD, but looking back that's clearly what was going on.
The *EARTH* is *NOT* *THAT* *FRAGILE*.
When people assert this as if it's an iron-clad fact, I'm always reminded of the people who said we could never run out of passenger pigeons, because God would never let one of his creations go extinct.
(This particular effect, though, does seem to be no big deal except for radio services. It's probably worth learning more about, though, so we can anticipate it for other big launches.)
Biting the hand that feeds IT © 1998–2018