Since the IP address of my PC is on a subnet local to my phone and my PC, different from the subnet used between my phone and the carrier, I would say NAT is used.
I'm telling you, and you can contact the network technicians at your carrier to verify this, that there is, indeed, NAT involved (in fact, we agree on this part). Carrier grade NAT on the carrier's network, not on your phone.
The IP space of the AdHoc LAN created when you wi-fi tether is, of course, local and quite likely on a different subnet than the one over which your phone talks to your carrier. That's how 1:1 routing works. 1 IP internal to your LAN routes to 1 IP external to your LAN. That is not NAT. It's also not pandering to the carriers; it's the very same thing dumb phones that allow tethering do. Back before 2G, it used to be circuit-switched data, your phone actually acted as a modem and you'd dial in, but tether-capable phones have been doing 1:1 routing for at least 2 decades now and today's phones use the same infrastructure. If your phone is doing NAT, it is doing you a disservice, because your carrier is, as well. The reasons for this have already been explained, but I'll take another go since you apparently completely missed that.
By the time phones technically gained the capability to perform NAT locally, we were already running out of IP4 addresses. There is not enough address space remaining to give each prone two separate public IP4 addresses, one for mobile and one for tethered data, the infrastructure is already in place to not need to do local NAT anyway, and doing local NAT would strain the CPU, RAM, and battery of your mobile device. As a result, mobile OSes don't implement NAT as we commonly talk about it, because it would be a battery life and performance killer and other solutions already exist; that's not to pander to the carriers, that's to make their products look better on paper and perform better in reality.
Actually... and I'm going to leave the above post as-is because it's all pertinent information... technically, yes, 1:1 routing involves translating network addresses. It's not NAT as we commonly talk about it, where one public IP is shared among many private devices, but yes, it is translation of network addresses, so from a very pedantic standpoint, it's NAT and yes, it is done on your phone. It lacks all of the connection and endpoint tracking functionality we commonly refer to as NAT, though; it's literally "anything coming in to this IP forwards to that one" and is commonly done in the radio chip itself, which certainly does not support what we commonly refer to as NAT.
Your phone is acting as a router, yes, but not NAT. Here's a test: Visit http://ip4.me/ from your phone. Now, tether your laptop or tablet and visit http://ip4.me/ from there. Same IP? If so, you're right, your phone is doing the NAT. Different IP? Different routing and carrier-level NAT. Simple test, really.
On T-Mobile with a Nexus 6 using the Android built-in tethering, my phone and tethered laptop have different public IP addresses. If the phone was doing NAT, the IP would be shared. Your phone is doing simple 1:1 routing of internal to external addresses, no NAT unless you're using something like Barnacle. Android itself does not do NAT. Period.
Which is funny to me, considering that our first interaction was you harassing me. As for whether or not I was on topic with my post earlier in this thread (in reply to the AC who first mentioned your name); the topic of the post I was replying to was how unwelcome your posts are and I was exactly on that topic. The AC may have been off-tpoic with his post, but my post was definitely in line with the topic he opened up.
Either way, I think the whole thing is rather amusing. Honestly, your constant posts about the hosts file are hilarious. Other than the repetitive and rambling nature of your posts, I don't see why people take such issue. Had you not attacked me in that other thread (unprovoked, at that) I'd have nothing against you; but you did, and I do. Mind you, I'd still have posted the same thing in this thread, but would have done so in a joking manner; now, I'm dead serious, you've really gotten that bad.
And yes, dozens of similar posts in a single thread is spam, I don't care how on-topic it may or may not be. To be clear, I'm referring to the following wording from that page: "Forum spam is the creation of advertising messages on Internet forums" and, as Slashdot (like other forums) is modeled after Usenet, "Usenet convention defines spamming as excessive multiple posting, that is, the repeated posting of a message (or substantially similar messages)".
The fact that they are routed to different APNs and identifiable by the carrier is itself a violation of net neutrality
Howso? I suppose, next, you'll claim that the fact that there is yet a 3rd APN involved in sending and receiving MMS messages is another violation, and a 4th network that routes voice, that's another violation, right? Better start filing those reports with the FCC then.
The reality is that there are actual technical reasons for routing tethering traffic through a different APN (which is done at the device leve, on your phone, by the way), not the least of which is that not every phone can do its own NAT (in fact, until a few years ago, the majority couldn't, the processing power and RAM just didn't exist in a device that portable) and even today where most technically can, the majority don't, partly because there is no need (the split infrastructure is already in place). When you use mobile data, your phone gets its own public IP address, which your carrier can manage because they know one subscribed device = 1 IP; when you tether, your session is routed over carrier grade NAT, sharing a public IP with many other devices, which is necessary because the carrier does not necessarily know how many devices you might tether at any given moment, far enough in advance to ensure that they have enough available IP addresses for everyone to use. Since most phones aren't also routers capable of performing NAT locally, this, again, is a necessity.
And they don't. The sell unlimited mobile (e.g. on your device) data, not unlimited tethering. The 2TB users are tethering past the 7GB limit they paid for.
They shouldn't even know if the data is coming from the phone itself or via tethering. Doing so is a violation of net neutrality, and is a bad thing.
Doing so by way of packet inspection is a violation of net neutrality. However, mobile data and tethered data are routed via two different APNs, which is the mechanism by which T-Mobile knows which is which. If you think it's not obvious what's going on when you go from 100MB mobile 7GB tether, to 7.1GB mobile 0GB tether, to 2TB mobile 0GB tether, either you're an idiot or you think everyone else is.
Indeed. And T-Mobile sells unlimited mobile (read: on your mobile device) data with limited tethering. The issue here is that people are bypassing the tethering limits.
I've found that there is usually a 3 month period between when T-Mobile releases a feature or plan and when the reps know anything meaningful about it. Consistently, I add features as they are released and have issues for 3 months thereafter, at which point everything is magically fixed, credits issues, a month of service comped, and life goes on.
It's gotten to the point where I'll add the feature and just expect to call them when the next 3 bills come out. I don't bother following up, I just make sure I've contacted them and it's documented; then, like clockwork, on the 3rd monthly call, everything is resolved, I'm credited for any overpayment or missed service, and given a credit for the next month's service as well. Like. Fucking. Clockwork.
Nope. Tethering, which is what is in question here, is not sold as an unlimited service, but as a limited add-on to an unlimited service. Subtle distinction, but important.
The unlimited LTE plan includes 5GB of tethering. So no, customers are not prohibited from tethering, but they're also not paying for unlimited tethering and the tethering limits are plainly displayed in marketing materials.
That's like a restaurant saying you aren't allowed to share the food you buy with someone else - only you are allowed to eat it.
I know they're harder to find now, but I seem to recall restaurants offering soup/salad bars as either a single trip alongside an entree, or as unlimited trips, as the entree itself. If you opted for the salad bar as a side, you could share off the one plate you were allowed; if you chose unlimited, you were not allowed to share. People still shared anyway and the restaurants typically looked the other way, so long as it was just a couple items here and there (after all, tasting might lead to additional sales in the future) and not one person ordering the unlimited salad bar and feeding a table of 4 with it.
So yes, that's exactly what it's like, and that's exactly what places like Eat'n Park have been doing since the 70's.
"Unlimited" usage of a 6 Mbit connection means that you can use the full 6 Mbit 24 hours a day, 7 days a week.
Indeed. And 5GB of tethering means you can use 5GB of tethering, even if you have unlimited LTE on your phone. And that's what T-Mobile sells: Unlimited LTE for your phone, 5GB of LTE tethering for devices that connect to your phone. They don't even really cut you off if you go over that; I've used ~20GB during a move when I had no other options and they didn't slow me until ~18GB. The issue here is that people are bypassing the tethering limits they accepted when they signed up for the service. Those people are thieves in the similar way to someone subscribing to basic cable and using a black box to get all the channels is a thief, but worse in that the bandwidth they're using incurs a cost for the provider that would not otherwise exist. So far, T-Mobile has been very gracious in their handling of these users. They threatened to terminate accounts over this last year but decided against it; that was a warning shot. This may not be.
And my achievements stand quite well on their own, people who need to know have copies of my resume. I sure do quite well in the technical field in which I consult for someone with "no demonstratable technical computing expertise", so I'll let you have that one as long as my bills are paid and there is food on my table and a roof over my head. And a comfortable lifestyle. Very comfortable. So much so, in fact, that I don't feel the need to trumpet all of my accomplishments to the world in some ego-maniacal tantrum.
You wrote a small application that pulls other peoples' domain block lists from the internet and assembles them into a hosts file. The hosts file, of course, being a throwback to ARPANET and not something you created. Now, that would be impressive.
who said my 30+GB/mo (not tethered) was abuse last time T-Mo said they were going to crack down on this (then didn't). It seems they've defined abuse and it is 2TB. Of tethering. Bite me.
I'm not sure how you think that was a correction. I didn't repeat what the article already stated, because the article had already stated it; that doesn't make me wrong or "in need of correction", it makes me "not redundant". Furthermore, you then go on to "invalidate" my mention that there was no patch for XP by stating that XP is no longer supported. Here's a newsflash: XP is still supported on POS platforms, which are widely deployed, so a patch is still necessary there. To top it off, XP was the most prevalent Windows version at the time of the incident, so your claim is really that Microsoft left the majority of their customers vulnerable.
Brilliant.
I know I'm not going to win this argument because you've clearly got nothing but time on your hands with which to craft your arguments so as to appear to be correct and on-topic while simply directing the argument away from what you perceive to be your opponent's area of expertise. That's fine, your tactics make you feel like a big man and the whole exchange is highly entertaining for me, because I know why it appears that I am losing.
Let me ask you this: If your application is so great, why do you have to spam Slashdot to sell it? Your high horse is complaining of back pains. You should get off it.
P.S.=> Spamming a forum about your "awesome" application and how great you think you are is never on topic.
And much would be gained. The companies expending such huge advertising budgets would suddenly no longer be spending that money. That means less cost, so lower prices for their products and more sales; or, more realistically, larger management and CxO bonuses. Either way, everyone wins.
Some checkums can only be guaranteed to catch a single bit error but, in practice, it is relatively rare that multiple-bit errors "line up" just right to create a correct checksum. There are better checksums available, as well, which will catch multiple-bit errors, as well as identify which segment of the data contains the error. Someone more familiar with the topic will have to name these, as I haven't studied them in years.
A phone operating under normal conditions isn't particularly useful for triangulating its signal.
Actually, TDMA networks (which GSM is) with extremely small time slices (which GSM definitely is), are so timing-sensitive that all stations (read: phones) are constantly adjusting their transmit timing to compensate for distance. If the phone in your hand can calculate round-trip time to do that, the tower it connects to can certainly do so to calculate distance. After all, we know how fast the signal moves.
Sure, it's not triangulation, which requires three points (it's all on the name), but it provides a radius along (not within) which you are located. Actually, an arc of either 120 or 60 degrees, depending on whether the tower uses 3 or 6 semi-directional elements. If your signal reaches a 2nd tower, that's all the information they need; you are located where the two arcs intersect. A third point (for triangulation) is only necessary when distance information is not available.
Slashdot Mirror
Since the IP address of my PC is on a subnet local to my phone and my PC, different from the subnet used between my phone and the carrier, I would say NAT is used.
I'm telling you, and you can contact the network technicians at your carrier to verify this, that there is, indeed, NAT involved (in fact, we agree on this part). Carrier grade NAT on the carrier's network, not on your phone.
The IP space of the AdHoc LAN created when you wi-fi tether is, of course, local and quite likely on a different subnet than the one over which your phone talks to your carrier. That's how 1:1 routing works. 1 IP internal to your LAN routes to 1 IP external to your LAN. That is not NAT. It's also not pandering to the carriers; it's the very same thing dumb phones that allow tethering do. Back before 2G, it used to be circuit-switched data, your phone actually acted as a modem and you'd dial in, but tether-capable phones have been doing 1:1 routing for at least 2 decades now and today's phones use the same infrastructure. If your phone is doing NAT, it is doing you a disservice, because your carrier is, as well. The reasons for this have already been explained, but I'll take another go since you apparently completely missed that.
By the time phones technically gained the capability to perform NAT locally, we were already running out of IP4 addresses. There is not enough address space remaining to give each prone two separate public IP4 addresses, one for mobile and one for tethered data, the infrastructure is already in place to not need to do local NAT anyway, and doing local NAT would strain the CPU, RAM, and battery of your mobile device. As a result, mobile OSes don't implement NAT as we commonly talk about it, because it would be a battery life and performance killer and other solutions already exist; that's not to pander to the carriers, that's to make their products look better on paper and perform better in reality.
Actually... and I'm going to leave the above post as-is because it's all pertinent information... technically, yes, 1:1 routing involves translating network addresses. It's not NAT as we commonly talk about it, where one public IP is shared among many private devices, but yes, it is translation of network addresses, so from a very pedantic standpoint, it's NAT and yes, it is done on your phone. It lacks all of the connection and endpoint tracking functionality we commonly refer to as NAT, though; it's literally "anything coming in to this IP forwards to that one" and is commonly done in the radio chip itself, which certainly does not support what we commonly refer to as NAT.
Your phone is acting as a router, yes, but not NAT. Here's a test: Visit http://ip4.me/ from your phone. Now, tether your laptop or tablet and visit http://ip4.me/ from there. Same IP? If so, you're right, your phone is doing the NAT. Different IP? Different routing and carrier-level NAT. Simple test, really.
On T-Mobile with a Nexus 6 using the Android built-in tethering, my phone and tethered laptop have different public IP addresses. If the phone was doing NAT, the IP would be shared. Your phone is doing simple 1:1 routing of internal to external addresses, no NAT unless you're using something like Barnacle. Android itself does not do NAT. Period.
They all have the Wifi hotspot function
Yes, they do.
which uses NAT
No, it doesn't.
So what are you talking about?
Clearly, something I know and you do not.
but, you guys constantly "harass" me
Which is funny to me, considering that our first interaction was you harassing me. As for whether or not I was on topic with my post earlier in this thread (in reply to the AC who first mentioned your name); the topic of the post I was replying to was how unwelcome your posts are and I was exactly on that topic. The AC may have been off-tpoic with his post, but my post was definitely in line with the topic he opened up.
Either way, I think the whole thing is rather amusing. Honestly, your constant posts about the hosts file are hilarious. Other than the repetitive and rambling nature of your posts, I don't see why people take such issue. Had you not attacked me in that other thread (unprovoked, at that) I'd have nothing against you; but you did, and I do. Mind you, I'd still have posted the same thing in this thread, but would have done so in a joking manner; now, I'm dead serious, you've really gotten that bad.
And yes, dozens of similar posts in a single thread is spam, I don't care how on-topic it may or may not be. To be clear, I'm referring to the following wording from that page: "Forum spam is the creation of advertising messages on Internet forums" and, as Slashdot (like other forums) is modeled after Usenet, "Usenet convention defines spamming as excessive multiple posting, that is, the repeated posting of a message (or substantially similar messages)".
The fact that they are routed to different APNs and identifiable by the carrier is itself a violation of net neutrality
Howso? I suppose, next, you'll claim that the fact that there is yet a 3rd APN involved in sending and receiving MMS messages is another violation, and a 4th network that routes voice, that's another violation, right? Better start filing those reports with the FCC then.
The reality is that there are actual technical reasons for routing tethering traffic through a different APN (which is done at the device leve, on your phone, by the way), not the least of which is that not every phone can do its own NAT (in fact, until a few years ago, the majority couldn't, the processing power and RAM just didn't exist in a device that portable) and even today where most technically can, the majority don't, partly because there is no need (the split infrastructure is already in place). When you use mobile data, your phone gets its own public IP address, which your carrier can manage because they know one subscribed device = 1 IP; when you tether, your session is routed over carrier grade NAT, sharing a public IP with many other devices, which is necessary because the carrier does not necessarily know how many devices you might tether at any given moment, far enough in advance to ensure that they have enough available IP addresses for everyone to use. Since most phones aren't also routers capable of performing NAT locally, this, again, is a necessity.
And a factual inaccuracy is still a factual inaccuracy. Which I was correcting. Your most resent comment would be correct, however.
If a customer consumes 2TB of data a month, does it matter if they were tethering?
Yes, because they're sold unlimited data on their mobile device and limited throttling.
And they don't. The sell unlimited mobile (e.g. on your device) data, not unlimited tethering. The 2TB users are tethering past the 7GB limit they paid for.
They shouldn't even know if the data is coming from the phone itself or via tethering. Doing so is a violation of net neutrality, and is a bad thing.
Doing so by way of packet inspection is a violation of net neutrality. However, mobile data and tethered data are routed via two different APNs, which is the mechanism by which T-Mobile knows which is which. If you think it's not obvious what's going on when you go from 100MB mobile 7GB tether, to 7.1GB mobile 0GB tether, to 2TB mobile 0GB tether, either you're an idiot or you think everyone else is.
Indeed. And T-Mobile sells unlimited mobile (read: on your mobile device) data with limited tethering. The issue here is that people are bypassing the tethering limits.
I've found that there is usually a 3 month period between when T-Mobile releases a feature or plan and when the reps know anything meaningful about it. Consistently, I add features as they are released and have issues for 3 months thereafter, at which point everything is magically fixed, credits issues, a month of service comped, and life goes on.
It's gotten to the point where I'll add the feature and just expect to call them when the next 3 bills come out. I don't bother following up, I just make sure I've contacted them and it's documented; then, like clockwork, on the 3rd monthly call, everything is resolved, I'm credited for any overpayment or missed service, and given a credit for the next month's service as well. Like. Fucking. Clockwork.
Funny, it's more reliable for me than most of the free wi-fi I encounter. Bay Area, including San Francisco.
because inevitably the ISPs like TMobile and Verizon...
are on different sides of the issue. T-Mobile actually supported (and still supports) Net Neutrality. Verizon, not so much.
Nope. Tethering, which is what is in question here, is not sold as an unlimited service, but as a limited add-on to an unlimited service. Subtle distinction, but important.
The unlimited LTE plan includes 5GB of tethering. So no, customers are not prohibited from tethering, but they're also not paying for unlimited tethering and the tethering limits are plainly displayed in marketing materials.
That's like a restaurant saying you aren't allowed to share the food you buy with someone else - only you are allowed to eat it.
I know they're harder to find now, but I seem to recall restaurants offering soup/salad bars as either a single trip alongside an entree, or as unlimited trips, as the entree itself. If you opted for the salad bar as a side, you could share off the one plate you were allowed; if you chose unlimited, you were not allowed to share. People still shared anyway and the restaurants typically looked the other way, so long as it was just a couple items here and there (after all, tasting might lead to additional sales in the future) and not one person ordering the unlimited salad bar and feeding a table of 4 with it.
So yes, that's exactly what it's like, and that's exactly what places like Eat'n Park have been doing since the 70's.
"Unlimited" usage of a 6 Mbit connection means that you can use the full 6 Mbit 24 hours a day, 7 days a week.
Indeed. And 5GB of tethering means you can use 5GB of tethering, even if you have unlimited LTE on your phone. And that's what T-Mobile sells: Unlimited LTE for your phone, 5GB of LTE tethering for devices that connect to your phone. They don't even really cut you off if you go over that; I've used ~20GB during a move when I had no other options and they didn't slow me until ~18GB. The issue here is that people are bypassing the tethering limits they accepted when they signed up for the service. Those people are thieves in the similar way to someone subscribing to basic cable and using a black box to get all the channels is a thief, but worse in that the bandwidth they're using incurs a cost for the provider that would not otherwise exist. So far, T-Mobile has been very gracious in their handling of these users. They threatened to terminate accounts over this last year but decided against it; that was a warning shot. This may not be.
The phone forwards tethered data over a different APN. Simple.
Where's your Mac version?
And my achievements stand quite well on their own, people who need to know have copies of my resume. I sure do quite well in the technical field in which I consult for someone with "no demonstratable technical computing expertise", so I'll let you have that one as long as my bills are paid and there is food on my table and a roof over my head. And a comfortable lifestyle. Very comfortable. So much so, in fact, that I don't feel the need to trumpet all of my accomplishments to the world in some ego-maniacal tantrum.
You wrote a small application that pulls other peoples' domain block lists from the internet and assembles them into a hosts file. The hosts file, of course, being a throwback to ARPANET and not something you created. Now, that would be impressive.
who said my 30+GB/mo (not tethered) was abuse last time T-Mo said they were going to crack down on this (then didn't). It seems they've defined abuse and it is 2TB. Of tethering. Bite me.
I'm not sure how you think that was a correction. I didn't repeat what the article already stated, because the article had already stated it; that doesn't make me wrong or "in need of correction", it makes me "not redundant". Furthermore, you then go on to "invalidate" my mention that there was no patch for XP by stating that XP is no longer supported. Here's a newsflash: XP is still supported on POS platforms, which are widely deployed, so a patch is still necessary there. To top it off, XP was the most prevalent Windows version at the time of the incident, so your claim is really that Microsoft left the majority of their customers vulnerable.
Brilliant.
I know I'm not going to win this argument because you've clearly got nothing but time on your hands with which to craft your arguments so as to appear to be correct and on-topic while simply directing the argument away from what you perceive to be your opponent's area of expertise. That's fine, your tactics make you feel like a big man and the whole exchange is highly entertaining for me, because I know why it appears that I am losing.
Let me ask you this: If your application is so great, why do you have to spam Slashdot to sell it? Your high horse is complaining of back pains. You should get off it.
P.S.=> Spamming a forum about your "awesome" application and how great you think you are is never on topic.
He's gotten so bad lately, I hear Google is talking about changing the file extension for Android apps.
And much would be gained. The companies expending such huge advertising budgets would suddenly no longer be spending that money. That means less cost, so lower prices for their products and more sales; or, more realistically, larger management and CxO bonuses. Either way, everyone wins.
Some checkums can only be guaranteed to catch a single bit error but, in practice, it is relatively rare that multiple-bit errors "line up" just right to create a correct checksum. There are better checksums available, as well, which will catch multiple-bit errors, as well as identify which segment of the data contains the error. Someone more familiar with the topic will have to name these, as I haven't studied them in years.
A phone operating under normal conditions isn't particularly useful for triangulating its signal.
Actually, TDMA networks (which GSM is) with extremely small time slices (which GSM definitely is), are so timing-sensitive that all stations (read: phones) are constantly adjusting their transmit timing to compensate for distance. If the phone in your hand can calculate round-trip time to do that, the tower it connects to can certainly do so to calculate distance. After all, we know how fast the signal moves.
Sure, it's not triangulation, which requires three points (it's all on the name), but it provides a radius along (not within) which you are located. Actually, an arc of either 120 or 60 degrees, depending on whether the tower uses 3 or 6 semi-directional elements. If your signal reaches a 2nd tower, that's all the information they need; you are located where the two arcs intersect. A third point (for triangulation) is only necessary when distance information is not available.