how-to block ads
|
Uniqs:
992 |
Share Topic
 |
 |
|
|
 BF69Premium
join:2004-07-28
Camden, TN | If you're not stealing then this doesn't affect you now does it? It's not affecting me. And don't tell me about "legitimate" torrent use. What's that, 0.1%? | |
|   kenn10
join:2003-09-10
Kennesaw, GA
1 edit | Re: If you're not stealing I agree with you, BF69. So long as my legitimate data and voice packets flow with minimal delay, I'm happy.
The obvious concern becomes over how Comcast can decide what is legitimate and what is not. We've seen cable companies try to block port 5060 for VOIP traffic as well.
The problem is with the torrent providers and they are who the internet services should go after. Its easy...you see torrents, shut down their Modem. | |
| | rugbyI think I know it all.VIP
join:2000-09-26
Indianapolis, IN | Re: If you're not stealing Your statement makes no sense. You make assumptions that torrents = piracy. That's not true, so get a new argument please.
ISP's should not throttle traffic period. If they don't have enough bandwidth maybe they should take some of their massive profits and invest it back into their infrastructure and stop being babies. | |
| |  BoogeymanDrive it like you stole itPremium
join:2002-12-17
Panama City, FL | Seriously, do a little research. Even the US government is using bittorrent to distribute large files. | |
|
rugbyI think I know it all.VIP
join:2000-09-26
Indianapolis, IN
1 edit | Please stop it with the self-righteousness. You know NOT of what you speak. Just about every large download available has a torrent.
Fixed typo. | |
|  dadkinsCan you do Blu?Premium,MVM
join:2003-09-26
Hercules, CA
kudos:18 | said by BF69:then this doesn't affect you now does it? It's not affecting me. And don't tell me about "legitimate" torrent use. What's that, 0.1%? But, I do use BT - for legal videos.
VUZE anyone?
This DOES affect me - and it's bullshit!
Percentage is irrelavent when *ALL* BT is getting hosed in a blanket action.
We won't mention all the other sources for "illegal" files that cruise right along without impedence.
--
Think outside the Fox... Opera | |
| |  ropeguruPremium
join:2001-01-25
Mechanicsville, VA | Re: If you're not stealing said by dadkins:We won't mention all the other sources for "illegal" files that cruise right along without impedance. Exactly... How about shutting down all the binary items in the text based news groups. That would probably relieve more traffic than torrent downloads. | |
| | | | said by dadkins:said by BF69:then this doesn't affect you now does it? It's not affecting me. And don't tell me about "legitimate" torrent use. What's that, 0.1%? But, I do use BT - for legal videos. VUZE anyone? This DOES affect me - and it's bullshit! Percentage is irrelavent when *ALL* BT is getting hosed in a blanket action. We won't mention all the other sources for "illegal" files that cruise right along without impedence. The issue isn't legal or illegal for Comcast. It is that ALL bittorrent traffic screws up the system. Want to blame someone - blame the developers of a protocol that was designed to be a bandwidth hog on purpose without any regard for any other internet traffic.
--
My BLOG .. .. Internet News .. .. My Web Page | |
| | |  funchordsHelloPremium,MVM
join:2001-03-11
Yarmouth Port, MA
kudos:5 | Re: If you're not stealing said by fAcEtIOUs:The issue isn't legal or illegal for Comcast. It is that ALL bittorrent traffic screws up the system. There is no evidence of this often-repeated claim.
said by fAcEtIOUs:Want to blame someone - blame the developers of a protocol that was designed to be a bandwidth hog on purpose All TCP applications will use all of the bandwidth that they can. The exceptions are programs that have a built-in limiter, which includes the top-two BitTorrent applications (uTorrent and Azureus-Vuze).
Furthermore, none of these applications do anything to abuse bandwidth. They can only use the bandwidth delegated to the subscriber by the modem and the CMTS scheduler.
said by fAcEtIOUs:without any regard for any other internet traffic. The applications don't "hack the modem" or DDOS other users in order to gain an advantage. They use the OS's network stack -- just like any other TCP application.
I want VOIP packets to have priority over BitTorrent, too. But I don't want Comcast, which should be a neutral transit provider, to make such decisions for me. I do hope that Comcast's eventual solution means that Comcast will acknowledge and react to the user's DSCP (QoS) flags in the IP header. Yes, some users will abuse these -- but such abuse is addressable by policy. If P2P file-sharers move their QoS flags above "best effort," I say throttle ALL their traffic to 256/64 for a month as a response!
--
Robb Topolski -= funchords.com =- Hillsboro, Oregon
FCC Public Hearing on the Future of the Internet - Thursday, April 17th - Stanford Univ., Calif. | |
| | | | RadioDoc58ef2c0Premium,ExMod 2000-03
join:2000-05-11 | Re: If you're not stealing Bravo! Excellent response and excellent description.
--
Toolmaster of La Grange. | |
| | | | openbox9
join:2004-01-26
Alexandria, VA
kudos:2 | said by funchords:There is no evidence of this often-repeated claim. There's no proof that the often-repeated claim isn't valid. You seem fairly intelligent with IP, do you really think that unmanaged P2P applications on a network aren't capable of wreaking havoc?said by funchords:All TCP applications will use all of the bandwidth that they can. They only use the bandwidth that they need. P2P apps (BT in particular) were built to use everything available without reservation. Go look at the statement from Bram Cohen regarding one of his objectives in creating the BT protocol.said by funchords:They can only use the bandwidth delegated to the subscriber by the modem and the CMTS scheduler. Great. I say limit the crap out of P2P (ab)users. I recommend ISPs draw a line in the sand, cap usage, and charge overages along with connection throttling across the board for the users across that line. Beyond that, open the floodgates and let consumers use their connections however they feel is appropriate and for however much they're willing to pay. Flat billing needs to disappear in preparation of "protocol agnostic" management practices. | |
| | | | | funchordsHelloPremium,MVM
join:2001-03-11
Yarmouth Port, MA
kudos:5
1 edit | Re: If you're not stealingIt is that ALL bittorrent traffic screws up the system. said by openbox9:said by funchords:There is no evidence of this often-repeated claim. There's no proof that the often-repeated claim isn't valid.
C:\Users\Robb>tracert www.dslreports.com
Tracing route to dslreports.com [209.123.109.175]
over a maximum of 30 hops:
1 <1 ms <1 ms <1 ms 192.168.177.251
2 * * * Request timed out.
3 10 ms 10 ms 17 ms ge-1-11-ur03.beaverton.or.bverton.comcast.net [68.87.218.89]
4 46 ms 10 ms 11 ms te-9-1-ur04.beaverton.or.bverton.comcast.net [68.87.216.50]
5 33 ms 35 ms 29 ms te-8-4-ur05.beaverton.or.bverton.comcast.net [68.87.216.101]
6 30 ms 30 ms 41 ms te-9-1-ur06.beaverton.or.bverton.comcast.net [68.87.216.98]
You seem fairly intelligent with IP, do you really think that unmanaged P2P applications on a network aren't capable of wreaking havoc? If you configure anything badly, it will operate badly. uTorrent or Azureus, configured according to the instructions provided by those teams, behaves quite well.
All TCP applications will use all of the bandwidth that they can. They only use the bandwidth that they need.
P2P apps (BT in particular) were built to use everything available without reservation. Go look at the statement from Bram Cohen regarding one of his objectives in creating the BT protocol. Regardless of what Bram said, it behaves no differently with regard to filling the pipe than FTP, HTTP, SMTP, POP3, IMAP, etc.. There are exceptions, but it's a designed-in exception -- BITS, for example. I don't know how I can explain it any better.
They can only use the bandwidth delegated to the subscriber by the modem and the CMTS scheduler. Great. I say limit the crap out of P2P (ab)users.
--
Robb Topolski -= funchords.com =- Hillsboro, Oregon
FCC Public Hearing on the Future of the Internet - Thursday, April 17th - Stanford Univ., Calif. | |
| | | | | | openbox9
join:2004-01-26
Alexandria, VA
kudos:2 | Re: If you're not stealing said by funchords:I'd say I've pretty much shown that my bittorrent traffic is not screwing up anything. Surely you aren't naive enough to believe that one instance of BT and your extremely limited scope of testing correlates to P2P applications not causing capacity issues on your provider's network are you? If so, then I take back what I said about you appearing to be relatively knowledgeable in your discussions.said by funchords:If you configure anything badly, it will operate badly. uTorrent or Azureus, configured according to the instructions provided by those teams, behaves quite well. I'd replace quite well with relatively well. There can be a huge difference. The problem is that end users have a choice in the configuration and contrary to your philosophy, most end users subscribe to the theory that they want everything and then some. A lot of users have no qualms about screwing everyone so that they can max about their BT client on this 8/512 connection with 150+ connections.said by funchords:I'm saying that TCP applications (actually IP applications) will fill the pipe until it has consumed all of the bandwidth it needs. One connection or 10 connections, it doesn't matter. I agree with this statement.said by funchords:Regardless of what Bram said, it behaves no differently with regard to filling the pipe than FTP, HTTP, SMTP, POP3, IMAP, etc.. There are exceptions, but it's a designed-in exception -- BITS, for example. I don't know how I can explain it any better. As you are well aware, it's not just about the amount of bandwidth consumed, but also the number of connections established. | |
| | | | | | | funchordsHelloPremium,MVM
join:2001-03-11
Yarmouth Port, MA
kudos:5
1 edit | Re: If you're not stealing Thanks, openbox9 -- nice conversation here.
said by openbox9:As you are well aware, it's not just about the amount of bandwidth consumed, but also the number of connections established. I am well aware of that.
But what most people don't realize is that the duration of that difference (congestion response) is measured in seconds not minutes. Over the course of a full busy day, perhaps cumulatively in minutes but certainly not hours. Unless the network is criminally oversold, the size of the problem no big whoop.
Like I've said on George Ou's comments, it's not an unfairness worth spending time to fix.
--
Robb Topolski -= funchords.com =- Hillsboro, Oregon
FCC Public Hearing on the Future of the Internet - Thursday, April 17th - Stanford Univ., Calif. | |
| | | | | | | | openbox9
join:2004-01-26
Alexandria, VA
kudos:2 | Re: If you're not stealing said by funchords:Thanks, openbox9 -- nice conversation here. I hope that's sincerity because I enjoy debating a relatively educated viewpoint.said by funchords:But what most people don't realize is that the duration of that difference (congestion response) is measured in seconds not minutes. Over the course of a full busy day, perhaps cumulatively in minutes but certainly not hours. Unless the network is criminally oversold, the size of the problem no big whoop. Except during the prime usage times during which a lot of people love to flaunt that "xxx ISP sucks, why don't they spend money to upgrade their sh!t network". | |
| | | | | | | | | |
| | | | | | funchordsHelloPremium,MVM
join:2001-03-11
Yarmouth Port, MA
kudos:5 | said by openbox9:said by funchords:I'd say I've pretty much shown that my bittorrent traffic is not screwing up anything. Surely you aren't naive enough to believe that one instance of BT and your extremely limited scope of testing correlates to P2P applications not causing capacity issues on your provider's network are you? If so, then I take back what I said about you appearing to be relatively knowledgeable in your discussions. Don't take it back. Remember, the statement was, "ALL bittorrent traffic screws up the system" (emphasis on "ALL" was his).
Really, what I proved was that it was an inadvisable thing for him to say.
--
Robb Topolski -= funchords.com =- Hillsboro, Oregon
FCC Public Hearing on the Future of the Internet - Thursday, April 17th - Stanford Univ., Calif. | |
| | | | | | | |  espaethDigital PlumberPremium,MVM
join:2001-04-21
Minneapolis, MN
kudos:2
 ·Clear Wireless
| Re: If you're not stealing said by funchords: Remember, the statement was, "ALL bittorrent traffic screws up the system" (emphasis on "ALL" was his). Really, what I proved was that it was an inadvisable thing for him to say. Ok. I'll bite -- how else would you describe it?
P2P networks like BT are really quite worthless with only a couple people. They are only really valuable when they become a "N-user" network (for large values of N), and coincidentally they also create the most impact in that scenario.
No single raindrop believes it is responsible for the flood. | |
| | | | | | | | openbox9
join:2004-01-26
Alexandria, VA
kudos:2 | said by funchords:Really, what I proved was that it was an inadvisable thing for him to say. One could also say the same thing about your statement. Without knowing the peers on your node and the current utilization, I'm not sure your example does anything but demonstrate that your configuration of BT doesn't affect your response times to DSLR on a limited basis given the current conditions. How does you example change during "prime time" with numerous other BT clients fired up on your node? | |
|
| | dadkinsCan you do Blu?Premium,MVM
join:2003-09-26
Hercules, CA
kudos:18
1 edit | Nothing wrong with the protocol... it's the use(over use) by some people that may overload the network.
Thing is, we all know it's BS!
Why sell a fast connection if it cannot handle a little use?
16/2 is 16/2 no matter what I use for transit.
Whether I am uploading files to my FTP, or sharing the videos I get via BT legally... I can only send and receive at the rated speeds I pay for, right?
Remember, I trash the connection every month - at 22-27GB per month!
But my BT is getting hosed also. WTF?
Comcast(and the other ISPs) throttling this protocol are doing so in a blanket action.
If me (or even you) are using a BT client to get MariposaHD legally, and that is "hurting" their network... fix the damn network! 
While Little Jimmy is lit up 24/7 downloading/sharing whatever via BT and *may* be using his/his Mommy's connection to it's limits, it's irrelevant to those of us that want the latest HD In Motion(legal Exotic car videos in HD).
But we are getting our share ratios screwed into the ground because of no seeding/uploading? Bullshit!
My shit through BT is LEGAL, period!
My shit is getting screwed along side of all of what you are calling "piracy".
Either discern and kill *ONLY* the "bad" things, or back the hell off completely!
Trashing all of it is lame!
Better start killing HTTP too, huh?
WAY too much traffic on that protocol!
FTP too! 
--
Think outside the Fox... Opera | |
| | | | funchordsHelloPremium,MVM
join:2001-03-11
Yarmouth Port, MA
kudos:5 | Re: If you're not stealing said by dadkins:My shit through BT is LEGAL, period! My shit is getting screwed along side of all of what you are calling "piracy". Either discern and kill *ONLY* the "bad" things, or back the hell off completely! Trashing all of it is lame! Better start killing HTTP too, huh? WAY too much traffic on that protocol! FTP too! All of the above protocols are TCP protocols. Comcast would solve the whole problem if they banned TCP! It's the illegalist of them all!     
--
Robb Topolski -= funchords.com =- Hillsboro, Oregon
FCC Public Hearing on the Future of the Internet - Thursday, April 17th - Stanford Univ., Calif. | |
| | | | | |
| |  Jovi
join:2000-02-24
Mount Joy, PA | Hey TK. I think funchords has the technology knowledge down pat. Quit while your ahead? I think your fight of pirates is making you argue for the sake of arguing.
--
"Where's my coffee? Oh. I guess it's my turn to make it."  | |
|
ropeguruPremium
join:2001-01-25
Mechanicsville, VA | said by BF69:then this doesn't affect you now does it? It's not affecting me. And don't tell me about "legitimate" torrent use. What's that, 0.1%? But if I am not stealing, their throttling IS affecting me. They DO NOT distinguish between the two.
Please cite your references for legitimate use only being 0.1%. | |
| | | | Re: If you're not stealing Furthermore,
If they impede the 0.1% that is legit, isn't that an anticompetitive move? If, I as a buisness owner, decide to sell copies of my software via BT, comcast is blatantly blocking my ability to service the market | |
| | | RadioDoc58ef2c0Premium,ExMod 2000-03
join:2000-05-11 | Re: If you're not stealing Change 'software' to 'video on demand' and you've hit it right on the nose. This is just the first assault against anything and everything that competes with their core video product.
--
Toolmaster of La Grange. | |
|
| | The thing is that there is no studies done on legitimate versus illegitimate P2P use, so it's anybody's guess as to how much P2P traffic is legit or not.
--
True Happiness Must Come From Within | |
| | | said by BF69:And don't tell me about "legitimate" torrent use, because it would nullify my imaginary point. Edited for accuracy. I've seen you post this garbage numerous times, and in turn get rebuffed numerous times, yet you essentially close your eyes and plug your ears so you can just keep beating the same moronic drum. It's high time you gave it a rest. | |
|  SeleniaI love DebianPremium
join:2006-09-22
Lanesboro, MA
kudos:2
 ·Verizon Online DSL
·AT&T Wireless Br..
·Verizon Wireless..
1 edit | Bandwidth is not the issue! Connections are! Does anyone here really know the general limitations of current cable infrastructure in general and DOCSIS 2.0? Cable modems have a terrible protocol for such use, which DSL does not suffer. Whenever data needs to be sent to a host(or peer), an RTS(Request to send) flag is sent. If 2 RTS packets collide in the same cycle, they are then dropped and must be retried before the modems can send data. Cable nodes work much like hub based topology as network sharing goes, While the collision window is quite small, do remember you might have hundreds of people in your neighborhood holding massive amounts of BT connections, sending out an assault of RTS packets. Routing equipment may also be under strain from large numbers of connections. Their state tables might very well overflow to the point where entries get dropped, thus connections get abnormally terminated, or proceed with more delay than traffic shaping at its worst might introduce. Now imagine these kinds of drops/delays on normally unshaped apps, such as online games. 1000 ms pings anyone? I'll pass.
I am not an advocate of Comcast's techniques. There are some more transparent methods of eliminating network strain. Sufficient network upgrades(don't forget DOCSIS 3.0 and less congested nodes/headend gear) would also significantly alleviate network strain. Some forms of QoS queuing with reasonable parameters(as in just enough to let important traffic pass in reasonable time) may also go almost unnoticed on torrents, which are very high ping tolerant with proper TCP settings. | |
| | funchordsHelloPremium,MVM
join:2001-03-11
Yarmouth Port, MA
kudos:5
2 edits | Re: Bandwidth is not the issue said by Selenia: Whenever data needs to be sent to a host(or peer), an RTS(Request to send) flag is sent. If 2 RTS packets collide in the same cycle, they are then dropped and must be retried before the modems can send data....you might have hundreds of people in your neighborhood holding massive amounts of BT connections, sending out an assault of RTS packets. Question: If, within such a cycle, I simultaneously open 5 TCP connections to 5 different peers, does my modem XMIT 1 RTS or 5? Can you lead me to a reference that describes this? (I honestly don't know the answer to this.)
If the answer is 1, then your problem statement is reduced.
Routing equipment may also be under strain from large numbers of connections. Their state tables might very well overflow to the point where entries get dropped, thus connections get abnormally terminated, or proceed with more delay than traffic shaping at its worst might introduce. Routing equipment traditionally is connectionless. The IP packet was designed for routing, and it contains nothing about state. The Internet was designed to preform routing based on the IP header. An ISP never needs to track state.
I can only think of 3 reasons to do full-time Deep Packet Inspection (DPI, inspect behind the IP header) -- none of them good ones.
1. To re-route or re-prioritize users' packets from their normal handling, contrary to the user's or application's instructions available in the IP header
2. To perform "marketing" via TCP forgery that does ad insertion or
3. To perform "marketing" by monitoring user's behavior and delivering different marketing pitches later.
DPI is a solution without a problem, and unless being used reactively to diagnose some kind of network problem, it has no business being turned on.
Some forms of QoS queuing with reasonable parameters(as in just enough to let important traffic pass in reasonable time) may also go almost unnoticed on torrents, which are very high ping tolerant with proper TCP settings. Absolutely. RFC 2474 -- it's tanned, it's rested, it's ready to go, and it's free.
--
Robb Topolski -= funchords.com =- Hillsboro, Oregon
FCC Public Hearing on the Future of the Internet - Thursday, April 17th - Stanford Univ., Calif. | |
|
| |
|
