AP Updates Its Story: Comcast Doesn’t Block P2P
You may have missed a key follow-up to the story the Associated Press filed Oct. 19, which asserted that Comcast “blocks” file-sharing application BitTorrent in some cases.
The alleged smoking gun: The AP tried to share a digitized copy of the King James Bible using BitTorrent software and Comcast’s cable modems, and claimed that two out of three attempts to share the file were blocked altogether.
The dispatch was absolute catnip to “Net Neutrality” advocates, who pounced on it as evidence that the federal government needs to start regulating the way Internet providers manage their networks.
But on Tuesday night, the AP issued an update to the story, as pointed out by IP Democracy blogger Cynthia Brumfield. Rather than blocking P2P traffic, the later AP article says, Comcast merely subjects such file-sharing to delays.
“In one case, a BitTorrent file transfer was squelched, apparently by messages generated by Comcast, only to start 10 minutes later,” the AP’s Oct. 23 article said. “Other tests were called off after around 5 minutes, while the transfers were still stifled.”
Well. That’s a fundamental difference: Like saying the Holland Tunnel is closed, whereas in fact traffic is subject to 5- to 10-minute delays.
So you can’t start downloading your pirated copies of Spider-Man 3, Knocked Up or Ratatouille right away? Boo-hoo, brother.
In any event, neither AP story addressed the fact that BitTorrent’s file-sharing protocol is not the most reliable piece of technology ever invented.
There are a number of reasons why an attempt to download a file via BitTorrent might fail. Sometimes it works; sometimes it doesn’t. Why? Could be the BitTorrent “tracker” that points to “seeds” hosting the requested file is offline, or it may have changed its address, or your computer may be firewalling the file-sharing protocol.
If you’re feelin’ holy (or holier than thou), by the way, the Bible’s already posted all over the Web.
Todd Spangler commented:
Mark, thanks for your well-reasoned post. Let's put aside the question of what kind of content traverses P2P networks. The issue, as I've noted in another post, is: Do Comcast's actions here warrant FCC intervention? I don't think they do. Our Holland Tunnel metaphor breaks down because Comcast is not like a regional transportation authority. Comcast has very strong competition. This is a case where the free market can, and should, sort itself out -- do you really want to empower FCC bureaucrats to decide how ISPs monitor and manage their routers?
Mark commented:
Todd, I agree with most of your positions and posts, but your tone on this one is awfully disdainful. The issue is real - Comcast is blocking/slowing/otherwise deliberately reducing network availability to P2P protocols. This has been confirmed by other organizations rather than AP. You don't have to build a wall of bricks in front of a tunnel to make it ineffective. To stick with your Holland Tunnel analogy, plucking random people out of cars along the way, so that when emerging from the other side, those cars have to back and make the trip again to actually get through with all of their passengers, probably wouldn't be considered good service. If the transportation authority advertised that all roads and tunnels were available, and denied any problems or deliberate slowdowns, while all the while preventing people from making their trips, there would be an outcry - just as there is over Comcast's P2P issues. My personal days of downloading pirated movies are long gone - I realized the practice was unethical. Currently I use P2P for occasional, legal downloads. Many software applications today use P2P to get updates and patches, for example. There's more to this problem than the cheaters - there's a big company pretending they're doing nothing wrong, and people like you pretending that what they're doing is like nothing more than a 5-minute traffic delay on the way to work.
Kyle Luna commented:
I wonder how many of those tests were behind routers without the ports forwarded, using ports blacklisted by most trackers (many times the default ports used by the bit torrent client), or relied on upnp, which many routers and/or clients don't fully implement correctly (not to mention many routers don't even have upnp enabled by default).
Also, if they're using Vista or XP SP2, they could have fell victim to the max half-open TCP connections thing that's there to prevent the spreading of worms, but also slows down bit torrent traffic, requiring people to use a patcher to modify the tcpip.sys file.
I have a feeling most of these tests were done with the default settings, and nobody bothered to configure any of this stuff
Ken commented:
Thank you. I thought the AP story was awfully heavy-handed for an organization purportedly devoted to objective (or at least thorough) reporting. There was no mention of who conducted the tests, methodologies, assumptions, etc., just rip-snorting, one-sided assertions without any real support. The story's bias was over-the-top and obvious -- pretty ridiculous for an outfit like AP.
