Sometimes I wonder if people really understand how peer-to-peer file sharing works. I listened to the CRTC discussions and I am following some of the “post-game analysis”.
Some of the rhetoric is just plain silly. Frankly, I think that there are a lot of outspoken people who might want to stop to listen and think before speaking out.
Like the call for a class action lawsuit that I read about on Tech Media Reports:
Yet Rogers disclosed that the upload speeds are actually identical for P2P traffic as they are both throttled equally. I find that extremely deceptive and can’t help but wonder whether they open themselves up to a class action lawsuit by Extreme subscribers who don’t get the service they think they do.
I’m no lawyer, but it seems to me that before someone sues someone else, you need to have someone who experienced some harm.
Let’s go back to look at the first principles for what is meant by P2P file transfer in order to understand why there is no harm to the uploader.
Most consumers subscribe to an asymmetric internet connection. That means that it works faster in one direction than the other. For most of us, we receive information (download) faster than we can provide it (upload).
So, lets say that I want to download a file that you have. If I get it from you directly, the fastest I could possibly receive it would be based on your upload connection speed, regardless of my faster download speed. I might get frustrated because I subscribe to a fast download capability and I might even blame my ISP, even though the real problem is that the source of the traffic is what is slowing down the transfer.
So, either you could upgrade to a very expensive symmetric internet service, or I might find a creative way to get the file faster. Many of the types of files that I want to get from you (music, movies, courseware, etc.) are sitting on other computers, not just yours. So some bright minds thought that I could get the file twice as fast faster if I got a few pieces from you, and a few pieces from another person at the same time. After all, you may only be uploading at say, 640 kbps and my download capacity is 8 times that. So, if two is better than one, why not try to get 8 streams working at once, or even more.
The P2P file transfer software tries to find people from around the world who have the file that I want and gets them all to provide pieces so that we all max out the upload and download capacity of our pipes. And, as soon as I get the first piece of the file, my computer gets identified as a potential source for other people.
You get the picture.
The software, by design, is supposed to keep the pipes full.
When I use the peer-to-peer file sharing application, my purpose is to have access to files found on other people’s computers. While I may altruistically want to contribute the files on my computer to the global pool of access points, I am trying to imagine a situation where I have a real concern about whether other people can get my files at full speed versus any kind of upload constraints that are imposed by my ISP.
After all, by design, the software will find other computers that are also sources for the file, so the person downloading from me isn’t harmed either.
I understand frustration by users in having an ISP manage download speeds. That could add to the time it takes to receive the whole file. But even those more aggressive management techniques don’t impact real-time applications like streaming video or other bulk file transfer protocols such as direct point-to-point downloads, such as those used by most companies.
P2P file transfer is not how you send an email or send your photos or update your blog or file your term paper, or do real time streaming. What is wrong with managing P2P upload speeds to ensure the proper operation of the rest of the applications – for you and everyone else served by your ISP?
You obviously haven't used P2P networks very much. The speed at which one can download content is very much a function dependent on one's upload speed.
First, if you aren't uploading, or your share ratio for the current torrent is too low, many clients will not allow you to download from them, or will limit the speed at which you can download. Since the number of connections is, in fact, finite, it can be very difficult to find satisfactory download speeds when one's upload speed is hamstrung.
Second, if one's share ratio remains too low between files, many tracker sites will simply refuse to send the torrent file. You then have to wait until enough data has been uploaded to permit downloading to commence once again.
Third, by your own logic, we should logically limit P2P upload speeds to 1KB/s, since it doesn't really matter anyway.
Yes, the purpose of P2P client applications is to maximize my use of my bandwidth, which I've paid for. Complaining that I am utilizing bandwidth beyond what the statistical models estimate is 'typical' is none of my concern. If my ISP needs to limit my bandwidth consumption, all they need to do is tell me what the business model is, and what it will cost me to download that next GB. They just need to make it predictable, consistent, and transparent, and let the market decide who the winners are. Anything less is false advertising, plain and simple.
You know, starting with "You obviously haven't used P2P very much" was insulting, untrue and didn't contribute to your otherwise interesting comment. I am intrigued by your feeling a need to lash out. Thanks for contributing to the view that I expressed last week.
With respect to your points:
1. change torrent clients to one that doesn't punish; as you indicate, not every client is as ill-behaved to encourage network congestion.
2. see number 1
3. that isn't my logic; such throttling isn't necessary to manage the network.
Your last point sounds like you must be a lot of fun to watch at a buffet. How many times do you regurgitate in order to get your money's worth? See number 1 above.
You're absolutely right Mark, and I regretted it as soon as I had hit the publish button. My apologies for 'lashing out'. I may often disagree with your comments, but that is no excuse. Sometimes we can forget there is a real person on the other end.
Let me respond to your responses:
1. It is not my choice of client that dictates the speed at which I can download from someone else. It is the choice of the other person's client application that dictates how much of their upload bandwidth they will allow me to consume.
2. See #1 above.
3. By your logic, I suffer no harm by limiting my P2P upload speed. 1KB/s may be an exaggeration, but where does one draw the line? As a user, I care not for my ISP's internal management of the network; I want MY internet service to function as advertised. If it can't, then one must change the service or the advertisement, not the user's expectations.
Your buffet comment implies moral judgement of the actions of the internet user. You are implying excess on the part of the internet user, when from my perspective, it is the service that is deficient. If an all-you-can-eat buffet really has problems with an excessive number of people suffering from an eating disorder (what a horrible analogy), then it is incumbent on the owner to specify a clear limit to the 'all' in 'all-you-can-eat'.
If my ISP advertises blazing speeds and unlimited access, and I pay for and am denied such service, how is this not false advertising?
Thanks.
1. we're talking about your upload speed, not the other person. With most torrents, there is more than one source (otherwise, torrent is not the optimal application) so the downloader is unharmed. If the source's ISP constrains uploads, then my earlier comment is valid.
2. see 1.
3. networks have always been managed by the operator, because networks have always been shared.
I don't think there is false advertising, but we are bound to disagree – hopefully, respectfully!
I think we're still missing each other on this first point…
If my upload speed is hamstrung, thereby affecting my share ratio, then another user, using a client that I don't get to choose, will limit the speed at which they send me data, thereby affecting my download speed. The downloader is harmed.
Most clients display this behaviour, so my odds of finding a seeder willing to share freely despite my poor share ratio are much reduced. I can download from multiple clients simultaneously, but since most reduce their available bandwidth as a result of my poor share ratio, I remain harmed.
Second, my choice of client application is immaterial to a tracker site. If my share ratio is too low, I cannot join the torrent, no matter which application I am using.
I understand all networks are managed to some extent, but such management has rarely prevented users from being able to utilize expected services, at least not to this extent. If nothing else, a little transparency and predictability isn't too much to ask for, I don't think.
There was a timely post today on Michael Geist's site regarding a submission from BitTorrent to the CRTC. I'm sure you've already read it Mark, but I thought I'd quote it here to put this post to bed:
"Rogers argued that users do not care about upload performances. BitTorrent says this is inaccurate given the correlation between download and upload for BitTorrent users."
We may not agree about the presence of false advertising, but can we at least agree that users do, legitimately, care about their upload performance?
Actually, I wanted to point out my ISP provider (Roger's) interferes with non-P2P traffic as well. (Even if I am not running any P2P protocol) as long as you are consuming bandwidth above an unspecified threshold, Roger's will throttled your service level.
As a digital nomad, I am paying $$$ for top-tier service for the privilege of telecommuting and enjoy digital content at any time, any format. The apparent murky business practice of Canadian Telecom Provider(s) is hurting (i) consumer experience, (ii) Canadian competitiveness.
Traffic Management as suggested by Telecom Provider(s) is very different from Traffic Management Standards(QoS) proposed by Internet Engineering Task Force. Both should not be confused.