You can have any answer you want except the one you’re looking for
[Update… OMG. I just finished this post and boom! – I got the CRTC’s letter to Jason and the CGO, released late this afternoon. After all the blood, sweat and tears, the Commission has the nerve to ask the CGO to provide “more specific information” relating to their recent complaint – whereas the Rogers reply says absolutely nothing about any of the technical details. And just to rub a fistful of salt into the wound, the Commission is insisting on getting the CGO reply on Thanksgiving – because 10 days is now the required response time! See below for the many problems with the Rogers letter. And stay tuned.]
Let’s review the bidding.
On Sept 16, i.e. before the new complaints guidelines were released, the Commission sent a kind of warning letter to Rogers. This letter (pdf) may look like the regulator’s getting tough – except that we’re now less than three weeks shy of the 2nd anniversary of Canada’s network neutrality rules (Telecom Regulatory Policy CRTC 2009-657 – pdf). And a lot of consumer-unfriendly things have happened in the interim (notwithstanding the updates to the guidelines issued by the Commission under the heading “How to make a complaint about Internet service”).
As for the Commission’s brief letter (barely 350 words), the meat of the matter is in the last couple of paragraphs:
“Commission staff requests that Rogers file a plan for resolving the possibility of misclassification of other interactive game traffic, by 27 September 2011, that includes specific steps and timelines for each step.
“Commission staff also requests that Rogers provide a detailed report to the Commission once the problem is resolved, demonstrating that the problem has been fixed. This report should include, at a minimum, the following information:
• an overview of the solution implemented,
• an explanation of how the solution resolves the underlying problem,
• a description of each step taken, and
• a description of the changes made to Rogers’ ITMP disclosures in order to accurately reflect resolution of this problem.”
These instructions don’t seem all that open to interpretation. Rogers is asked to file a plan that incorporates steps and timelines. Unless you believe in some kind of planning other than “forward” planning, that would be a look into the future – to explain what the ISP is going to do to make its ITMP problems go away. The request is also clearly based on the surmise a) that a misclassification problem exists today, and b) that there is a reasonable likelihood similar problems could arise at some later date.
The second half of the request concerns what Rogers is to do after the plan is put into effect: Fix what’s broken and report back on exactly what you did to accomplish the fix. The letter also asks for “a description of the changes made to Rogers’ ITMP disclosures.” The status quo is not an option; we need to see a new approach.
Here’s the setup for Rogers’ response (p.2):
“Rogers is pleased address these issues. In doing so, we believe it would be helpful to provide the Commission with some background on our ITMPs before turning to why, in rare cases, some customers have been impacted when playing certain online games and the measures we have put in place to resolve those issues.”
You don’t have to go too deep into the sub-text to see that, in its chipper and deferential way, the letter is saying, All due respect, but we have no ITMP problem. If we did, that was in the past, water over the dam, and we’re also reminding you the problems were rare, and now that they’re resolved, let’s not lose any more sleep over it.
The reader is offered a number of variations on the theme of “not guilty.” They include predictable platitudes such as, “Rogers implements ITMPs to ensure the best experience for all of our customers.” Well, the company actually does so to ensure the best experience for all of its shareholders, but let’s not split hairs.
Equally predictable, the storyline is quick to take refuge in Big Media’s favorite defensive manoeuvre – demonizing “downloaders,” the customers who initiate peer-to-peer transfers and are presumed guilty of both bandwidth hoggery and copyright infringement. Rogers doesn’t call it demonizing, however, they call it “background”…
Our network is not designed for machine to machine uploading of content from a customer’s home such as is the case with P2P file sharing uploading.
So before we’ve got past even two paragraphs, the defendants have cleverly turned the spotlight on the people who should really be taking the rap: a gang of anti-social, trouble-making customers. As you can see, dear reader, the “problem” really has nothing to do with the way we run our business or even with our Cisco gear (though Cisco is certainly up for second place in the lineup of suspects).
Next question. Why is all this space devoted to explaining why Rogers uses ITMPs and why it targets P2P? Why is Rogers preaching to the choir – i.e. the very Commission staff who likely had a hand in devising the ITMP rules in the first place? The manifest purpose of the explaining is to support a resolution of the gaming problem; but I’d say its latent purpose is to pass the buck back to the customer – a sentiment this Commission will instinctively understand and appreciate, given that their guidelines do exactly the same thing. And btw, the ITMP rules state that the preferred option for resolving traffic issues isn’t to “manage” customer bandwidth, but to expand it through investment: “Network investment is a fundamental tool for dealing with network congestion and should continue to be the primary solution that ISPs use…” In this case, however, neither the Commission in its request nor Rogers in its reply makes even a mention of network investment as part of the solution.
Demonizing P2P: the cost-benefit analysis
For years, the established players have blamed North America’s shitty broadband on bandwidth hogs, whose suspicious activities have been used to explain why the incumbents have crippling congestion in their local access cloud, and thus why the world needs more ITMPs to manage it. Those players have included the outgoing chair of the CRTC, who last I heard would simply not give up his specious theory that light broadband users unfairly “cross-subsidize” heavy users (maybe the PMO’s power-brokers are now looking for a replacement chair who believes in evidence-based policymaking).
Naturally the incumbents have attributed most of their bandwidth problems to the use of P2P and especially Torrent transmissions. Indeed, Rogers claims that all the problems with World of Warcraft, and other games, have only occurred in the presence of an associated P2P transmission:
Rogers has experienced a small number of cases of gaming traffic being misclassified as P2P file sharing traffic. In these cases, gaming customers have only been affected when running P2P file sharing simultaneously with a misclassified game. The typical game requires less than 80 kbps and so would not be affected even if a misclassification were to occur. It is only when the games are running in conjunction with P2P file sharing that our ITMP would be deployed. This has been confirmed by repeated testing in our lab. We have currently resolved all of these cases (pp. 2-3).
Yes, we have no bananas, and, yes, Bob’s your uncle. Everything’s been resolved – except of course ridding the world of P2P and the pirates that ruin the customer experience for everyone else. Before I get back to the main story, let me indicate how the P2P whipping-boy strategy has fouled the air well beyond this fuss over WoW and other games.
When they first turn up, the digital “natives” I teach have no clue about what P2P technology really is. Most of them were about 10 years old when Shawn Fanning created Napster (in 1998), and they usually assume Fanning invented P2P – and that he did so to make the theft of digital IP easier and more fun. So as part of getting them to see the bigger picture, that back-of-the-envelope sketch directly opposite pops onto the screen in the first class or two. It depicts the first 4-node implementation of the Internet’s predecessor, ARPAnet (as in the DoD’s Advanced Research Projects Agency). Four nodes and no central machine like we have in the later client-server model. While this isn’t the place to get into network topologies, the model in the diagram reflects the essence of the Internet paradigm: load-sharing made possible by the equal distribution of resources among all nodes; redundant links to overcome traffic issues like congestion; and no central server acting as a potential point of failure for the whole network. (Btw, those 4 nodes were at UCLA, UCSB, Stanford and Utah.)
Why does the bad rap for P2P matter? Well, neither consumer welfare nor post-secondary learning was ever helped by disinformation of any kind – especially the kind that makes the incumbent ISPs look like Captain America defending light users, music labels and movie studios from the scourge of hoggish downloaders. The bad rap skews the whole public debate on congestion and ITMPs in favor of the status quo. Meanwhile, I’m told by some of my students that Rogers actually sends out cease-and-desist letters to certain subscribers who are heavy downloaders, alleging infringement of third-party copyrights. (I’d like to hear from anyone who has information about this issue.)
Customers blame us because they don’t understand Internet connectivity (doh)
Jason K, Geist and others have been saying that the biggest single flaw in the CRTC’s guidelines is they still put the onus on consumers and consumers aren’t equipped to deal with ITMPs, DPI and misclassification bugs. And in a bizarre twist, that’s exactly what the Rogers letter goes on to say, once it’s done blaming P2P for all our woes:
Based on our testing, we know that some of the complaints that the CRTC has received from gaming customers have blamed unrelated problems on our ITMP. Online gaming problems can be caused by many things unrelated to our ITMP which, however, all cause the same symptoms as misclassification. Unfortunately, customers cannot tell specifically what has caused their problem. It could be, for example, the home pc, the home network, in house wiring, gaming servers, the game itself, Rogers’ network and/or external networks. We encourage customers to contact us so we can help to determine the cause of their problem.
Booyah! Customers can’t tell! That’s our line Rogers is stealing. It’s precisely because customers can’t tell what’s causing technical problems that the CRTC should not have based its whole framework on getting information that complainants are extremely unlikely to have on hand – so says Rogers, which oughta know whereof it speaks.
But wait. The paragraph just cited does encourage customers to contact Rogers. Isn’t this progress? Not in the slightest. As the CGO notes in their press release on Rogers reply (posted Wednesday at OpenMedia.ca), Rogers hasn’t complied with the CRTC order; hasn’t acknowledged a real throttling problem persists; and apparently hasn’t teamed up with Cisco to resolve this issue after all – while the appended flowchart (pasted in up top in this post) simply doesn’t reflect what the CGO membership has seen or experienced.
What gamers have experienced is captured painfully in a posting by a Rogers customer on the Rogers tech support forum. Ressy is apparently tech-savvy and a prolific poster. Ressy’s post dates back to January 2011, just as Jason was filing the first CGO complaint about Rogers with the CRTC…
Attempting to call Tech Support results in 1 of 3 of the following:
1. “Its not our problem, its the game.” It’s not. Other users in Ontario play fine (I can play fine at home in Toronto, but I dc constantly at my sister’s in Ottawa).
2. “We don’t support gaming.” Fine, update your website to show that you can’t game on any of your packages from Rogers Lite and above.
3. “Sorry, I can’t transfer you to a supervisor.” Oh… REALLY???
Ressy points out something else very useful in his post, namely that the all-out ISP war on P2P is very likely interfering with online distribution methods that use or mimic P2P. His case in point is Blizzard’s Repair on Demand feature, a clever way to patch game software on the fly without disrupting play. Except that it disrupts Rogers DPI software, which in turn disrupts whatever that customer is trying to do online.
If that’s just one man’s experience, there’s plenty of other evidence that Rogers is not exactly waiting with open arms for its customers to lodge complaints. And always remember: if the complainant doesn’t get through to Rogers, there is no complaint, and no complaint means no problem and thus no resolution.
You recall the complaints Geist wrested from the Commission under his ATI request in July? In Canada’s Net Neutrality Enforcement Failure, he documents among many other things that 17 of the 36 complaints were directed at Rogers. Take these for example:
• Case CRTC #486625 (March 2010). “Throttling affecting corporate VPN connection… Rogers policy says it does not affect VPN. Contact Rogers. If unsatisfied and you have evidence of rule violation, provide evidence for further consideration.”
• Case CRTC #494836 (July 2010, no link). “Problems with SIP (session initiation protocol) on Port 5060… Admits problem but argues it is not a traffic management issue.”
• Case CRTC #496562 (August 2010). “Claim throttling speeds on xBox usage… Rogers denies the claim.
• Case CRTC #513298 (February 2011). “Throttling of World of Warcraft online game. Leads to regular disconnection… Denied. Rogers says throttling practices have no effect on online gaming (later reverse)
I’ll pass on the other 13 complaints; I assume everyone gets the idea.
A new dawn?
Hard on the heels of Rogers’ not so persuasive response, SVP Ken Engelhart stirred up a full-blown customer love-in for his interview with the Wire Report (Sept 28). And grabbed the story headline:
No one is more interested than we are in providing good service for online gaming, Rogers says.
Continuing in the vein of the reply letter, Ken pointed out that “Our well-being and profitability as a corporation depends on us giving our customers the best possible Internet service we can.” I’m glad we’ve run out of space, because I couldn’t bear to start digging back into Rogers’ price structure, and how their data caps and burn rate manage to screw everybody (especially those disloyal customers who’ve tried to make Netflix a substitute for our fine, made-in-Canada TV walled garden). If you want to check out some analysis of mine from about a year ago, try Why they’ll never tell you about your data cap burn rate (4). Ken bookends the article with a reassuring denial:
“Engelhart said the suggestion that the company seeks to slow down its gaming customers is counter-intuitive and personally frustrating.”
I’m not aware how the question was worded that earned this response. Nevertheless, while there is a distinction between intended and unintended consequences (it can turn murder into manslaughter), I don’t think anybody cares much at this point what Rogers has sought to do through this whole fiasco. What we do care about is that Rogers has fucked up their Internet access service, and failed to own any corporate responsibility for letting it happen.
ps: In part 4, we’ll look at how, in its letter today to the CGO, the CRTC has made this already unacceptable situation far worse for Canadian consumers.