June 2007 Archives

Brooks Davis gave a talk at this year's BSDCan conference on building clusters using FreeBSD. I couldn't attend, but the title of this talk caught my eye; he links to his 2003 talk.

One thing that caught my eye was the potential for a cluster's intercommunication to affect the rest of the network to which it is attached. Considering we're looking at building a small cluster on a Blade Centre, this might be something to keep in mind. He has some other notes on naming conventions (particularly FQDNs) and the effect they could have on other pieces of software as well. Interestingly, while we're looking at solutions for NAS type devices that don't include NetApp, he's talking about moving *to* a NetApp - granted, this talk is now 4 years old so maybe he'd do something else instead. (Our Sun clusters use a Sun 5210 NAS, which exports its shares via NFS.)

We've run into the same issues he describes with PXE booting, although mostly on commodity class workstation motherboards. Our Sun PC hardware has been very good at this, and we use PXE booting to re-image the workstations we use for our grads - those all run on Asus motherboards of one flavour or another, so BIOS support has gotten better since.

I hadn't known that N1GE could run on FreeBSD, that's of potential interest to me.

I feel their pain with regards to remote console access; as they say, it's a stone cold absolute must when diagnosing issues, particularly with clusters. I think the only real recourse is to bite the bullet and purchase hardware that's intended to run as such; our SunFire X4200s in our 3 Sun clusters are very nice this way. I don't think the LOMs on the 2200s are anywhere near as nice though, and a 4200 is very much a Cadillac solution.

All in all, an interesting read for anybody interested in designing their own HPC cluster. There's a lot more work involved in setting up this sort of approach (vs, say, Rocks clusters) but in the end, the system may be more manageable.

Probably common knowledge to most people who need to admin OS X machines a lot in anger, but it was new to me. Best way to set one's hostname (short of editing /etc/hostconfig):

sudo scutil --set HostName servername.example.com

One of the various Suns brings us news that Dick Tarnstrom may be back in Edmonton.

Good news for the Oilers if true (even if this isn't the first time I've heard this this summer), but the money quote is here:

If he signs, the Oilers blueline corps will include Jason Smith, Steve Staios, Matt Greene, Ladislav Smid, Tom Gilbert, Jan Hejda, Tarnstrom and another veteran puck mover acquired via trade or free agency.

Yes, I know Kevin Lowe has been quoted often as saying he intends to pursue such a player this summer. Of course, we all knew he needed to last summer and all last season, but it never happened. I'm not holding my breath.

Nevertheless, good news if it's true. Tarnstrom isn't Pronger, or even Spacek, but he's a vet and performed well in his brief tenure as an Oiler.

Seth Godin's post today on reasons and excuses for companies reminded me about something I've been thinking about for a while.

The terms reason and excuse are frequently used interchangeably, and the term "excuse" often has negative connotations to boot: "that's just an excuse." I think that's wrong. An excuse should excuse you - it may also be a reason (and likely is), but the perception of it as a negative thing ought to be incorrect. A reason is just an explanation for why something happened.

This may be prescriptive grammar-GodwinsLaw-ing, but I think there's some merit to the approach. There ought to be a distinction between a reason and an excuse, and I don't believe that the English language as currently used really allows for that - which is ridiculous, given how many words we have.

(Today's probably useless entry brought to you by a long put-off dist-upgrade on a group server from Ubuntu 5.10 to 6.06. Boring.)

After reading some comments and posting to a month-old thread on another blog, I got to thinking (for the nth time) about the practise of "necroposting." This is posting to old threads, where the definition of "old" varies. I've seen it as long as a couple of weeks, but generally it's a week - singular - or even days. I've seen ridicule heaped upon the heads of those who dare to respond to a thread that's been dead for a week, and I've personally been asked why I thought responding to a 3 month old post was appropriate.

I can understand why some bloggers close comments on old posts - I get a lot of emails as a result of spammers trying to post to a couple of my old posts in particular. I don't appreciate it though, and I tend not to comment at all on sites that have such a policy. (In fact, I tend to avoid them altogether.)

I believe that the rationale behind criticizing necroposting goes something like this: "that discussion was already had and it's over with." Follow-ons may be of the form "beating a dead horse" or "we're no longer interested." While that has some merit on topics that are only of interest in a particular time period - say, current events, or 'what should Kevin Lowe do this summer' (obviously that's no longer interesting if we're talking 2006, he's done it already), I don't feel that it ought to extend beyond that.

If a given subject is interesting enough to discuss in the first place, it ought to be interesting enough to continue discussion. The notion that one only "gets" a few days or a week in which to formulate one's thoughts, then to post and debate them, is most charitably described as naive and misguided. (I would go so far as to use the word "stupid" in most applications of the term.)

Needless to say, I deliberately keep comments open on all posts here, and will for the foreseeable future, because I would like to think that by and large, the things I write are not rooted temporally. Obviously discussions about Game 5 of the 2006 Stanley Cup Finals are unlikely to be of any interest at all, but I've seen it happen on other weblogs where people raise interesting points beyond what is directly applicable to the original topic. I suppose at that point it ought to be worthy of another post, but that line is not always easy to judge. I do not subscribe to the view that topics can be eternally closed, and do not like discussing things with people who do.

I read this Ethical Hacker article on rainbow tables a few months ago. It provides a pretty good introduction to what they are, how they work, and gets you started on how to use them.

The key takeaway is in the conclusion (where it belongs) - which I quote in entirety here:

As you can see Rainbow Tables and RainbowCrack are powerful password auditing tools. The best course of action to protect yourself is to not allow the storage and use of LAN Manager (LM) passwords on your network if you don’t absolutely need to and to create and enforce a strong password policy that will force the storage and use of passwords as NTLM and not LM. Additionally, the time to compute and space requirements of complex Rainbow Tables should limit the use of them to only determined attackers or auditors. A strong password policy, strong domain security policy, and keeping up with your patches and updates are your best safeguards against password attacks.

Excellent advice.

(Based on a post from April, but re-written.)

ITSecurity brings us 103 free security applications for the most important platforms today. I haven't tried every application on this list, but the ones I have, are generally worthy of inclusion.

Normally my approach to spyware is "take off and nuke the site from orbit, it's the only way to be sure," but sometimes that's just not possible, or at least not possible right away, and it's always worth spending anywhere between 10 minutes and an hour to see if you can figure out what's up. For that, the trifecta of Windows Defender, Ad-Aware, and Spybot S&D have rarely disappointed.

I'm not personally a big fan of antivirus software - it usually protects you best against threats you already know about, and generally speaking, once you've got virus software on your system you have a bigger problem (how did it get there?). In about 18 years of using personal computers, I've never once gotten a virus that I didn't deliberately give myself, despite engaging in some extremely high risk behaviours as a teenager. (Isn't that what being a teenager is for?) Generally, I run pants-down without av, particularly on non-Windows platforms. Knock on wood, of course. Having said that, not everybody is as careful as I am.

Rootkits, chkrootkit has generally been good to me when investigating things - as much for what it told me wasn't there as anything else.

I'm not sure how they slipped Thunderbird into the list, but they ought to have included Enigmail given that they did.

WinSCP is an excellent file transfer utility for Windows, but they missed Fugu for the Mac. Back to Windows, they included PuTTY, but left out Pageant, which is available from the same site - it's an ssh-agent for Windows. WinSCP works with it too. PuTTYGen (again, same site as PuTTY) can be used to generate keys, or to convert keys generated on other platforms. Again, not sure why Firefox made it into the list, and I'm not sure what they mean by "uses SSL browsing by default" - if you go an http URL, it's not SSL, and Firefox doesn't somehow magically save you from this. I've used SiteAdvisor, it was ok, but generally speaking I just don't go to sites I suspect.

While I agree that knowing what services and hosts are up and down is important to security, I'd have difficulty classifying Nagios as a security application, even with something like a check_nmap plugin. nmap and Nessus are definitely list-worthy though. They also missed Metasploit. Arguably that's an application that's too complex to include a list that generally seems geared towards beginners, but having spent a fair bit of time wrangling Nagios configs, I'm here to tell you that it's no easier, and Nessus and Snort aren't exactly lightweights either. I'm just getting into Snort, but I suspect it's much easier to use when paired with something like Sguil.

HijackThis is an excellent tool for its purpose, although I'd have put it up under spyware detection and not "temporary files."

I've not used Kismet much, but I've used KisMAC a fair bit, and that is an excellent tool - so much better now that it supports passive mode on the Airport Extreme.

In class we're discussing models of describing cognition. One thing that strikes me is that humans seem capable of retaining two beliefs that are inherently contradictory. How do you model (as a for instance) racism in an artificial intelligence? Is this even desireable? If you believe the assertion that most people - even those who are otherwise perfectly rational - possess at heart some base level of an -ism based on race, class, nationality, or some other relatively artificial division, is it possible that in order to create a true artificial intelligence, we would need some way to program these presumably negative biases in. (Indeed, the Turing test may even require it: if the person I'm talking to always exhibits perfect logic and rationality, I do not believe that they are a person. They are either a living saint, or a computer.)

Strictly rules-based systems cannot model this - they're insufficiently flexible. One could train up a neural network, but would that even be sufficient? We don't even know what causes this inherently irrational behaviour in humans, so how can we model it? We can make educated guesses about social influences and perhaps an inherited tribalism that was formerly essential to survival, but those are just theories, and still doesn't help us when we want to code up our Turing-test AI.

What other sorts of inherently irrational behaviours and beliefs might we want to give to an AI?

(I originally wrote this post in March, but just now unearthed it. I'd previously thought I would polish it up, but I think it's ok as is.)

I was going to make this an anti-rant. I was going to make a list of the Oilers players of 2006-07 and speculate about the future, and point out a few things about drafts, both this one and past, in a reasoned manner.

I will stipulate that I found Kevin Lowe's moves last night odd. I honestly didn't think they'd go for Gagner at 5, thinking one offense-first Knight was enough. Plante at 15 is, as Lowetide said, defensible but still a bit weird. I will also admit that I'm no expert on the current draft class, outside of the players meant to go in the first round.

Having said that, the "Oilogosphere" is all awash in cries of "but we needed skill and scoring, not a big defensive defenceman and some guy I've never heard of before!"[1] "Lowe's a freaking idiot for not taking [somebody, right now the current hot two are also Pierre Maguire's favourites, Cherepanov and Esposito]."[2] Yes, it's true, the Oilers last year lacked scoring. It's also true that if Petr Sykora - I think he'll be back - Shawn Horcoff, and Ales Hemsky can't be counted on to at least get 1.25-1.5 times last year's output (which would put those three solidly in the 66-79 point range, not too much to ask), the Oilers are in a lot of trouble anyway. If Raffi Torres and Jarret Stoll don't continue their development and get at least 20-30-50, the Oilers are in a lot of trouble anyway. If they can't consistently get the puck safely out of their own end, they're in a lot of trouble anyway. Cherepanov won't help with that, and neither would Esposito. That would require airlifting in at least 5 crack veteran players, which just plain isn't going to happen. Does anybody seriously think that a player such as the young men I've just mentioned would put the Oilers back into Cup contention? If so, you've been smoking more illegal substances than you've accused Kevin Lowe of having done.

Will Esposito fit in with the Penguins? No doubt, they've done pretty well with their young talent - Crosby, J. Staal, Malkin, Whitney, M-A Fleury. I was hoping for the Penguins in last year's playoffs, I'll admit - I even said they reminded me of the next best thing to the 1981 Oilers. But they're not the 1981 Oilers, and this isn't the NHL of the early 1980s. Those Oilers smoked one of the best teams in hockey three games straight, and it wasn't even close. They lost in the second round in a well-fought series against the team that would eventually go on to win the second of 4 Stanley Cups in a row. The Penguins these past playoffs? Not so much. No doubt, they'll do some damage in the future, but I have great difficulty seeing them winning it all at any point in the next couple of years, not the least because Fleury, budding young star that he is, is no Grant Fuhr, and the Pens don't have a young Andy Moog either. They don't even have a combination approaching the 1-2 of Billy Smith and Kelly Hrudey.

Fine, fine, so what, get to the point. What is this all in aid of saying? Look at the past drafts. 1999, for instance. Atlanta took Patrik Stefan at #1OA, and while Stefan's had a decent career, he didn't cover his draft position. Vancouver did all right with the Sedins, trading up to ensure they could draft both of them, but it's only been last year and this year that they've really come into their own. (And now they're only a few years from potential UFA.) #4? A young Czech, scored 245 points in his last two years of junior, tons of offensive potential, drafted by the New York Rangers (who are now getting tons of positive press for taking a chance on a young Russian). Pavel Brendl. Oops. Those of us who remember Jason Bonsignore can feel the pain.

Next up was another highly touted junior player, Tim Connolly. Again, a useful, even good player, but he's not come close to covering his draft number either, and now his career could be over just when it was finally getting interesting. 7th that year was Kris Beech, the guy Mario Lemieux wanted to play with, came over to Pittsburgh as part of the Jaromir Jagr trade. That's two blown calls - Washington for taking him that high, and Pittsburgh's for getting him at all. He scored 10-15-25 his first year as a Penguin, playing with the best player in hockey at the time.[3]

Get my drift now? The draft really is a lottery. Even picking first overall is no guarantee of quality - while you're more likely to get Mario Lemieux than Alexandre Daigle, you're just as likely to get Patrik Stefan. Even if you do draft a player, chances are excellent he'll be a Sedin or a Taylor Pyatt, not a Zach Parise or Jordan Staal. You might get the double whammy, a guy like poor Alexei Semenov or Doug Lynch, players who do well for just long enough to tease you, then disappear into the ether, er, the ECHL.

Drafting for needs now? Fool's game. Cherepanov or Esposito might well turn out to be the next Alexander Ovechkin and Sidney Crosby, but they're much more likely to be Michael Frolik and Alexandre Picard - first rounders with plenty of promise and undeniable skill, but still very much works in progress. Good news, but not if you want to get back into contention in time to make a run in 2008. If you think anything different, I have no problems calling you a naive fool, right to your face.

In Lowe I Trust? Not quite, but do I really have a choice? I could cheer for the Penguins, I suppose, but even Great Ones aren't quite perfect.

[1] First of all, right off the bat: I'm not a member of the Oilers organization. Neither is anybody else that frequents these weblogs, as far as I'm aware. "We" don't need anything except our paycheques and the occasional beer. No matter how much we might pretend or like to think that we're integral members of the Oilers, somehow smarter and more specialer than Kevin Lowe, Kevin Prendergast, et al, we're not. We would perhaps like to see more whatever in the organisation for which we allegedly cheer, but *we* don't *need* bupkis. Unless you legitimately wear the Oilers uniform, or have an office at Rexall Place or wherever their corporate HQ is, THERE IS NO WE. Got it? Good.

[2] I didn't remember how exactly to spell Cherepanov's name, so I went and looked it up. All you guys who like to talk prospects, you'd sound a *lot* goddamn smarter if you'd only do the same thing. How are we supposed to take you seriously if you can't even spell the guy's name, and don't care enough to take 5 seconds to do it correctly? "You knew who I was talking about" is a lousy excuse. Rob Schremp doesn't get a goal credited for waving the puck around on his stick like it was a magic wand, but we all knew what he meant. Get a dictionary, and keep a couple of windows open for TSN and hockeydb.com. If you can't be bothered to at least spell the kids' names correctly, you'll excuse me if I can't be bothered to take your opinions seriously.

[3] Hell, Jamie Lundmark was another kid that the Oilogosphere was all over Lowe to get several years ago, but his career so far of 71 points in 232 games wasn't worth a late round pick for Lowe to trade, never mind his 9th overall selection.

In what will no likely be a blow to guys like Lain, it sounds like the Predators have traded Scott Hartnell and Kimmo Timmonen to the Flyers; in return they get back the same first-rounder they gave up to get Forsberg. Flyers win that trade, I think, unless one of those players was just not going to re-sign with the Preds - the Flyers have a lot more cap room to play with.

The Senators fired Muckler. Rumour has it it's because he couldn't get Gary Roberts and instead got Saprykin. What a stupid move. Granted, Muckler's not getting any younger (he's 76), but it's not like he's done a terrible job with the team, and I'm in his camp: if the owner says "do this at any cost," I do what I think is best for the team long-term, and if he doesn't like it, well, I'll take my buyout. And if I'm a team owner, I let my GM do his job - that's what I hired him for. (And if I decide to acquire a particular guy at any cost because I think he'll help my team win the Cup this year instead of just going to Game 5 of the finals, it's *not* Gary Roberts. It's Ryan Smyth.)

Error messages like this:

(II) NV(0): Mode "1600x1200" is larger than BIOS programmed panel size of 1280 x
1024. Removing.

can probably be solved by installing nvidia-glx and switching to the nvidia driver. Or by installing whatever gets you the nv driver, and using that one. (Mine was the first case.) The Dell doesn't have any place in the BIOS to program a panel size.

I also had:

Option "ExactModeTimingsDI" "TRUE"
Option "UseEdidDpi" "FALSE"
Option "ModeValidation" "NoEdidDFPMaxSizeCheck,NoVesaModes"

in /etc/X11/xorg.conf, but I don't know if those did any good. Got the hints from ubuntuforums.com.

Martin McKeay writes:

We're never completely secure, there are too many aspects of our enterprise to cover to be certain we've taken everything into account. There's always going to be a vulnerability somewhere, it's a fact of life.

We can't afford to have too many illusions as security professionals. We have to see our network for what it is so that we can take the appropriate steps to safeguard our resources. We often have to do the best we can with the resources at hand, because proving the threat/risk to cost ratio to management is difficult in the best of situations. We have to hope that what we've done is good enough to keep the bad guys out of our network.

I like Martin, as much as somebody can like somebody else that they've never met, and his opinion usually matches mine enough that I consider him to be right-minded (/tongueincheek) but in this case I think he's totally incorrect.

Security professionals cannot afford *any* illusions about their work. It's true that there's no such thing as perfect security, but telling yourself lies to allow yourself to sleep at night is exactly what it sounds like: self-delusion, and it's dangerous. One should always be perfectly aware of what the problems are surrounding one's network, and one should obviously strive to eliminate or negate those problems. "Good enough" is not static; the unspoken words afterwards are always and forever "for now."

Yes, we have to have illusions to live our lives, but those illusions have no place in network and computer security. If being able to honestly tell yourself "I did the best I could" is insufficient to generally allow you to sleep at night, you need to find a new line of work.

Something I've been pondering for a long time: Distributed John. Now my plans to dominate the school and eventually the world are that much closer to fruition, and I am yet again saved having to figure out this whole "programming for myself" thing. An MPI-aware version would be better I think (at least for my porpoises) but this will do for now. I think some sort of message passing interface would solve the problem of individual nodes having to be the same specs as well.

(Update: 02 October, should have been earlier - there's an MPI john.)

The Wall Street Journal Online has an article on social computing. Studying social computing isn't a new phenomenon (which they admit); I remember papers coming out about MUDs in the mid-90s and at least one person did a thesis on the subject - Masters, I think.

One bit that jumped out at me was this:

As part of the project, researchers will look at archived versions of Web sites to gauge, for example, whether blogs have contributed to polarized political opinions in recent years. "These digital records leave unprecedented opportunities for social scientists to study human interaction," he said.

While the latter is indisputable, I think it's dangerous to assume that what happens on blogs is necessarily the way humans interact. It's a great way to study *online* human interaction, true, but everybody online knows at least one or two people who are normally reasonable people that turn into raving lunatics when they're behind a keyboard.

Furthermore, I'm not sure that political opinions have become polarized only in recent years. The first example that sprang to my mind was Anne of Green Gables - at one point in the first book, I think, Anne asks somebody if they're a Tory or a Whig. (See, fiction can be a good insight into human interactions too.) Even now in Canada, we have 3 "official" parties (NDP, Liberal, Conservative), and at least one other that seems to be up and coming (Green), but the NDP and Greens are currently fringe. The question is essentially still "Are you a Whig or Tory?" One could even lump them together: Liberal + NDP = Whig, Conservative + Green = Tory. Granted, I've not studied politics much, beyond "what everybody knows," so maybe I'm wrong - maybe there was some period post-Anne and pre-now that could have been considered to be de-polarized, but I really doubt it.

If there was any such time, at least in American politics, I would expect that it was in the mid-90s, post-Soviet collapse and before, say, 2001. I would also consider that it was more likely that that period was something like The Phoney War of 1939/40. The polarization was still there, just unacknowledged. Giving blogs credit (or blame) for the apparent recent polarization of (North) American politics is like blaming somebody else for how you're feeling.

Seth Godin writes so succinctly that it's difficult to quote any of him without quoting all of him. However, in a recent post about the skills required in a modern workplace, he writes: "And yet, we're still trying to hire people who have shown an ability to follow instructions."

I agree that the emphasis in jobs that require at least a bit of creativity[0] ought to be on people who can get stuff done on their own by visualizing the outcome, but that's not really the same as somebody who always colours outside the lines. If your employees are more interested in colouring outside the lines than they are in following directions, they will drive you insane, this I guarantee you. What you need is somebody who can do as Seth describes, but they also need to be able to recognise that sometimes, you just need to buckle down and turn the screws on a project to get it done.

0 - And there's more than you think; try stifling all your creativity at your current job and see how happy you are about it.

Science Daily had a writeup about a study which suggests that forgetting things is important to how human memory works. I believe it's possible that we will one day be able to create an artificial intelligence with a perfect memory, but I do not believe that we would be able to consider that AI to be human-like if its memory does not behave as ours does. We will at least require an associative memory, and I believe (with no real reason for believing such yet) that it will be necessary for it to be able to forget things as well. Memories fading and disappearing appears to be part of the human condition; it will be necessary for a human-like AI as well. (Will it be necessary for the AI to be 'mortal'?)

Last night a friend in IRC says "Cool, new hi-res maps of KW," so I grabbed the latest version of Google Earth and fired it up to have a look. Neat, there's my building, and I can tell whose cars are in the parking lot because one of our neighbours drives a large dark car and couldn't park straight with a gun held to her head. (Our car was not home, nor was it on campus at WLU. Hm, Linda, where were you?) I can tell when the photos were taken - late last summer or early last fall. Let's look at the campus I work at now - yup, pretty nice.

Wait a tic, that's not "Library at University of Waterloo," that's Chemistry. That's not "Math and Computer Sciences Building," that's the Davis Centre, and there is no M&CS building. There's a Math and Computer building though, but that's not what they've labelled and that's not its name.

So now I'm looking to see how to correct marker placements and labels, but it's not easy. Google itself returns lots of people grumbling about inaccuracies in community markers. sigh.

(Update: looks like one can comment at panoramio, which is where the photos appear to be pulled from. Yay!)

Now I've read the news, and wowzers. Yashin's been bought out, but I don't expect it will be long before he has a new home somewhere else. Buyout makes sense from an Isles POV.

Could he fit into the Oilers? Why not, they're bleeding everywhere; if the price is right, sign him. Who cares if he disappears in the playoffs, they'll be hard-put to make 'em anyway. The Isles are surely going to be gunning for Smyth now. Can he play for MacT? I doubt he's as bad as he's sometimes been made out to be. Hey, playing the wing is easier than C, and even if the Oilers pull off a double-deal and get Smyth back and sign Yashin, they'd have their #1-2 LW without needing to rely on streaky Torres.

What I said about interesting about 10 minutes ago in the previous post just got even moreso.

(And hey, I get to beat all the Leafs beat writers, who are likely so busy chortling over Ottawa choking again that they're not going to write about how the Leafs should go after Yashin just like they should go after every single UFA with a heartbeat until they wake up - hungover as hell, I hope - tomorrow morning.)

Apparently there was a hockey game tonight. (I even watched the third period; I likely would have watched the other two but my stepdaughter had a birthday party tonight and it would have been pretty antisocial given that most of the attendees were adults.)

As an Oilers fan, I suppose I should be frothing that "FCP", "CFP", whatever - Chris Pronger - won the Cup. I don't actually care, really; in fact, Comrie badmouthed both the city and team on his way out, which in my mind makes him worse than Pronger. Not that I really care much once a guy's been gone for a year. Once you get used to Gretzky, Messier, Kurri, and yes, even Lowe and Anderson leaving, even a player of Pronger's stature is almost ho-hum. So yeah, last summer I felt betrayed, now I don't even care. Hell, I'm even happy for the guy - two Finals appearances in a row, he won the Cup, and he's not in an intense media spotlight to boot, just what he wanted. And they beat the Wings, and Chris Chelios that smirking bastard I've hated for a *long* time. Pronger's a great player, even if he is allegedly a jerk and obviously somewhat domineered by his wife, and he did some fucking *outstanding* things last year, all the while battling whatever home problems he was having. Last year should have been his, not this one. There's not a single pair of players on the Oilers roster that I wouldn't trade to get him back, although I wouldn't part with both of Horcoff and Hemsky. Throw in any one of the blueliners and you've got a deal, Mr. Burke.

Niedermayer the better won the Conn Smythe - well, not surprised, he's the captain and didn't play badly, plus he's an old dude and won't be in the running for too many more, so he gets the sympathy vote. Pronger better get the Norris though.

I don't have much to say about the game itself, or previous ones - I watched the last two in a haze of bacterial-induced fever - save that Anaheim was the better team nearly full-time, and the third period tonight they simply dominated the Sens.

Was Ottawa Canada's team? I don't think so, and I don't mean just cos a lot of Leafs fans who were Oilers bandwagoners last year were conflicted this year and so chose not to choose. That Cup is going to be spending more time in Canada now this summer than if Ottawa had won. That's Canadian enough for me, thanks. I don't care if most Americans don't care about my favourite sport - the less the better, in fact. The fact the schedule was shifted to suit NBC's wishes and they couldn't even bother to run over for OT speaks volumes, as did the viewership in the Finals this year and last. Sorry Wade, but you'd have better hockey options now than I did as an Oilers fan in Nova Scotia in the mid-1980s, that's for sure, where if it was the playoffs or the Leafs, their games just didn't get shown. No intertubes to keep up either.

So yeah, I'm glad it's over, because now all eyes are on the draft, and it's shaping up to be an exciting one (at least in the fabled Chinese sense) for the Oilers. It's been quite the ride the last few years, and I hope it continues.