hhw

There's far more room for improvement when per capita is high than there is when per capita is low.

And the climate certainly does care about cumulative CO2.

Would have, could have, should have. But didn't. Trying to hold others accountable without getting one's own house in order just comes off as the blatant hypocrisy it actually is. Until we do better ourselves, we are in no position to call others out.

Maybe it has to do with the US having double the per capita carbon footprint, and double the cumulative total amount of CO2 of China, and many multiples of India:

https://www.worldometers.info/co2-emissions/co2-emissions-pe...

https://www.carbonbrief.org/analysis-which-countries-are-his....

Equinix doesn't own many of the buildings they have facilities in. Aside from Dallas Infomart, I don't really know of any where they do, but I do know many buildings they have facilities in that they don't own. Digital Realty much more commonly owns carrier hotel buildings.

Network at the wholesale level is always measured in Mbps using 95th percentile, not in data transferred (average sustained, equivalent to 50th percentile because we're talking about 5 minute samples of interface counters over the course of a month). Note I used a small b in Mb. Depending on the variability of traffic patterns, that usually works out to be on average ~200GB* transferred per Mbps of 95th percentile over a period of a month. Meaning a TB would work out to about $0.35.

*A long, long time ago, I looked at about 1000 co-location customers' MRTG stats and compared their monthly 95th percentile Mbps to their average sustained data transfer in GB, and something like 90% of them were between 150GB-250GB per Mb and 98% of them were between 180GB-220GB. Many people assume 324GB which would require their traffic to be perfectly flatlined throughout the month, which obviously rarely ever happens.

Pretty much any colo or dedicated/bare metal provider. $0.07 to $0.12 per Mb is the going rate for most carriers at any appreciable volume, and even the higher end carriers are less than 3x that.

To be fair, big tech despite their massive volume pay much higher rates than small networks because the carriers charge them enough to fully cover their costs to build out their networKs, while they make all their profits from selling their excess capacity to the little guys for pennies on the dollar.

I think the point he was trying to make is that big tech can use privacy regulations to keep out new competitors, rather than encourage them, as compliance with privacy regulations can create a higher barrier to entry

I think financially literate people in Canada don't pay anywhere near $200. I've always paid around $50/mo for the last 20 years, without ever going out of my way to shop for promotions, and have been able to have all the data usage I could reasonably use (previously excluded watching videos when not on Wi-Fi, but the most recent plan of a few years has enough data that I no longer need to), plus extra bells and whistles like free roaming in the US. This is with both a corporate plan with Bell the last 6-7 years, and a regular single individual plan with Rogers all the years before that, so wasn't even with a lower cost provider.

There are huge promotions multiple times a year because the competition is fierce between the top providers. When long-term contracts that subsidized phones were still around up to a few years ago, the buyouts to switch providers were so aggressive that you could end up with an extra few hundred dollars in your pocket on top of a new phone every 2 years when switching providers, or staying with the same provider and getting the loyalty/retention department to match offers. Yes, there may be better deals to be found down south sometimes, but not by enough of a margin to deal with cross-border banking, currency conversion, and much worse consumer protection laws for most people.

Yes, additional competition might potentially help drive prices down, but the low ROI on the huge amount of infrastructure required for such a small population might also result in worse economies of scale for all players resulting in the need to cut corners on coverage or service quality to remain competitive.

Also, I'm not convinced coverage is better down south. Anecdotally, I seem to hit way more deadspots driving down I5 through Washington and Oregon than I do on Highway 1 across BC to Alberta despite having much larger swathes of populated areas. I'm also shocked everytime I go to New York and get zero cell signal in every subway station including near Wall Street, when every underground transit station in Vancouver has coverage (admittedly Toronto does not have this though).

Sorry, the website is pretty outdated. We're almost exclusively rolling out AMD EPYC3's these days, and we'd price any of those older configurations much lower than what the website lists them at. Nobody, other than spammers, ever orders through our website (although to be fair, our website may be to blame for that also). We get all of our business through word of mouth, and keep busy enough on that alone, so the website hasn't been a priority.

From our experience, if we price too low we get people who expect the world for bottom dollar. $199 is more the minimum price point at which we’re generally willing to take someone on a customer, than a reflection of the price of a base configuration server. Anyone e-mailing us for a quote, if they seem like they’re on the up and up and we like what they're about, we will usually give a pretty good discount. Most business nowadays are for people ordering several servers at a time and they will always request a custom quotation anyhow, and we're pretty aggressive with larger volume orders.

(Disclaimer) As a bare metal provider, I hope more people become aware what I've been saying for years: cloud is great for scaling down, but not that great for scaling up. If you want to have lots of VM's that don't warrant their own hardware that you can conveniently turn up and down, then cloud is fantastic. If you have a big application that needs to scale, you can get further vertically with bare metal, and if you need to scale horizontally, you need to optimize better higher up in the stack anyway, and the much lower cost for equivalent resources (without even taking any virtualization performance hit into account), more flexibility and thus more/better fitted performance of bare metal should have the clear advantage.

No matter what, SK Telecom is going to be eyeball (ingress) heavy and no amount of engineering of Netflix traffic is going to help with that, as it's all inbound anyway. Not that Netflix's upstreams like Level3 or Telia would be peering with SK Telecom anyway. Reducing the amount of Netflix traffic can only help.

1) It's impractical to test everyone for antibodies. There's enough empirical data to suggest that immunity wanes over time for most people that it makes sense to just give everyone a 3rd, booster shot. 2) It was always expected that vaccine immunity would wane over time. Nobody (of any relevant repute) ever suggested otherwise. 3) I don't doubt that you had side effects from the vaccine. But there's a strong probability that the side effects of an actual infection would have been much, much worse for you. You may potentially have an undiagnosed health issue that could have potentially been a comorbodity in an actual infection.

Why focus on China and India, when on a per capita basis they have a small fraction of the carbon footprint that we have in Canada and the US? When we can get our per capita carbon footprint down below theirs, then maybe we'll have a right to criticize their environmental policies. But in the meantime if we're being at all objective, trying to point a finger at them makes us look like absolute hypocrites.

Not all servers are web facing or live in data centres. There are tons of small businesses that run servers in a backroom. Those are predominantly Windows.

This is not imperative to do in-house. My company (*shameless plug for Astute Internet) has expertise in those areas you describe, and will provide and/or support the edge network infrastructure for small ISP's in exchange for selling them some of those services (we're a mini-carrier and value added reseller) and charging a nominal retainer to be an available path of escalation 24/7. Due to our purchasing volume and industry knowledge, we are able to sell these services for the same or less than what our customers would be able to negotiate on their own, while getting a lot of support from us in the process.

Although this type of knowledge and expertise is rather specialized, we are able to focus on these specific areas without getting involved in the intricacies of the downstream side of our clients' networks, thus avoiding the need for us to have too much client specific or institutional knowledge. This has worked out very well with the handful of eyeball networks we're currently working with in the Pacific Northwest, and we intend to more actively extend this line of business down the coast in the near future.

I've never heard of any data centre burning down (and I work in this industry), so never hearing of an AWS DC burning down isn't really saying anything about AWS.

Customer service in Japan is hands down superior to in the US.

Maybe other factors like culture matter more than tipping?

The side effects have also been known for a long time, which is why dosage is carefully monitored in Lupus patients and blindness is an anticipated side effect.

My understanding is that another side effect of HCQ is that it extends the QT cycle, and given the duress COVID-19 puts on the heart, there is not an insignificant risk of cardiak arrest. There may not be as much data to show this as conclusively or prevalently as some might like, but the underlying mechanisms are well enough understood that it would be reckless to continue prescribing HCQ until it had been sufficiently demonstrated that the benefits outweighed this risk.

There really wasn't sufficient reason to think that HCQ would work in the first place, based on any understanding of how the drugs works. It was just the original, now discredited paper from China that opened up the floodgates in the first place. The antiviral properties are far too weak and would require far too high (dangerous) concentrations in vivo, far in excess of what's commonly prescribed for other purposes, to match the earlier in vitro results. If there's any benefit to HCQ, it would be from immunosupression. But dexamethasone serves that purpose much more effectively, and has shown much better results so it doesn't make much sense to continue exploring HCQ.

My claim was that the site pushes a blatantly false narrative, not that science is a popularity contest.

This paper you've cited supposedly showing positive results is a preprint and has not been published yet, which suggests it hasn't satisfied the peer review process in whichever journal it is intended to be published in. The study was also suspended partway through so it's unlikely it'll ever be completed. I don't entirely disagree that science is a pure numbers game (only papers that stand up to peer review, not only from the journal but from the wider scientific community are worth consideration), but you're undermining your own argument with this poor selection of a paper. If it does eventually end up published, and doesn't have too many glaring issues with its methodology pointed out (some of which are already indicated in the comments on medrxiv), it may be worth mentioning in the future but it certainly isn't right now. And the existence of this paper doesn't make the site's analysis any less dishonest. Suggesting 100% of papers of some particularly category when only one (in actuality none) exist is clearly wrong, and intentionally misleading if not dishonest.

I also don't entirely agree that numbers don't matter. Given a set of papers that are published, properly peer reviewed, and don't have any glaring issues with their methodology, it certainly matters if the vast majority demonstrate something while a tiny number suggest something else. This would indicate that tiny number are outliers, and may have had some problems that weren't immediately obvious. Otherwise, their results would have been reproducable which is a key indication of the validity of their empirical data.

There are also other, better studies that show negative results or no benefit. Why did you pick this particular one?

That site appears to be making some obviously false claims in its analysis. For instance, tt makes this claim: "100% of Randomized Controlled Trials (RCTs) for early, PrEP, or PEP treatment report positive effects, the probability of this happening for an ineffective treatment is 0.002."

Note that the only RCT with positive results is https://www.medrxiv.org/content/10.1101/2020.03.22.20040758v...

Which was late, not early. Meanwhile, quickly searching through all the RCT's listed, there's 5 that are negative and 10 that were inconclusive (i.e. showed no benefit).

So there's 15 to 1 against the effectiveness of HCQ using the highest standard of study (RCT's), yet somehow that site attempts to represent that as 100% of 'early' RCT's showing positive results.

The New York Post doesn't have a great track record on being factually accurate though, whereas the Washington Post can reasonably assume articles published by the New York Times have undergone some reasonable standard of fact checking (even if imperfect), so it's not a reasonable comparison.

This is a much better and thorough critique than I could write up:

https://respectfulinsolence.com/2020/04/03/zelenko-smith-aba...

https://www.nytimes.com/2020/05/01/us/coronavirus-doctor-zel...

Doubts about the efficacy of convalescent plasma already started popping up before Trump started advocating for it, with many papers being published in June and July.

Zelenco has been thoroughly disgraced for ethics violations. His work was so riddled with issues, that it was the subject of ridicule for weeks afterwards by legitimate scientists. It's not worthy of discussion in a good faith discussion on the efficacy of HCQ.

In vitro means a petri dish. Bleach has antiviral properties in those conditions.

In vitro results are much less meaningful than in vivo results. It can be a useful early validation step, but is far from telling us anything about real world efficacy.

There's enough in vivo studies at this point that in vitro studies should no longer be part of the conversation.

Because with presumably feasible level of efforts and resources, you can make a claim that's not yet true, but has a reasonable probability of becoming true by the time you need to deliver your product/service. So you can make that claim in good faith, even if you're not 100% certain it will hold true.

That's very different than making a specific claim that you already have a feature right now, that you in fact don't. That claim cannot possibly be made in good faith, as it's currently outright false, and you can never retroactively apply end to end encryption on conversations that have already happened.

A philibuster is nowhere near the same level of obstruction as refusing to call a vote at all.

It also hasn't been used to anywhere near the same frequency.

Correct me if I'm mistaken, but isn't Brazil currently losing huge tracts of protected rainforest in those areas, just because of Bolsonaro's politics? How is he managing to undermine these protections so totally?

https://www.nytimes.com/2019/07/28/world/americas/brazil-def...

Which smaller ISP in the US is able to resell lines from Comcast, TW, AT&T, and Verizon with nationwide coverage?

I'm not aware of the US mandating lines be resold at controlled costs dictated by the FCC, but I don't deal with consumer/eyeball networks so perhaps I'm missing something? From the frequent complaints I see about duopolies in most US networks, I'm quite skeptical of this.

In metro Vancouver, I'm able to purchase 1G VPLS fiber links for around $800/mo, and dark fiber in the $1000-$2000/mo range for the most part, with similar pricing in Calgary and Toronto. Quotes I've seen in Seattle and Los Angeles have been typically 3-6x that range.

I can purchase municipal dark fiber in Coquitlam or New Westminster (suburbs of Vancouver) for $400/mo, and I believe Montreal is $500/mo. I'm not aware of being able to buy metro dark fiber for anywhere near that anywhere in the US.