Share good ideas and conversation.   Login, Join Us, or Take a Tour!
veen's comments
veen  ·  5 hours ago  ·  link  ·    ·  parent  ·  post: Pubski: November 22, 2017

Every year Thanksgiving catches me by surprise...or, as the rest of the world sees it, "That Weekend the Internet is Much Quieter than Usual." Now let's hope the Internet doesn't stay that quiet... ;)

veen  ·  5 hours ago  ·  link  ·    ·  parent  ·  post: FCC plans to vote to overturn U.S. net neutrality rules in December

    Cable TV greatly improved access to television content compared to over-the-air broadcast, with more and more channels and packages available year after year.

More and more packages - for more and more money, with the good ones of course spread over multiple packages. Check out that graph! It rises 3 times faster than inflation since 1998.

My argument is that the same cost hiking is bound to happen with internet if the FCC is gonna do this. The airline industry might serve as a good corollary to this. Airline seat pricing is time-dependent and, if the airlines had their way, customer-dependent. Do you know the concept of willingness to pay? It's the bread and butter of airline pricing: each person has a dollar value in their head that represents what they are willing to pay for a service: anything above and they won't buy a ticket. The only goal that shareholders want an airline to pursue is to get every person in every single seat to pay as close to that price as possible.

The most lucrative passengers are people who fly for business reasons, since the cost/benefit calculation is nearly always positive. If airlines could charge you more for traveling as a business-passenger they would, but they're not allowed to directly discriminate like that. But pretty much all business passengers want to be home on Friday or Saturday, so one of the best ways for airlines to figure out if you are a tourist or a businessman is to offer a cheap ticket that has your outbound flight before Saturday night and the inbound after. This is called the saturday-night stay, and while good competition can destroy it, the airline industry in the US has consolidated so much that it is pretty much standard now.

My "nightmare" scenario is that price-practices like this will also be adopted by ISPs. They have your internet history anyways, so they can totally figure out how rich you approximately are. Net neutrality also prevents discrimination between customers, if I understand it correctly.

Infrastructure costs are important, but it's not like cable companies aren't making plenty of money - the problem is that they let the customer pay for that kind of stuff, because capitalism. Also, wouldn't it be an argument for net neutrality if internet penetration is larger than TV cable penetration?

veen  ·  6 hours ago  ·  link  ·    ·  parent  ·  post: Pubski: November 22, 2017

OMG DAT TABLLLE

Man, it totally sucks when you do the friend rate and it just becomes a shitshow like that. I've had some design projects like that go south, but never a "you're infuriatingly insulting" kind of going south. I think the only way forward is to sternly let him know that you will never do that again because you can't save the Titanic by calling in a friend. Maybe there's a slim chance he'll understand where you're coming from...

veen  ·  7 hours ago  ·  link  ·    ·  parent  ·  post: Pubski: November 22, 2017

Tripel Karmeliet is great. Or my memory of it is, I'm not sure... Belgian tripels are not for everyone, but I prefer it over whiskey or scotch any day.

veen  ·  7 hours ago  ·  link  ·    ·  parent  ·  post: Pubski: November 22, 2017

    It's not the final document, but a 90%-done document that determines whether you can graduate or whether you need to put more work into it. That meeting is next week

It went well! My committee thought that the core of my thesis is incredibly good but that I really need to streamline the shit out of my thesis. My goal was to strive for comprehensiveness, but they argued that it is much better to have a concise main story and leave the comprehensive details for the appendices because I was boring them to tears. Fair enough - I'm rewriting most of it anyways because I'm porting it all to LaTeX. I found a neat thesis template that, after some fiddling around, checks all the boxes.

The one downside is that I won't be graduating December 21st like I had planned because one of my thesis councilors will leave the country a few days before that. The next available date is halfway through January, which means that I didn't achieve my goal of graduating this calendar year. A bit of a bummer, but also a blessing in disguise because it gives me the breathing space to polish and perfect my thesis. Like, it gives me the time to have my thesis printed as a softcover 7x10.5" book, circulation of ~8. Partly because it'll make a splash, but mostly because it is cool af. (Especially because LaTeX allows me to format everything in Garamond beautifully.)

----

After some more job conversations I think I finally have a great answer to the age-old "so what do you want to do" question. It's a long answer, but the short version is a crossover between business development & innovation, data science, and urban planning. The great thing is that I can do that at all three companies that I'm in serious talk with. Tomorrow I'm gonna discuss details with my so-far-favorite. (I count my blessings that it's a close race though!)

----

Aldi and Lidl are know for temporarily selling random, cheap stuff in between the rest of the groceries. I found a 10-hole harmonica for like €4. I couldn't resist its sillyness. So if life ever blows I can blow that.

----

edit: I forgot to mention that that Punch Brothers album I posted in the music thread is really growing on me. It's eccentric and creative and I love it.

Devac  ·  58 minutes ago  ·  link  ·  

    Aldi and Lidl are know for temporarily selling random, cheap stuff in between the rest of the groceries. I found a 10-hole harmonica for like €4. I couldn't resist its sillyness. So if life ever blows I can blow that.

If you found those sold in small black boxes, then you should buy it. It's been almost five years since I got mine from Lidl and it still maintains the pitch.

veen  ·  8 hours ago  ·  link  ·    ·  parent  ·  post: Pubski: November 22, 2017

    "My Daughter Bought Bitcoin. It Went Up, Now She Thinks She Is A Genius"

It's from a JP Morgan exec but I'm pretty sure it represents the vast majority. Including myself, until I actually did my homework.

ButterflyEffect  ·  7 hours ago  ·  link  ·  

I typically let the other geniuses tell me when to buy...mk, insomniasexx...a bit facetious, but, you know.

mk  ·  7 hours ago  ·  link  ·  

That was part of Jamie Dimon's market manipulation. Seems that he thinks he's a genius.

veen  ·  1 day ago  ·  link  ·    ·  parent  ·  post: America is now an outlier on driving deaths

I have to admit, part of me also wanted that simple answer. My tiny roadtrip from San Diego to LA and back felt like a long-ass drive. And in that six hours, I got cut off by an asshole in a BMW thrice. But it's not like the rest of the world are saints: the Dutch are known to drive selfishly, rarely making room when you want to merge for example. Germans are nicer but are worse at tailgaiting.

What I did find worrisome was that driving in the U.S. and Canada was way easier (in terms of mental energy needed) than I expected. I did not feel like I needed to pay attention as much as I do over here, like once I'm on the right avenue I just have to follow the guy in front of me and not run a red light, even in dense urban areas. It was tempting to check my phone because driving was legitimately more boring than I was used to. My impression is that distracted driving is a much bigger issue over there than here.

veen  ·  1 day ago  ·  link  ·    ·  parent  ·  post: The GOP's fractal incompetence problem

Trainer's point is also that were it not for Trump, this backlash against harassment would never be as strong as it is. She also worries about what the backlash against #metoo might be, and now I kinda worry about that as well. Let's hope this resetting the bone doesn't break even more.

veen  ·  1 day ago  ·  link  ·    ·  parent  ·  post: America is now an outlier on driving deaths

To put some numbers behind it, the theoretical capacity for a regular roundabout is 20-25k vehicles/day, whereas the above can process up to 40k/day, mostly dependent on where the largest flows come from. (A regular intersection with traffic lights can handle between 20-35k.)

goobster  ·  1 day ago  ·  link  ·  

Ooooh! Numbers! I love data!

Do you have numbers for similar intersections with traffic lights, or stop signs, to compare against the roundabout numbers?

veen  ·  1 day ago  ·  link  ·    ·  parent  ·  post: FCC plans to vote to overturn U.S. net neutrality rules in December

Alright, let's talk plans. I consider them not-bad partly because of my frame of reference. Another thing is that they are far removed from my imagined 'worst nightmare' non-neutrality situation. I'm sure you've encountered this argument before, but I'm gonna bring it up anyway: what if internet is going to go the way of cable tv?

In this imagined world, companies have slowly persuaded people to get bundles for their most popular apps (like the kinda good deals in the Portugal example). Because people don't want to increase their internet fees too much (or at all), they downgrade their base bundle. So most customers then have a few bundles and a small base bundle. Each bundle has a data limit (because fair use).

There's a bunch of things I don't like about that scenario, which I think is likely to happen. Most notably, it is a world that heavily favours incumbents. You're not going to make it with your film streaming site Flatnix if Netflix is already in a popular bundle. The ISPs would probably ask a nice fee from each company that wants to join a bundle, one that will only make it harder for David to beat Goliath. Hubski would never be in a bundle, so that means I can only 'ski with that smaller bundle.

Secondly, with more bundles to keep in check, there will be more overcharge fees, which are probably super lucrative for ISPs. My mom is on social media all the time, and I had to explain to her that she really needs to turn off LTE and turn on wifi whenever she is at home or at work. Her bill almost double her actual plan for multiple months when I found that out.

Neither of those things are better at satisfying customer demand than our current, net neutral world.

    They already provide the very minimum (measured by cost to provide) that they can get away with, before too many customers switch to alternatives.

This is a good point and I don't have a good comeback to it. But as the above kinda shows, if profit is the main motivator (which we both agree on) customer demand is not the driving force. I think that the current local optimum of value per dollar offered to customers is higher than what we would end up with in a world where all ISPs would do a bundle-like thing like I described. I would much prefer ISPs increasing the price for all customers a little, instead of them increasing the price of visiting sites that they don't care about or don't favor.

wasoxygen  ·  1 day ago  ·  link  ·  

    what if internet is going to go the way of cable tv?

Cable TV greatly improved access to television content compared to over-the-air broadcast, with more and more channels and packages available year after year.

The FCC used a light touch, requiring a "basic tier" of local channels and a few extras, but otherwise practicing literal neutrality: leaving decisions about pricing and packaging of additional content between the providers and their customers.

Cable TV companies created a smorgasbord of options to try and appeal to a broad cross-section of customers. Big content creators like Disney negotiated to get less-popular channels bundled together with popular channels like ESPN.

Customers always complain, but wired cable penetration passed 70% by 2000. Canadians objected to the bundling, so the CRTC required "skinny" TV capped at $25 per month, with additional channels available individually. Customers were not impressed with the results.

    "Am I allowed to laugh?" said Gilda Spitz when asked for her reaction to the prices for the new line-up of stand-alone channels offered by Rogers. Most cost $4 or $7 each.

    ...some industry experts are not surprised by the pick and pay prices. That's because, they say, TV providers are for-profit companies, and their main objective is to protect the bottom line.

    "What did you really expect?" says telecom expert Gerry Wall.

Today, only half of U.S. homes are wired for cable, as more and more households opt to cut the cable and stream online. We have pay-per-view, "catch-up" TV, monthly subscription models, near video on demand, push VOD, and all kinds of telecom bundles and tie-ins with mobile and home telephone service.

Even if you only pay for broadband internet, it is impossible to run out of free stuff to watch.

Is this the kind of "nightmare" scenario you are worried about?

veen  ·  5 hours ago  ·  link  ·  

    Cable TV greatly improved access to television content compared to over-the-air broadcast, with more and more channels and packages available year after year.

More and more packages - for more and more money, with the good ones of course spread over multiple packages. Check out that graph! It rises 3 times faster than inflation since 1998.

My argument is that the same cost hiking is bound to happen with internet if the FCC is gonna do this. The airline industry might serve as a good corollary to this. Airline seat pricing is time-dependent and, if the airlines had their way, customer-dependent. Do you know the concept of willingness to pay? It's the bread and butter of airline pricing: each person has a dollar value in their head that represents what they are willing to pay for a service: anything above and they won't buy a ticket. The only goal that shareholders want an airline to pursue is to get every person in every single seat to pay as close to that price as possible.

The most lucrative passengers are people who fly for business reasons, since the cost/benefit calculation is nearly always positive. If airlines could charge you more for traveling as a business-passenger they would, but they're not allowed to directly discriminate like that. But pretty much all business passengers want to be home on Friday or Saturday, so one of the best ways for airlines to figure out if you are a tourist or a businessman is to offer a cheap ticket that has your outbound flight before Saturday night and the inbound after. This is called the saturday-night stay, and while good competition can destroy it, the airline industry in the US has consolidated so much that it is pretty much standard now.

My "nightmare" scenario is that price-practices like this will also be adopted by ISPs. They have your internet history anyways, so they can totally figure out how rich you approximately are. Net neutrality also prevents discrimination between customers, if I understand it correctly.

Infrastructure costs are important, but it's not like cable companies aren't making plenty of money - the problem is that they let the customer pay for that kind of stuff, because capitalism. Also, wouldn't it be an argument for net neutrality if internet penetration is larger than TV cable penetration?

veen  ·  1 day ago  ·  link  ·    ·  parent  ·  post: The GOP's fractal incompetence problem

    I don't think that without "grab 'em by the pussy" that we'd have people like Glenn Thrush and Charlie Rose being suspended.

I found Ezra Klein's conversation with Rebecca Trainer very enlightening. She remarks that the whole awakening we're going through now is not just about unearthing the sexual harassment itself, but probably more about finally seeing repercussions for that harassment. As in, the difference between Chris Rock and Harvey Weinstein is that only one of those two careers is destroyed because of the awful things they've done to women.

What people hoped was that Trump's sexual harassment would have repercussions on his political career and that it didn't says a lot about how far we still have to go.

b_b  ·  1 day ago  ·  link  ·  

What I'm saying is that without Trump, there is no Weinstein. If Trump were to have lost, like we all expected, then we would have patted ourselves on the back and said, "See, the system works! You assault women, you pay." His victory gave rise to the women's march and similar events, which were supposed to be about power and not just Trump. So when another Big Bad Wolf is exposed, but this time a liberal, it's up to liberals to put up or shut up. It looks for now as if we're not making a distinction between political leanings, which should of course be the case. I heard one guy say that the Trump presidency is America resetting its broken bone. This is part of that.

veen  ·  1 day ago  ·  link  ·  

Trainer's point is also that were it not for Trump, this backlash against harassment would never be as strong as it is. She also worries about what the backlash against #metoo might be, and now I kinda worry about that as well. Let's hope this resetting the bone doesn't break even more.

veen  ·  1 day ago  ·  link  ·    ·  parent  ·  post: America is now an outlier on driving deaths

All directions can use the heuristic 'if you go left, take the left lane, if you go right, take the right lane.' It looks harder than it actually is! You're right though, it's a solution that does not always fit the problem.

Dual-lane roundabouts usually create more confusion because people might not change lanes in time, like here:

veen  ·  1 day ago  ·  link  ·    ·  parent  ·  post: America is now an outlier on driving deaths

What the above table shows is that when corrected for sheer volume of driving (the red line) or when corrected for sheer volume of vehicles (the 'per 10k vehicles' stat), the U.S. fatality declines less than other countries. It doesn't say anything about the interaction effect of having both a large volume of vehicles and miles driven. That is your argument, right?

I couldn't find numbers that corrected for both. It's a shame China doesn't provide reliable numbers, since they are also experiencing a very large rise in vehicles and vehicle miles.

kleinbl00  ·  1 day ago  ·  link  ·  

And that second graph I linked shows it pretty clearly - US traffic deaths are clearly not declining at the rate of everyone else which is why the line crosses over Britain and Sweden in like '93 and Japan and France in 2005. My argument hinges around why.

The linked article says exactly what rd95 and goobster want it to say - "because Americans are assholes." I have never not seen that sort of argument be overly-simplistic at best and wrong at worst. The Brits, the French, the Swedes and the Japanese are assholes, too but that's just motive. It's not method, means and opportunity.

Newsweek, of course, is happy to oversimplify the issue. If you get a little more rigorous on it you discover that there isn't even a lot of consistency of how many drivers are tested for drunk driving, which skews the accident statistics. Here's what I know:

I dated a Serbian girl who every year would make a big road trip to the coast with her family. They'd pack up the car, buy snacks, get the maps out and settle in for a mind-blowing three hour drive. In Girl with the Dragon Tattoo, Steig Larssen has his hero travel from Stockholm to the "back of beyond", an island an hour north of Gavle. That makes it two hours by car.

here in the United States, you can drive about a thousand miles a day. Two hours? I have friends that commute that to work every day. and back.

IF: Driving is inherently dangerous

AND: American culture simply involves more driving

THEN: it's entirely possible that the reason our death rate is going down less fast than the rest of the world is we're closer to the asymptote. And I don't think it's fair to wave hands and say "it's because Americans are assholes."

goobster  ·  1 day ago  ·  link  ·  

Psh. FACTS.

Doesn't change the fact that Americans are assholes! ;-)

veen  ·  1 day ago  ·  link  ·  

I have to admit, part of me also wanted that simple answer. My tiny roadtrip from San Diego to LA and back felt like a long-ass drive. And in that six hours, I got cut off by an asshole in a BMW thrice. But it's not like the rest of the world are saints: the Dutch are known to drive selfishly, rarely making room when you want to merge for example. Germans are nicer but are worse at tailgaiting.

What I did find worrisome was that driving in the U.S. and Canada was way easier (in terms of mental energy needed) than I expected. I did not feel like I needed to pay attention as much as I do over here, like once I'm on the right avenue I just have to follow the guy in front of me and not run a red light, even in dense urban areas. It was tempting to check my phone because driving was legitimately more boring than I was used to. My impression is that distracted driving is a much bigger issue over there than here.

veen  ·  1 day ago  ·  link  ·    ·  parent  ·  post: America is now an outlier on driving deaths

Preach! They're not great if you are prone to get car sickness, but they are safer by lowering the speed for everyone and in most low-to-medium-volume intersections they can process more traffic from more directions. What I've heard (as in, anecdotally) is that drivers in the US are not taught what it is and that pretty much nobody knows how to properly yield.

Did you also get to see a turbo-roundabout? Instead of a circle, it consists of two interlocking spirals:

My hometown of 100k already has more than a hundred roundabouts, but added a dozen of those in recent years because they can handle a much larger capacity. It obviously has a larger footprint, since two directions need to pick a lane, but it's pretty darn cool.

goobster  ·  1 day ago  ·  link  ·  

Making a left in that turbo roundabout seems more finicky than it needs to be.

It seems like this is going to be more trouble for anyone that doesn't know exactly where they are going. It also only works where there are 4 potential exits. Add an exit, and the middle lane becomes a wasteland of lost tourists, wandering aimlessly... ;-)

Well ok, maybe not THAT bad! But I think in the UK the majority of the roundabouts had more than 4 exits.

Heck... maybe that's why roundabouts came up in the first place... to make it easier to join multiple roads, rather than just a crossroads...

kleinbl00  ·  1 day ago  ·  link  ·  

I suspect, like roundabouts in general, it's maddening and/or frustrating until you use it and then you go

...oh...

And then you feel silly for getting so worked up about something that's actually really fucking effective.

veen  ·  1 day ago  ·  link  ·  

To put some numbers behind it, the theoretical capacity for a regular roundabout is 20-25k vehicles/day, whereas the above can process up to 40k/day, mostly dependent on where the largest flows come from. (A regular intersection with traffic lights can handle between 20-35k.)

goobster  ·  1 day ago  ·  link  ·  

Ooooh! Numbers! I love data!

Do you have numbers for similar intersections with traffic lights, or stop signs, to compare against the roundabout numbers?

veen  ·  1 day ago  ·  link  ·  

All directions can use the heuristic 'if you go left, take the left lane, if you go right, take the right lane.' It looks harder than it actually is! You're right though, it's a solution that does not always fit the problem.

Dual-lane roundabouts usually create more confusion because people might not change lanes in time, like here:

veen  ·  1 day ago  ·  link  ·    ·  parent  ·  post: America is now an outlier on driving deaths

I did some digging in a textbook that I have on traffic and transport safety. It takes a whole bunch of statistics from the OECT/ITF yearly IRTAD Road Safety Annual Report. It's 568 pages (because it has a chapter detailing each OECD country), but the juicy bits are in the first chapter.

    Looking at the longer-term developments since 2010, the number of road deaths has decreased in all countries with validated data except in the United States, Chile and Sweden. In the United States, fatalities increased by 6.3% between 2010 and 2015 and indications suggest that the situation has not improved in 2016.

The real data I think you're looking for is all in Table 1.3, Road Fatalities per 100k inhabitants, 100 billion vehicle-km and 10k registered vehicles:

If you compare the US with deaths per billion vehicle-kilometers, you can still see the US declining less fast:

          1990    2015

FR 25.9 5.9

SL 65.1 6.7

UK 12.8 3.4

US 12.9 7.0

Sidenote: it puts seatbelt use for U.S. front seat passengers at 91% (similar to other countries) and rear seats at 70% (on the low end), see page 30.

kleinbl00  ·  1 day ago  ·  link  ·  

I appreciate the digging, but my point is that if Russian Roulette has a 1 in 6 chance of killing you every time you play, and Americans play Russian Roulette twice as often as Britons, then twice as many Americans are going to get killed playing Russian Roulette regardless of the safety of the guns they're using. That top graph indicates that Americans play Russian Roulette a lot more than anybody else and regardless of how much they spin the cylinder or how long they pull the trigger, the fact of the matter is, Americans spend a lot longer putting themselves in harm's way... and that while vehicle use in the rest of the world is leveling off, vehicle use in the US is increasing. Check this one out:

veen  ·  1 day ago  ·  link  ·  

What the above table shows is that when corrected for sheer volume of driving (the red line) or when corrected for sheer volume of vehicles (the 'per 10k vehicles' stat), the U.S. fatality declines less than other countries. It doesn't say anything about the interaction effect of having both a large volume of vehicles and miles driven. That is your argument, right?

I couldn't find numbers that corrected for both. It's a shame China doesn't provide reliable numbers, since they are also experiencing a very large rise in vehicles and vehicle miles.

kleinbl00  ·  1 day ago  ·  link  ·  

And that second graph I linked shows it pretty clearly - US traffic deaths are clearly not declining at the rate of everyone else which is why the line crosses over Britain and Sweden in like '93 and Japan and France in 2005. My argument hinges around why.

The linked article says exactly what rd95 and goobster want it to say - "because Americans are assholes." I have never not seen that sort of argument be overly-simplistic at best and wrong at worst. The Brits, the French, the Swedes and the Japanese are assholes, too but that's just motive. It's not method, means and opportunity.

Newsweek, of course, is happy to oversimplify the issue. If you get a little more rigorous on it you discover that there isn't even a lot of consistency of how many drivers are tested for drunk driving, which skews the accident statistics. Here's what I know:

I dated a Serbian girl who every year would make a big road trip to the coast with her family. They'd pack up the car, buy snacks, get the maps out and settle in for a mind-blowing three hour drive. In Girl with the Dragon Tattoo, Steig Larssen has his hero travel from Stockholm to the "back of beyond", an island an hour north of Gavle. That makes it two hours by car.

here in the United States, you can drive about a thousand miles a day. Two hours? I have friends that commute that to work every day. and back.

IF: Driving is inherently dangerous

AND: American culture simply involves more driving

THEN: it's entirely possible that the reason our death rate is going down less fast than the rest of the world is we're closer to the asymptote. And I don't think it's fair to wave hands and say "it's because Americans are assholes."

goobster  ·  1 day ago  ·  link  ·  

Psh. FACTS.

Doesn't change the fact that Americans are assholes! ;-)

veen  ·  1 day ago  ·  link  ·  

I have to admit, part of me also wanted that simple answer. My tiny roadtrip from San Diego to LA and back felt like a long-ass drive. And in that six hours, I got cut off by an asshole in a BMW thrice. But it's not like the rest of the world are saints: the Dutch are known to drive selfishly, rarely making room when you want to merge for example. Germans are nicer but are worse at tailgaiting.

What I did find worrisome was that driving in the U.S. and Canada was way easier (in terms of mental energy needed) than I expected. I did not feel like I needed to pay attention as much as I do over here, like once I'm on the right avenue I just have to follow the guy in front of me and not run a red light, even in dense urban areas. It was tempting to check my phone because driving was legitimately more boring than I was used to. My impression is that distracted driving is a much bigger issue over there than here.