rpedroso

Suppose that we eliminate net neutrality. Further suppose that, in a few years, I start a video streaming service called "Notflix".

Because ISPs are no longer required to be dumb pipes, they've started charging their customers a premium for access to video content. This is bad for Netflix, of course, because it cuts into their subscriber base.

They crunch some numbers and come up with an amount they're willing to pay the ISPs to exclude their subscribers from this segmented pricing scheme. The ISPs accept this mutually beneficial deal, and everyone's happy, right?

What about Notflix, though? We don't have the capital that Netflix has to pay into this protection racket, so our subscribers must choose between paying the ISP video premium + our subscription fee, or Netflix's subscription fee.

That's an uphill battle and straightforwardly anti-competitive, but is also the precisely the promise of a world without net neutrality.

I also disagree with your assessment that anybody is subsidizing internet giants. All those players pay for the bandwidth they consume to dump their data into the network. The end-user pays for the other end of the connection. Everything is being paid for, proportionally to the usage.

The last-mile ISPs simply want to double-dip. Comcast thinks Netflix should pay both Level 3 AND Comcast for using the pipes, which fundamentally undermines the architecture of the internet.

IANAL and I don't have an answer to this, but I would be deeply alarmed if this were the case. I can understand them making the case that anything on your personal is searchable (though I disagree that this should be allowed).

By asking you to sign in and sync, they're not just requesting access to information on your person -- that's an enormous expansion of their search powers.

Pai makes a lot of questionable arguments:

The FCC’s regulations weren’t about protecting consumers’ privacy. They were about government picking winners and losers in the marketplace. If two online companies have access to the same data about your Internet usage, why should the federal government give one company greater leeway to use it than the other?

1. Is he implying that the previous FCC was intentionally boosting content providers over ISPs? What motivation would they even have for doing this?

2. Content providers and ISPs do not directly compete with each other in the course of their primary business, so in what respect does the rule create winners and losers in the marketplace?

I suppose he could mean in the advertising marketplace. I guess I would prefer that both Facebook and Comcast be losers in that marketplace though.

3. Two online companies do not have access to the same data. Facebook has data + metadata for any interaction you have with their platform. Your ISP has metadata for interactions you have with any platform, along with data for any plaintext interactions. This is a massive difference in scale.

Users have the choice not to use Facebook, and thus not provide them with data. Users do not have the choice to access the internet without making use of a public utility (i.e. their ISP).

He cites an expert saying:

Rather, the most commercially valuable information about online users . . . is coming from other contexts,” such as social-media interactions and search terms.

I can see how user-shared data might be considered more commercially valuable, but this misses the point about the differences in the kinds of data we are talking about. When someone publishes on social media, they're making an explicit choice to share information with the public.

When that same person visits a website, it is not commonly understood that this act might reveal personal information (though end-users ought to be more aware of this). Consider a teenager doing research on pregnancy tests -- they do not intentionally publish information, but the metadata (webpages visited, time of visits, etc) is potentially revealing.

I primarily write python web-based APIs for a web application + 2 mobile apps. Just the other day, I was dealing with an endpoint that had to update hierarchical data (i.e. a collection of trees).

Due to the circumstances, normalization wasn't an efficient option. I ended up throwing together a barebones tree with a 5-line DFS implementation to traverse it. It handled inserts, updates and deletions (for my use-case) in linear time.

The details aren't so important as the fact that adding a dependency would have been overkill for my needs. This isn't to say that efficient graph implementation libraries should not exist or be used, but I was able to produce this code faster by having that basic CS knowledge.

I write python professionally and I use vim exclusively, but I have pylint + jedi plugins and could not imagine working without them.

I'm also not convinced that depriving new programmers of real-time syntax checking actually helps them learn. Wouldn't immediate feedback help them memorize the correct syntax?

They're programmers, the fast idiot is a computer.

There are a few problems with this explanation:

1. FDR did not propose a wage cap until 1942, 3 years after the Great Depression had ended. It was a temporary wartime fundraising effort (and was quite popular with the public). In fact, health insurance really first came into existence during the Great Depression (mostly to ensure physicians and hospitals got paid).

2. FDR's wage cap did not pass.

What actually happened was the passing of the the Stabilization Act of 1942, which gave FDR a power (that he invoked) to freeze wages and salaries during the war. The freeze was deeply unpopular with the labor movement, who threatened mass strikes. As a compromise, congress exempted health insurance (and other benefits like PTO and pensions) from the freeze.

This was certainly a contributing factor (proportion of US population with healthcare roughly doubled from 1940 to 1945), but ignores the other factors that prompted the rise of employer sponsored healthcare. Indeed, by 1945, less than a quarter of Americans had health insurance.

Rather, the two largest contributors to employer-sponsored health insurance was the demand of health coverage by labor unions and the 1954 legislative change that made health insurance tax exempt.

I have no suggestions for the latter two, but number of commits / user / repo is doable in vanilla git:

  git shortlog -n -s

The Electrician was functioning as an Electrician, not as a government agent.

I don't think your example vindicates Best Buy, but rather, simply repeats the central question of the case: was the Geek Squad employee simply functioning as a Geek Squad employee?

The prosecution has argued that an employee who happens to stumble on images of child pornography (analogous to your electrician stumbling on a drug lab) is not acting as an agent of the government. I'm inclined to agree with that judgement.

On the other hand, if the employee was conducting extra thorough searches, scrutinizing the files on any customer storage media, or otherwise performing surveillance tasks that had nothing to do with his job, then it seems apparent that the employee was acting as an agent of the law in accordance with a financial incentive from the FBI. It doesn't help, in this case, that the employee lied about having been compensated by the FBI.

Ultimately, this particular case will come down to the details of how and why the employee stumbled across these images. As a general principle, however, it seems wrong to me that any computer in for miscellaneous repairs (touchpad/screen replacement, battery refurbish, etc) should be subject to a search for illegal data pursuant to an FBI incentives program.

I think the correct tradeoff is to judge the malfeasance of the product based on what security precautions were reasonable at the time the product was created.

Granting that "reasonability" is a very fuzzy standard, it seems obvious that a product with 30 year old crypto should not be subject to lawsuits because someone got solved integer factorization on real hardware.

One unfortunate side-effect of traffic capacity expansion is that usage increases to compensate:

http://www.brown.edu/Departments/Economics/Faculty/Matthew_T...

The underlying assumption is that a multitude of users would switch to devices produced by such a manufacturer. This, I think, overestimates how much most users currently care about security.

As it turns out, there are more secure devices in the marketplace than the affected phones, but they cost more. All other things equal, a contractual obligation for security policies would increase the cost (and thus price) of devices, and users would likely stick with cheaper options.

That's a good point, and thanks for the recommendation!

Public auditing of voting terminals strikes me as a place where open source of government-used software matters.

Well, notwithstanding all the ordinary people who got blood tests done by Theranos that were later invalidated.

For those who prefer CLI git: I've used gitk a fair bit, but lately I've been gravitating towards tig, which implements a tree view via ncurses.

Actually yes:

https://en.wikipedia.org/wiki/Synthetic_cannabinoids

They typically cause more severe side-effects than traditional marijuana. There have been a few rounds of subway ads related to synthetic marijuana.

To clarify, PEP 8 nonetheless recommends that lines be limited to 79 characters, and even recommends capping docstrings and comments to 72 characters. The python standard library follows those guidelines.

The 100-character allowance is for code "maintained exclusively or primarily by a team that can reach agreement on this issue".

Yes, it is a feature of the default browser stylesheet, but many websites override it. DuckDuckGo, for instance, keeps all of its links black and uses the :visited pseudo class to show the checkmark.

Even if it's default browser behavior, keeping it around is a specific design decision.

Little details like checkboxes next to places I've already been are super helpful and I miss those little touches in google.

Google does do this, and in a less subtle way -- it applies the classic, purple styling to visited links.

Still, I find your comments compelling enough to give DDG a shot. I found the results inadequate when I last tried a year or so ago -- I found myself using !g more often than not, which led me to abandon DDG altogether.

As the article mentions, the polygraph results are still occasionally used in court. Regardless of its utility as an interrogation tool, there are serious concerns about the scientific validity of its output. The use of polygraph output to bar people from jobs and as evidence in court deserves scrutiny.

Carbon neutrality does not mean that atmospheric CO2 is not released at any point, but rather, that any release of carbon is offset in some way. Offsets might include planting trees, purchasing carbon credits, etc.

I honestly take issue with the way you phrase your argument. You refer to the poor with phrases like "these people", trivialize their ability to survive without a week's paycheck because "they've done without before" and accuse them of spending frivolously without any real evidence.

My point in writing my comment was to emphasize that the poor pay disproportionately more of their income towards basic needs than the middle class. My hope was that I would provoke some empathy when discussing the economics of poverty.

I don't disagree with you that bail is a tricky issue -- for it to accomplish its goal, it has to mean something. But on the flip side, there are people living in New York City who cannot afford bail and get stuck in the justice system for years without recourse, without a speedy trial, and that seems unjust to me.[1]

A little empathy goes a long way.

[1] http://www.newyorker.com/magazine/2014/10/06/before-the-law

Yeah I figured I would be off by a fair bit off when it came to the income tax math. I'm not sure whether there would additionally be some tax burden due to NY state income tax and NY city income tax. You're right that $2000 is huge to a minimum-wage earner -- that's almost 11% more disposable income.

On the flip side, minimum-wage earners are more likely to be employed part-time instead of full-time, but only 10% and 2% respectively of those groups are paid minimum wage or below, so it's pretty tricky to come up with a good estimate here.[1]

My main goal was to illustrate that the poor spend disproportionately more of their income on basic needs compared to the middle class, and inspire some empathy when we discuss the economics of poverty.

[1] http://www.bls.gov/cps/minwage2013.pdf

The minimum wage in New York is $8.75. Someone working 40 hours/week for 52 weeks/year, which is a generous estimate due to how many minimum wage employers hire and schedule employees, would earn $18,200 every year. After 15% in federal income taxes, that leaves us with $15,470. Let's assume they pay nothing in state/local income taxes. Let's also assume they live alone and have no children or family to care for-- all their expenses are their own.

Assuming that they pay $1000/mo for their apartment (let's be generous -- this includes rent and utilities), that leaves them with $3,470 every year to cover all other living expenses.

If I spend $50/week on food (which is conservative for NYC), that adds up to $2400, which leaves us at $1070. At that salary, if you spend more than $70 in a YEAR on clothing, healthcare, toothpaste, public transit), then you won't be able to afford $1000 in bail. I spend more than that every year on public transit alone (MTA 30-day pass costs $116.50 -> $1398.00/year).

It's really not hard to see how a $1000 bail can be out of reach for the poor in New York.

Certain politicians in Iceland wanted to ban a particular subset of pornography (hardcore, violence, etc). I don't remember hearing that any progress was made on that legislation, but I could just be uninformed on that front.

There was some concern about the security of the product:

http://2014.hackitoergosum.org/bittorrentsync-security-priva...

These concerns were amplified by the difficulty of auditing a closed-source product. Their argument that hashes are one-time secrets and not permanent keys is difficult to validate without access to the source.

The main reason I bought Spotify premium (other than access to the full mobile platform) was to ditch the ads. I don't mind ads so much, but Spotify ads are particularly annoying. I seemed to get a lot of ads from fledgling hiphop artists, including extended samples of their songs, which was really disruptive and unpleasant when listening to folk/americana. After one two many focus sessions were ruined by obnoxious ads, I went ahead and bought premium.

Perhaps this was intentional :-)

If you want to enforce this more stringently, you can even prefix methods with __ (double underscore). Python will mangle the name:

    __mydef -> _myClass__mydef

BuzzFeed definitely still publishes Mashable-like "top-n" list pieces and similar entertainment pieces, but their news section[1] has improved significantly over the years.

[1] http://www.buzzfeed.com/news