Hacker News Reader

> And let's not get started about the legal acumen of the site as a whole.

This has been disappointing, not because I expect everyone to be lawyer, but because I expect HN commenters to be able to use an internet search engine. It'd be one thing to miss details that you need years of training to understand correctly, but a huge proportion of the comments in these threads strongly suggest that the person posting them has not spent even five minutes researching the subject they're posting on, and yet has somehow arrived at strong opinions on the subject anyway.

Yes! It's like reading and tracking down sources is a kind of superpower here; sometimes, it feels like threads treat that as a kind of unfair advantage. In fact, if pressed, I could cite examples of commenters on threads asserting that.

I don't have time to hunt for sources, honestly. I'm not attempting to be a bastion of truth when I interact on the Internet, I'm attempting to explain my take on an issue, or ask a question based on what I already know.

Maybe I'm the problem, but I'm not going to change. I just don't have the resources to be 100% right every time I say things online.

Tracking down sources are not always as simple as spending a few clicks on a search engine. I particularly find talks to be problematic, as the content is not indexed, nor is it easy to remember which n'th talk the speaker said a particular fact. You basically have to re-watch them all, which for a few comments I have done.

Worse is legal case findings which for whatever reason, the media did not pick up. Take the statement that in Swedish law, people who produce or run a webservice can be made liable if the majority of users use the service for illegal purposes? That facts is basically impossible to find using a search, even through it is written plainly as a simple Q&A in the appeal court judgment of the TPB trial. If the case had been that I forgot where I read it, a search query would not have helped me in tracking it down.

The site has been equally embarrassingly bad about taking certain claims at face value, like, "no direct access to the servers", when it is painfully clear to anyone running a colo how NSA PowerPoints could talk about data direct from BigCom servers at the same time as BigCom denies giving direct access to its servers, with both 100% "technically true."

What HN could use is a bit less knee jerking towards belief based discussion, and a bit more analysis: we have these two claims, assuming both parties are self interested, could both be true, and if so, how.

I see "of course Google is/isn't giving server logins" but I don't see as much "here are ways a third party could get data directly from servers, for these various definitions and implementations of 'directly'."

That stuff does get said here more than other places I'm reading, but still clearly not enough as I haven't yet seen that kind of analysis get noticed and picked up by the reporters increasingly sourcing their tech digests from here.

This comment neatly characterizes the kinds of discussions we have on this issue; it starts with innuendo about how the NSA could have what any reasonable person would refer to as direct access to servers, then retreats to a broad, abstract position crafted to make the innuendo harder to rebut.

Thank you.

You seem to frame it in a very odd light. I don't see anything in that comment could be refereed as an "innuendo", nor do I see any source for comment to be crafted as to make something harder to rebut.

From that, I can only ask if you are arguing against a honest intellectual discussion, based on facts as well as rational arguments in which the criterion of the truth is not sensory but intellectual and deductive?

In this specific case of "direct access", the facts are the few press-release made, the leaked slides, and any contributing leaked report. The rational arguments is mostly around the definition of "direct access". The truth is thus depended on the quality of the facts, and the derived result of discussing the rational arguments.

Is this bad for HN, and if so, what should be done about it?

Thanks. This thread is a meta discussion about HN, not a technical discussion on concrete data collection methods.

I agree NSA's slides are not innuendo. Neither were BigCom denials. To me it seems reconciling those is neither a rebuttal nor a retreat, it's advancing the conversation from two disputing sides (NSA lying vs corporate collusion) to a third "this is likely what's meant, as seen in multiple concrete cases from 2006 to today, and makes sense of currently available info."

The 19 July story by Pete Ashdown, CEO of XMission, disclosing one in flight data capture practice well known to the data center community, was not abstract. The latest xkeyscore reveal fits this model as well.

"Are they or aren't they" isn't the most productive debate to have anyway. If they aren't, they could, and if the end result is the same, what should we do about it?

As you suggest, the discussions in this area I do appreciate at HN are on what honest rational policy should be, and on how technologists can assist in ensuring trust in confidentiality and non-repudiation in the cloud of services HNers are building.

>In fact, this site has been embarrassingly bad about the technical issues of surveillance, for reasons ranging from gullibility to capriciousness.

Poor analysis by some of the users (there are a lot of non-technical commenters on here) doesn't negate the higher degree of technical discussion that has indeed been present here. Just because opinion and fallacious arguments are present doesn't mean that good technical discussion isn't. Outside of dedicated infosec communities, I am not sure what online community has had more purely technical discussion on these issues over time. Feel free to list them though, because without sarcasm, I would be happy to know of them.

I don't think tptacek is saying that good posts don't exist on HN, of course they do. The problem is that the signal-to-noise ratio is so low nowadays that it's hardly worth wading through all the hyper-politicized vitriol just to get to it.

The amount of horribly bad posts on HN has reached a proportion where they're no longer exceptions, they characterize the site as a whole.

Personally I've noticed my participation drop in the last few months because of this. There are fewer and fewer people interested in engaging in a discussion, and more and more religious zealotry where it's clear poster has zero interest in opposing views, and will stoop consistently to hostility and fundamental indecency when confronted.

The problem is that many HN users think that because they are qualified to talk about tech, they are also qualified to talk about everything else. In reality, they're often not qualified to talk about anything.

While I agree with you that all of those things are ridiculous (and I was vocal in arguing against them), is it fair to characterize the entire forum as being so easily swayed by ant-state sentiments?

During the two week "freak out", I saw a lot of linkbait about the NSA having massive conspiracies, but I saw relatively fewer actual comments where people were clearly being swayed by anti-government sentiment. For every comment I read that was outlandish, fallacious and clearly media spoon-fed, I have to say I can recall a thread of people saying, "No, that doesn't make sense, you're trying to disprove a negative", etc.

tl;dr - My point here is that I think the baseline intelligence of Hacker News is higher than we might think it is just by observing the front page, and that there are actually a lot more savvy people gaming the front page who are just driven by a relative few who act as the passionate, vocal majority.

That's just my opinion. I could be wrong. But I like to think there's a lot of under the radar intellectual activity, and people are just being really opportunistic for karma or some such.

As for legal acumen, I agree completely - I don't have nearly as much as, for example, 'rayiner. But that's exactly why we have people with niche expertise or domain knowledge. It's a real problem when people get frenzied and decide they know Constitution without having read it.

I haven't been here as long, but I believe that we have sampling bias from the hugely outspoken minority who know it's trendy to be anti-state.

EDIT: I want to submit my experience about the site being gameable - it's true it's easy to get the top comment for news stories that are heavily politically loaded, but I have to say it's easy to karma farm even if you're not anarchist/cynical/conspiracy mongering. I do not try to game the forum to get high comments, but I can still personally attest to having some top comments in the high 40s during the NSA scandal while being incredibly vocal against the "popular opinion" that Google was directly aiding the government. I probably had the top comment on at least half of those stories, arguing against the tone of the story profusely. I don't have a sockpuppet ring, so those numbers of people who upvoted me are to the best of my knowledge genuine. They may not have been as vocal in their agreement with me as the detractors who replied to my comments, but they certainly exist.

I guess I just want to try to dispel pessimism. I don't think all is lost regarding the political climate of Hacker News :)

Sometimes I wish comments were shorter so you could easily tell what's going on in a discussion.

The parent comment says X, the one below says the opposite, and then someone says X again. Do people click and write mainly to get karma? Should posting also cost karma?

The longer you write the easier it is to say something that's not true and harder for people to follow it accurately.

(I know I'm guilty of this.)

Haha, I'm assuming you're speaking to my comment length...sorry, that's pretty par for the course with my comment history :) I like to write with a level of verbosity.

But...I do agree with you. It becomes harder to sift through facts when a post is very long. I do it because I enjoy writing long prose on topics I'm interested in - I don't think it's particularly correlated with getting high karma. I've seen very high comments that consist of a little paragraph (albeit packed with technical information).

But I think a lot of people do just click and write for karma. As long as there is a karma system, this is somewhat unavoidable. I really wish we could do away with the entire karma system entirely, but your suggestion about posts "costing" karma sounds really neat, I'd definitely test that on a small forum...not sure how you'd deal with throwaways though, and how would new users accrue karma?

One problem with posts costing karma is that it would all but guarantee to stifle long discussions amongst any but the highest karma members -- especially if you end up autobanned when you run out.

That could be a good thing. It's very rare to see a deeply nested thread actually worth reading.

I know pg's added things like a progressive delay to the reply box further into threads. The curve of quality going into a thread is an interesting problem, I can see how the subjects of discussion would go from general to specific (and thus likely less interesting to general readers) the further down branches you go, but also sort of by definition those branches become more and more relevant to the people involved. Maybe a flat or hybrid flat/tree layout would help keep discussions more linear?

I don't know if giving that much more power to older posters is necessarily the answer, although it might help reinforce the perception of the community maintaining a certain tone in discussion, if the same posters are more likely to be heard and heard more often. On the other hand, with that scenario, karma would actually mean something (though that just brings up the possibility of karma-farming posts.)

> For every comment I read that was outlandish, fallacious and clearly media spoon-fed, ...

If you look at the timeline, the CIA had such a spectacular failure that an ambassador was raped to death and Hilary Clinton kicked to the curb. Almost simultaneously the IRS was caught embezzling money from anti-statist campaigns.

Every time those stories threatened to gain traction, every leftist organ would run another 48 point headline about Snowden or the NSA. The coincidences piled up until it is impossible that the NSA story's popularity was not largely a political creation, and just barely might be a false flag operation to punish the intel community.

Likewise, I was downvoted to oblivion every time I pointed out that the NSA story was not a revelation, that it wasn't even news. My first awareness of the NSA was their Echelon spying efforts, where it was openly discussed that they wanted to vacuum up all the worlds' communications. The weakness of the DES cipher was widely recognized to be a NSA plot to make it easy to intercept domestic comms. The Clipper chip and key escrow programs were a naked domestic snooping plan. This was widely covered by the trade press, a fair bit by the mainstream media, exhaustively by Slashdot and Ars Technica, and obsessively by the Computer Underground Digest, the Hacker News Network, the Cipherpunks, Telecom Digest, and many others.

Hacker News has also started importing the Reddit Censorship ethos. Downvoting rings censor many politically correct or just unpopular comments, comments that in many cases are correct but counterintuitive. The endless September seems to have finally arrived at HN.

Politics on HN, Exhibit A for the prosecution.

My comment is a non-partisan analysis of why most of the NSA story is astroturf. Astroturf can only be debunked by describing the conspiracy. This does not make me a partisan either for or against the astroturfers. If HN stories were showing up simultaneously and with the same headlines as press releases from the John Birch Society, I would direct my flamethrower in their direction.

And you ignored the other half of my comment, about how the NSA story is not news. It is merely new to excitable young people who mistake unfamiliarity for exposé. If I can convice them to take the red pill, they will learn that parts of signals intelligence are profoundly more important than even the astroturf claims, and at the same time more mundane.

Still, I think Thomas is right: your comment is an example of what needs to be avoided to prevent further devolution. The issue isn't whether you are right or wrong, whether the comment is political or non-partisan. Rather it's whether HN or any online community can take on such issues without destroying itself. Historically, the odds are poor. If we want to keep quality of the technical discussion high, I think that comments such as yours need to be reserved for elsewhere.

If we had a vi-versus-emacs debate, most people would know not to state their views too strongly. Or even if someone did, other people would refuse to take the bait.

With politics, there seems to be no such restraint.

Perhaps. Infosec is critically important to our society, and HN appears to be responsible for radicalizing a large fraction of practitioners. It is not even a radicalization of substance, but collateral damage from a forgettable unrelated political campaign.

Now it is just about too late. For the next 10-20 years, national infosec policy will be driven by the radicals' memory of their principled stand against the NSA "revelations". "Abuses" will be "curtailed" without regard for legitimate security needs.

The is no elsewhere to reserve this discussion for. If HN is credulous enough to believe staged CNN sound bites, there is no hope for other venues.

> that the NSA story was not a revelation, that it wasn't even news

I still disagree on this (but wouldn't downvote you for expressing that opinion).

I now design and review systems with the assumption that the GPA (global passive adversary) is real. It's not a political thing; it's an observation of technical reality.

To explain why that is a shift in thinking, note that basically every web-site password reset mechanism in the world (apart from those that employ 2FA) is broken in this scenario.

Sensible people cannot expect Tor to provide the fig-leaf of safety it seemed like it offered.

GPA was not a default assumption in threat models before.

How recent do you think this shift is? I don't remember when I learned how juicy a target international telephony is, but it had to have bern the late 90s. Certainly defense contractor salesmen have been treating the hotel telephone with great suspicion for a long time.

I did't realize that anyone outside of movies even used hotel telephones anymore (except to call Housekeeping or the front desk).

About the other stuff, I only recently realized that the IRS scandal, the US spy who was caught in Russia, and Benghazi have basically disappeared from the news, while the one thing that the White House has the least control over and is the most distanced from is the one that is now most talked about.

Another thing to think about is that when the IRS story broke, a lot of new agencies were calling it a "controlled or planned leak" meaning that the white house and IRS had coordinated on how and when to break the story, timing it with new info on Benghazi for information-overload, and finally Snowden was just a freebie, while I'm sure they're not happy about the facts coming to light, nothing internally will really change, they'll continue spying on us, they'll just be more careful who they allow to access the information.

Well, since Snowden. Name a site with secure password reset.

> If you look at the timeline, the CIA had such a spectacular failure that an ambassador was raped to death

That's doubtful. http://www.snopes.com/politics/military/stevens.asp

I understand what you're saying. You're suggesting that I might (perhaps unintentionally) be cherry-picking. I disagree. I think the kinds of commentary I referred to aren't outliers, but rather characterize the site.

They characterize the site now, that's for sure.

But it didn't before, as best I can tell. The switchover seemed to happen in 2013, but it wasn't Snowden's leaks that caused me to notice the change.

Fair enough. I did mean unintentionally.

I am so fucking happy someone people may listen to has been able to articulate visibly what I've been thinking for months now.

>The site isn't merely the boy who cried wolf; but rather a boy with a wolf-oriented case of Tourette's.

I almost hate coming here now, for that reason. Which makes me sad.

>Witness for example several weeks of intense belief that Google had allowed NSA logins to its own servers in order to pull information off of them

At least HN is a place where (presumably) there are users informed enough to set the record straight, rather than having the theories perpetuate. As someone who frequents Reddit, I appreciate that much.

You seem to think you successfully defended Google and the NSA. My technical chops are just as good as your, if not better, and I found your efforts utterly buffoonish. If you don't like people disagreeing with you or criticizing the gov, time to pack it in.