Hacker News Reader

Comment by bitL | Hacker News Reader

It looks like everything in tech got poisoned, smart TVs taking screenshots, web apps tracking and matching user clicks, smartphones tracking locations realtime and who knows what else, desktop apps monitoring other apps and peripherals, creepy companies building profiles on everyone, health institutions selling data of their users... I want out, I didn't get into this field, keeping myself up-to-date and super capable via top universities, to be just another cog in building a toxic monstrosity this industry is becoming just to make somebody with a limited lifespan feel powerful and rich.

View on HN

"just to make somebody with a limited lifespan feel powerful and rich."

Wow that last sentence really puts things into perspective. How can be reverse course and throw a wrench in the system? We are the makers, we should be able to wrestle back control and do it democratically and get politicians on our side to legislate this ad industry into the ground.

Quickest thing you can do is quit your jobs en masse. At some index level of number of people * avg skill of people quitting, the industry becomes nonviable. Then you state your terms.

Sounds like a union? Also sounds like Galt's Gulch. Weird dovetail, there.

The kicker is that tech workers are in a FAR better position than the other groups that are pushing or considering a general strike. I suppose that makes the prospect more viable, but also more dangerous to the stability of the overall economy. I guess it's up to you if a shake-up now is worth stopping or forestalling the rising waters.

This. Tech workers need to organize. Unfortunately you have to convince tech workers that moral duty to society is better than their comfortable high-salary lifestyles, and good luck with that.

This is a non-starter for achieving the end result.

There's enough of a delta in both the money paid for online advertising of a target nature and in the better results that yields for the advertisers for them to fund a very rapid turn-over.

There's enough money in it that if they must, they'll be able and willing to buy their way past any unionization issues.

Sorry to break it to you but most of this "we" already sold out. Much like the corporate hippy boomers.

> "just to make somebody with a limited lifespan feel powerful and rich."

> How can be reverse course and throw a wrench in the system?

We start by taking the guillotines down to Sand Hill Road.

Good grief. Many of the people in privileged positions to complain about this stuff got to those positions through Sand Hill road.

Telling some poor person that they have to pay $10 per month to use Google or send Facebook messages isn’t democracy. Don’t want surveillance? Don’t use the products the employ it, but some people, especially those that don’t have means, might be perfectly fine to trade privacy for a “free” service. Destroying the ad industry is elitism and tone deaf — ads are imperfect, but they have enabled people to do things that would have been impossible or unaffordable 20 years ago.

This pitchforks and guillotines talk is ridiculous. Build something better if you don’t like the way things are.

Do you use ad blockers, VPN, etc?

> Don’t want surveillance? Don’t use the products the employ it

That’s as simple as saying “don’t like crime? Don’t be near criminals.”

Data’s being stolen and we’re being watched whether we like it or not. Only sometimes can we easily opt out and have those decisions respected.

It's not their fault they're exploiting a system that allows what they are doing.

They're not good people, but if it weren't this set of people, another would take their place. The world is rife with opportunity for people of low morals.

It is society's fault that we have not explicitly codified what will not be allowed and constructed the right laws and enforcement to ensure that violators are ruinously punished.

First step would be to kill advertising-based models. Get them banned. Because it's the advertising industry's race to the bottom that poisons everything, and fuels the creation of surveillance infrastructure. With ad-related snooping gone, it will be much easier to rein in the remaining few players who misuse data in pursuit of optimizing their business models.

(And yes, I know ads enable a lot of free content on-line. But as countless problems like this show, it's a bad tradeoff.)

Do you know of anyone working on such an effort?

Self-hosted apps/DApps come to mind however they are still alpha-quality. It might take some time (~5 years) to replace 90% of expected FAANG functionality.

So go back to the 90s? Microsoft being the reigning king? Not sure if this time period was honestly better.

Or alternatively, give users ownership of their data like the EU has taken steps to do. Advertising is here to stay and provides its own use. But there could be something that forces transparency.

>Not sure if this time period was honestly better.

Canter and Siegel was in 1994.

How about no free stuff on the web? Charge for services and make it illegal to sell private information without the subject’s consent and remuneration.

Making it illegal won't solve much. E.g. FB doesn't sell private information, they sell the ability to target demographics. However I understand the underlying principle which you are espousing.

The genie is out of the bottle at this point. IMHO the only method forward is how do we as a society responsibly allow for coexistence such that all parties are satisfied.

>> How about no free stuff on the web? Charge for services and make it illegal to sell private information without the subject’s consent and remuneration.

> Making it illegal won't solve much. E.g. FB doesn't sell private information, they sell the ability to target demographics.

You just need to appropriately specify what's being made illegal. You're right that FB could weasel out of a law that outlawed selling private information. They couldn't weasel out of a law that outlawed monetizing personal data without user consent.

Precisely. And then extend it further. Make advertisers responsible for the behavior of the ad marketplaces, ad tech, ad venues they're paying.

Make their only defense be cooperation in prosecuting the offending party.

The way to criminalize it properly is to create a presumption of criminal activity when an advertiser benefits from ad targeting. Literally anything more than the advertiser choosing what sites to appear on should be made criminal. No other criteria allowed.

Nothing short of inquisition-style inspections of company servers, code, and networks is going to fix it on that front. We'll have to do some really radical invasions of company privacy before it results in more than fines-as-a-business-expense.

Not really, you just need to incentivise the behaviour. Pass a law to make it illegal to show ads based on tracking, with executive imprisonment for violations. Executives will then demand their teams to put controls in place to ensure that no one is using tracking to show ads. Violations of HIPAA have some serious fines, and I've watched executives at all levels take it VERY seriously

Rich executives and therefore the media tend to push back on such limits. And they tend to be more equal than the rest of us.

Wouldn't this executive and Media companies will run afoul of the GDPR act. At least in EU they can be fined and may be prosecuted if they are repeat offenders.

It will be nice if all the EU residents in one go send a request to remove their information from all the US based media and technology giants who under the guise of changing the world are indeed working to sale more to consumers and make themselves rich.

Awesome idea, HIPAA for protection against the advertising industry and their digital intrusion/violation/surveillance of our privacy.

Something like APPAA - Advertising Privacy Protection and Accountability Act

What would you guys call it? what language is necessary to cover all the edge cases for the deceptive and dirty-playing advertising industry?

Luckily 1/3rd-ish of those are happening over HTTP(S) over IP on PC.

There was a time before now where advertising was effective and didnt track you - it was the advent of targeted advertisement which really killed things off.

Can we go back to the days where an advert was was just an image and a hyperlink? Where advertisement paid by the pixel and location on the website? Where JavaScript was unused unless in some rare and warranted cases?

I still believe the web can be a free and open market place of ideas.

Most ads that I remember when I was new to the internet were all running flash and had really annoying circus-game-like objectives such as "Can you shoot the basketball through the hoop?!?"

I hate the means by which advertising is targeted today, but I would be lying if I said the format of the ads themselves were more annoying or less useful than the past.

There are two troubles:

1. The uplift of targeted advertising is unbelievable until you see the actual statistics. It's like slowly sipping a cup of coffee to wake up versus waking up to snort a line of crack.

2. Advertisers were abused and defrauded by adtech. Which has inspired all kinds of surveillance hellscape because the advertisers finally caught wise and have renegotiated to pay for actual performance only -- not clicks -- actually closed sales. But adtech wants paid if you do your research online, respond to an ad online, and then buy in store. And a whole lot of adtech now allows for that. Attributing an in-store purchase with no customer interaction to a prior web session by that same party.

The benefit of those two factors to the advertisers is such that we can't have a serious discussion about this shit going away without a law which assigns criminal penalties for being a beneficiary of the scheme.

This is just a non starter and a pipe dream. It goes against the principles of capitalism and free enterprise.

You need to think about incentives. You have two tablets, identical in functionality and performance. But the one without the ads and surveillance features costs twenty percent more. Which one do you buy? Actually Kindle did this for a while (maybe they still do...). I ended up buying the cheaper one with the ads.

Can confirm that Kindle does in fact sell devices that are ad-supported.

For a $20 one time fee you can remove this feature (which can be done after purchase at any time). But most people won't notice the tiny option select on the amazon purchase page that defaults to "with Special Offers".

To make it even more confusing, they make it sound like the better option to pick is the ad-supported version. I mean you are choosing between the model "with special offers" or the one "without special offers"? Most people that don't know any better will leave the default "with special offers" options selected.

Source: https://www.amazon.com/All-new-Kindle-now-with-a-built-in-fr...

Yes, you're right. I know people who bought Kindle with ads just to save that 20% or so.

But that's why I wrote that advertising-based business models need to be banned. Not discouraged, not badmouthed, but banned. They're anticompetitive and poisonous; when one company starts doing it, others in the whole sector are drawn to follow suit (it's e.g. why it's hard to actually sell apps on mobile or make subscriptions for publications on-line profitable; ad-supported operations create a baseline cost of zero).

Ads enabling free online content is a bit of a glazier's fallacy. Yes, when ads are an option most content producers will be funded by ads. it's simple to implement and does not require end user consent but as soon as the degenerate strategy is removed new strategies become viable such as micro-payments.

I think the potential of micro-payments was never reached because transactions are currently too large, and uni-directional. If we could process payments of $0.01, it would create a culture were you are actually willing to pay for content. Also, if you make a insightful comment, a product review, or even a upvote, you should receive some credits to spend back on consuming content. Doesn't have to be blockchain, it would be effective even if limited to one site or content provider.

What we actually need in a practical sense is Kindle Unlimited but for web content.

And furthermore it could be a plurality of those kinds of providers aggregating content.

Deploy single-sign-on schemes, and websites might participate in a plurality of programs from different vendors.

But at the end of the day, you'd pay one or two "providers" a monthly sub, they pool the funds, take their cut, and do prorata distribution of the pool based on views, eyeball time, popularity of content, lots of ways.

No need to perform microtransactions from a banking perspective. You're going to eat $20 of web content this month, and so will lots of others. And then those views can participate in the pool and get paid monthly or something.

The entire targeted ad model has to die.

As in, being a beneficent party of a targeted ad campaign becomes a presumption of criminal activity.

We have to make the advertisers culpable for the behavior of the companies serving up their ads.

Why legislate it? Why not make a thing that doesn’t need advertising and charge people money for it. The way to “end” the industry you don’t like is democratically — using dollars as votes.

The idea of using government to crush an industry is a bit totalitarian — it “the people” agree with you, they should be happy to pay you for your product. If they don’t agree, then there isn’t anything democratic about using a government to shut down an industry you don’t like — that’s not democracy, that’s fascism.

Here's a radical idea. Embrace that technology will bring everything to light and essentially destroy privacy. Instead, let's ensure the powerful people in our society also have no privacy. Let's ensure we can watch the watchers with more eyes than they watch us with.

A person with a private security army, a 40 acre compound and a personal helicopter has very little to fear from the general populace knowing things. They never have to worry about the price of insurance, or getting a job, or securing a loan vs being homeless.

This does not solve the problem, not even close.

They might have to worry about the aspiring plans of that security army. It's always something.

Doesn't help at all.

Privacy of varying levels is and has been a functional requirement for smooth working of free society.

Economic disparities make the impact of lower levels of overall societal privacy have a disproportionate negative impact to those on the lower end of the scale.

Working on a Kubernetes based clone of Instagram.

My plan is to deploy it and a VPN tunnel and give certain folks access to keep in touch. I’ll have instructions for self hosting and VPN key creation/sharing (Wireguard ftw)

There’s absolutely no reason to bother with cloud services. They’re nothing but big corp coopting our problem solving.

It always comes down to be gatekeeped but no one having the guts to gate keep a rich douche whose money to buy security goes away as soon as we do

My biggest challenge is communicating with non-technical family and friends. Teaching them Whatsapp was a challenge, but things like IG and FB are simple to install and log in to. I've never been able to quit them without suddenly being isolated.

I never jumped on the “using social media means I am connected” band wagon.

Still just using email. It’s web scale, and just needs UX love.

But really even that is overkill. Self hosting is too easy and cheap for me to justify cloud services privacy and just generally being in the habit of externalizing every aspect of utilitarian life.

I’m not talking webscale loads. And it could be a hub for IOT. My data streams are not Google scale. But don’t take your eye off them sticks & options. Ooo shiny

Share my data with my doctor directly over local area WiFi, by making it adhere to a specific format. No data middle men needed.

For a culture constantly climbing up its own ass about austerity in economics, we sure enjoy selling ridiculously uneconomical means of communication.

It’s almost as if it’s a purpose built emotional response but who could believe so many people would fall for an emotional mass delusion?

This is an awesome project, please post this to HN if you open source it (please do!)

Create enough noisy data so that the data these companies receive is so unreliable that they give up. Have a Chrome extension that runs and clicks on things randomly or a separate app that controls the mouse that feeds bad data. If it ran on every computer maybe Google AdWords might become useless.

IDK, my go-to strategy for now is to support products that respect my privacy and recommend them to friends and family. My friends and family don't know this is going on or that alternatives exist, so I'm their reference for what is and isn't good.

I refuse to work for companies I don't agree with, which hurts me financially. I will never work at a FAANG company, for example, or most of the other heavy weights that are funny l functionally similar.

For things without a clearly superior alternative, I have a list of business opportunities. For example, smart devices are becoming popular, and they're horrendous for privacy. That means there's a market for devices that don't spy on you, and the open source options after inconvenient enough that a packaged deal is attractive, even if it could be DIY. For example, I think there's room for a Ring competitor that is E2E encrypted, provided the app is well designed and the device is unobtrusive. Privacy respecting services and devices are unlikely to take over the alternatives, but merely existing puts pressure on the major players to act better.

Thanks for taking a stand against FAANG, I know my wallet is hurting for doing the same.

On the bright side, as biotech advances eventually you’re gonna get rich powerful immortal bosses.

Oh yes, the best/worst case scenario all at once.

Only best. Aged meat tastes better.

hello Altered Carbon

I'm 100% like you feel. I also did a top uni for years, it was super stressful but I eventually managed. Now I just got a work at a tech company similar to Wacom. I thought okay, we all know how unethical are Google, FB & friends, but a company selling hardware accessories only, c'mon no way it should not be too evil... Well that Wacom story is cute in comparaison to what they do. Ofc they also log all apps used like wacom, but they are also implementing systems to analyse if you have health issues from your peripheral logs, analyse your emotions and stress level from microphones & cameras, facial recognition, eye gaze, etc.

I love technology and computer science but tech is so screwed up in terms of ethics.

I wish we'd see more people coming together that care about this (like truely care, not the #Tethics of the sillicon valley) to make some open and private alternatives to all this toxicity. But it is super hard to make things change.

I'll work in that direction in my free time, but I feel so alone. HN seems the only place people care a bit about that. Around me at uni or at work, the level of ethics and care for privacy is so low, it's depressing. It's not only that "rich boss" telling its employees to exploit people's data, it's also engineers themselves being happy collabo of this because they make huge salary working for those companies.

Don't forget your employer selling your paycheck records, FFS.

The Work Number's entire sales pitch is... "We'll accept the data from you for free. You can give our our number and website to anyone who calls into your HR looking for employment reference, employment verification, etc. We'll field those calls for you."

They don't pay for the paystub data. The employers give it to them.

Although it's an invasion of privacy, to be sure, it actually does have some benefits for the employee.

In places outside San Fran, where people actually get conforming mortgages, having your data in The Work Number's database automates and cuts out the employment & income verification so that you don't have to track down records and submit manually and can potentially skip multiple must-connect phone calls between the lender and employer.

I don't think we have to quit our day jobs to change the world. What we need to do is (1) change the expectations of the public via conversations, activism, migrating away from FB/Google/Microsoft products, etc. and (2) make it easier for people to prioritize privacy by building open-source apps and hardware.

Inspiring examples that I use daily include Linux, git, and Bitwarden.

The general public are too disinterested to make it possible to convince the majority to change the way they make decisions about software and hardware.

Legislation is the only effective course.

I'm not sure how you can propose legislation without the majority approval.

A functioning democracy legislates according to the will of the people. That means you HAVE to convince the majority first.

In a broken democracy, you still need to convince the powerful (although this might actually be easier). But you still have to convince them to apply the same standard to those without power, which is likely a hard sell.

I'm convinced the parents of the US, and in turn broad bipartisan support, could get behind some sane legislation on this stuff if a massive dump like the New York Times one were deanonymized and focused on kids.

Is Wacom's driver an example of excessive monitoring?

While the author presents the graphics tablet as a glorified mouse, tablets usually offer many more features. How those features interact with various applications is important, and they have to prioritize which applications they support. The data collection that the author describes may be viewed, internally, as part of that process.

Now I am not claiming that Wacom is doing the right thing, nor am I claiming that they are doing the right thing in the wrong way. Yet it is entirely possible that they feel justified in collecting that data for product development without having ulterior motives. Their failure may simply lay in the failure to recognize that many people are sensitive to data collection due to real, potential, or perceived abuses by other parties.

It’s 2020. We’re starting to develop a pretty solid framework around what is respectful to users and what isn’t, how to disclose things, ask for permission, etc. (All knowledge that has been acquired the hard way - through data breaches, users getting their personal data stolen because of a lazy software company, etc)

Wacom is a $500M company. They don’t get the benefit of the doubt.

Starting? It's been known for many years, it's just that $$$ talks louder.

Or the failure may be on your end, believing that "many people are sensitive to data collection" while in fact, most people don't give a dusty fuck about it and happily share everything for saving a few bucks a month.

Hackernews is NOT the people. HN represents a TINY TINY fraction of users.

You're saying this group of "most people" knows what's being sent, that they're exercising informed consent? Surely you aren't hanging this argument on "common sense."

You are saying that most people doesn’t? Any polls or data supporting that assertion? Bring data to the discussion.

Here’s some: https://www.pewresearch.org/fact-tank/2018/03/27/americans-c...

91% of Americans feel that they have lost control of their personal data and privacy. The logical conclusion is that at least that many understand what they have gotten themselves into. That would indicate that a majority of people are exercising informed consent, despite the vast majority of Americans feeling that way, they continue to use the gamut of products and services.

The sad thing is, they don't know, and they don't care. Like at all.

It's true, until you show them how much data is collected and who is buying and selling the information without their consent. Then a significant portion start to recognize the threat. There's a reason none of these companies mention it.

One thing I've noticed about far too many people in tech is that they all seem to believe that those not in tech are stupid.

Many of them are not stupid. On average, half of them are above average. They're just uninformed and busy with their own lives.

They don't care. Until you can show them you know how often they're on Grindr and where their tricks live.

Or that they got a prescription filled. For Valtrex.

What would be helpful -- but that I am adamantly against -- would be tons of data drops, in communities across the nation, of local church leaders and local community leaders.

That's why we should give a fuck on their behalf. In a techno-capital society, it's too much to expect normal people to have to know the technical details of all of these things.

I don't know anything about water treatment or nuclear power, but I still expect the people working in those industries to be held to extremely high standards of competence, virtue and accountability.

We should have the same standards. We don't, so instead we need to demand regulations for these monsters.

Civilized societies don't tolerate "vampires" and cannibals walking among them (or lording from on high). They eliminate them. Eventually the people will wake up.

It's just one more example, but one that is rather unexpected. It forces me to ask questions about why? and if we are so deep down that road that this level of monitoring/surveillance became the new normal and an expected outcome of production cycles? However, I think all the gaming companies have much more intrusive monitoring to prevent cheating, NVidia might need to do the same for their GeForce Experience and per-game customization, not forgetting creepy antivirus software like Avast etc.

I have been a Wacom customer for years.

I'm not a graphic artist, but I hate mouse cords and hate having to recharge mice or deal with batteries.

So a series of Wacom "puck" mice on one (over the years, several) of their digitizer tablets has been my mouse substitute at my desktop. I bought the high end ones. On an average of every 3 to 4 years.

They stopped making the puck several years ago. Mine was starting to wear out, so I finally made the leap to Logitech's G703 and the Logitech G PowerPlay inductive mat. So same benefit -- the mouse is just magically always charged.

If I hadn't already switched, I would have anyway after the Wacom selling data thing...

Well put, I’m ashamed at the current state of our industry.

I wanted to add on the list of privacy invading examples. It's better I make the counter claim.

The data collected has massive potential to improve medical research. Being able to validate database wide experiments on hundreds of millions of people at once is pretty incredible. There's likely to be a decade of insights to be found in this rapidly filling digital ocean of information.

In several years the clamour to get off the known web will empower a lot of security apps (not "privacy" apps, that are the opposite of their name) that are growing behind the scenes.

You have a valid point, I did some Deep Learning-based medical research as well as analytics on top of FHIR and having more data is absolutely essential for e.g. finding correlations/covariate conditions/risk factors etc. However, there are massive privacy risks, economically-ruinous or literally life-threatening.

Seems like folks should get to choose what experiments they participate in? Medical ethics 101.

I recently encountered a company that uses ad-tech pioneered identification techniques to snake around patient privacy laws and clinical trial anonymity to sell re-identified data back to insurance companies.

So no, fuck that.

I'm only one voice, but for my part, I am willing to forego the advancements that may come to medicine via this route.

The societal costs of surveillance capitalism are only just starting to appear, and it's going to get so much worse before it gets better.

And it's not all bad, but there's no preserving the little bit of good without canning the tons of bad.

Poisoned tech anecdote: I recently noticed that in one of my favorite bars the coffee grinder has a screensaver.

Considering what should and shouldn't be done is much less popular than finding ways to do it.

I think one of the great signals of good ethical behavior is the person who'll refuse to build a thing that they know should not exist, even when offered a great reward if they'll do so.

s/limited lifespan feel powerful and rich/limited lifespan powerful and rich/

It's a whole attitude. They're aware of their limited lifespan and intend to either buy their way into more and better lifespan (if possible), but in any event become actually powerful and rich.

At least on a certain scale, they're not wrong. It does work.

This is not to say, however, that they're not slugs deserving of a good salting.

> It looks like everything in tech got poisoned...

> ...building a toxic monstrosity this industry is becoming just to make somebody with a limited lifespan feel powerful and rich.

That's what you get when you combine shareholder-focused capitalism, the attitude that everything's OK as long as it's legal, and the attitude that we should avoid making things illegal through regulation (so as not to stifle innovation such as this).

IMHO, all of those ideas should be rejected or greatly curtailed because of the poisonous tree that's grown from them (which bears delicious poison fruit).

The real kicker, though, is the scale at which software allows that combination of ideas to affect people.

you're getting downvoted just for pointing out why the problem exists.

The discussions around these issues always follow the same pattern that reminds me of a dialogue I recently saw posted somewhere, where an Amish person and a non-Amish person talked about technology and the amish person asked the other one, "do you think having the television on is good for you and your family?" and the other person responded "no, but we don't want to get rid of it because it may be useful", and the amish guy responded with "that's the difference between us, if something is bad for our family we throw it out."

The discussions around tech are the exact same. We all agree the modern internet is screwed, large companies put ads into everything, we're getting screwed over, non-profit domain spaces are being sold, everyone's unhappy, and we do .... nothing. Because of 'innovation' or some other conjured up fantasy term.

> everyone's unhappy, and we do .... nothing. Because of 'innovation' or some other conjured up fantasy term.

The only reason "we" do nothing, is that "we" have no agency. Other reasons are just rationalization to cope with our powerlessness: we pretend ideological debates among the people decide the faith of the country. The only vote "we" have is voting with our wallet, which only works in a truly free market.

How much of this is feature creep when teams feel obliged to keep adding more and more to justify their jobs? Let's say you work for Samsung in their TV software division. Other than incremental upgrades in screen technology coming through for you to accommodate, what else can you keep pushing through the pipeline to impress your managers? Maybe smart TV features are the low hanging fruit for these employees? I mean, I dislike that functionality, but maybe the pushback from the public isn't strong enough?

It's not to impress managers, it's to make more money. And most tech employees just don't care enough (or even have enough of a moral compass) to protest. Paycheck first, ethics second.

You are implying that capitalism is the problem? What does privacy look like in non-capitalist places? What was the privacy like in East Germany? Or China? Or Cuba? Capitalism isn’t the problem, it’s the solution: build things people want and the market rewards it.

> You are implying that capitalism is the problem?

Certain kinds of capitalism, yes. There's not just one kind, and the kind we have now isn't the best kind. Perhaps we should try to discover a better one?

> What does privacy look like in non-capitalist places ... like ... China ...?

China is very capitalist now, if you weren't aware.

In any case, the main problem in those countries (at least with regards to privacy) was authoritarianism, not non-capitalism.

If only there was an system somewhere between the United States and East Germany. Sheesh.

It's important to also acknowledge the crucial role culture plays in this problem, regardless of the political system it cradles.

The public is simply ignorant about surveillance technology issues. Not that long ago we used to tolerate sawdust in our bread[1], and that's food, something humans should be pretty knowledgeable about. People would revolt if this happened now, whether they live under a capitalist or communist system. A free market might accelerate the transition, but education about the issue is still the underlying factor of change.

[1]: http://www.theoldfoodie.com/2011/03/sawdust-bread.html

Reply on news.ycombinator.com