biologist
1679 stories
·
11 followers

Things that happen in Silicon Valley and also the Soviet Union

jwz
4 Comments and 14 Shares
Anton Troynikov:

  • Waiting years to receive a car you ordered, to find that it's of poor workmanship and quality.
  • Promises of colonizing the solar system while you toil in drudgery day in, day out.
  • Living five adults to a two room apartment.
  • Being told you are constructing utopia while the system crumbles around you.
  • 'Totally not illegal taxi' taxis by private citizens moonlighting to make ends meet.
  • Everything slaved to the needs of the military-industrial complex.
  • Mandatory workplace political education.
  • Productivity largely falsified to satisfy appearance of sponsoring elites.
  • Deviation from mainstream narrative carries heavy social and political consequences.
  • Networked computers exist but they're really bad.
  • Henry Kissinger visits sometimes for some reason.
  • Elite power struggles result in massive collateral damage, sometimes purges.
  • Failures are bizarrely upheld as triumphs.
  • Otherwise extremely intelligent people just turning the crank because it's the only way to get ahead.
  • The plight of the working class is discussed mainly by people who do no work.
  • The United States as a whole is depicted as evil by default.
  • The currency most people are talking about is fake and worthless.
  • The economy is centrally planned, using opaque algorithms not fully understood by their users.

Previously, previously, previously, previously, previously, previously, previously, previously, previously, previously, previously, previously, previously, previously, previously.

Read the whole story
notadoctor
11 days ago
reply
Oakland, CA
popular
15 days ago
reply
Share this story
Delete
4 public comments
freeAgent
15 days ago
reply
Genius
Los Angeles, CA
laza
15 days ago
reply
So great!!!
Belgrade, Serbia
BLueSS
15 days ago
reply
Good thing silicon valley isn't influencing US elections... Oh wait.
kbrint
16 days ago
reply
Kinda!

We Should All Be Feminists

1 Comment and 2 Shares

This morning I got up and the sun was shining, normally a rare enough occurrence in Scotland, but for the last couple of weeks a startling daily reality. I put on shorts and a t-shirt to make the most it for my morning Whisky walk.

The t-shirt was on the top of my drawer, and I didn’t think about it as I put it on. It has

We Should All Be Feminists

in large type on the front.

As I headed out the door, Matt (who bought me the t-shirt) said,

“If anyone gives you any hassle in that, beat them up”.

I dismissed the idea out of hand, and strode out the door.

Being a Saturday morning, Whisky and I took a slightly different route, to a different park than our usual. He was having a great time, meeting new dogs, smelling new smells, and generally running around like a crazy thing.

As usually happens I got speaking to the other dog owners, in this case two men. We talked about the weather, looking at the beautiful blue sky.

Man 1: “The sky used to be bluer than that. See those airplane trails, know what they are?”

Man 2: “Are they dumping fuel?”

Man 1: “Who would dump fuel? No those are chemicals. Who knows what they are.”

Nervous laughter, and as he continues, it becomes obvious that the guy is deadly serious. Maybe he has issues. Or maybe he’s right – who knows.

Man 1: “The real question is what they are hiding up there. Maybe space ships…”

More nervous laughter from me and Man 2. Our dogs continue to play.

Man 1: “Anyway that’s my rant over for this morning.”

Me: “It’s always good to get these things over and done with early,” I say and smile (it’s not even 9 am).

Man 1: “What’s that written on your t-shirt?”

Me: “We Should All be Feminists.” I say, again smiling

“Man, the world is crazier than I thought,” he replied.

Read the whole story
notadoctor
11 days ago
reply
🤣
Oakland, CA
Share this story
Delete

Digg's v4 launch: an optimism born of necessity.

2 Shares

Digg was having a rough year. Our CEO left the day before I joined. Senior engineers ghosted out the door, dampening productivity and pulling their remaining friends. Fraudulent voting rings circumvented our algorithms, selling access to our front page, and threatening our lives over modifications to prevent their abuse. Our provisioning tools for developer environments broke and no one knew how to fix them, so we reassigned new hires the zombie VMs of recently departed coworkers.

But today wasn't about any of that. Today was reserved for the reversal of the biggest problem that had haunted Digg for the last two years. We were launching a complete rewrite of Digg. We were committed to launching today. We were agreed against further postponing the launch. We were pretty sure the new version, version four, wasn't ready.

The day started. We were naive. Our education lay in wait.

If you'd been fortunate enough to be invited into our cavernous, converted warehouse of an office and felt the buzz, you'd probably guess a celebration was underway. The rewrite from Digg v3.5 to Digg v4 had marched haphazardly forward for nearly two years, and promised to move us from a monolithic community-driven news aggregator to an infinitely personalized aggregator driven by blending your social graph, top influencers, and the global zeitgeist of news.

If our product requirements had continued to flux well into the preceding week, the path to Digg v4 had been clearly established several years earlier, when Digg had been devastated by Google's Panda algorithm update. As that search update took a leisurely month to soak into effect, our fortunes reversed like we'd spat on the gods: we fell from our first--and only--profitable month, and kept falling until our monthly traffic was severed in half. One month, a company culminating a five year path to profitability, the next a company in freefall and about to fundraise from a position of weakness.

Launching v4 was our chance to return to our rightful place among the giants of the internet, and the cavernous office, known by employees as "Murder Church", had been lovingly rearranged for the day. In the middle of the room, an immense wooden table had been positioned to serve as the "war room." It was framed by a ring of couches, where others would stand by to assist. Waiters in black tie attire walked the room with trays of sushi, exquisite small bites and chilled champagne. A bar had been erected, serving drinks of all shapes. Folks slipped upstairs to catch a few games of ping pong.

The problems started slowly.

At one point, an ebullient engineer had declared the entire rewrite could run on two servers and, our minimalist QA environment being much larger to the contrary, we got remarkably close to launching with two servers as our most accurate estimate. The week before launch, the capacity planning project was shifted to Rich and I. We put on a brave farce of installing JMeter and generated as much performance data as we could against the complex, dense and rapidly shifting sands that comprised the rewrite. It was not the least confident I've ever been in my work, I can remember writing a book report on the bus to school about a book I never read in fourth grade, but it is possible we were launching without much sense of whether this was going to work.

We had the suspicion it wouldn't matter much anyway, because we weren't going to be able to order and install new hardware in our datacenters before the launch. Capacity would suffice because it was all we had.

Around 10:00 AM, someone asked when we were going to start the switch, and Mike chimed in helpfully, "We've already started reprovisioning the v3 servers." We had so little capacity that we had decided to reimage all our existing servers and then reprovision them in the new software stack. This was clever from the perspective of reducing our costs, but the optimism it entailed was tinged with madness.

As the flames of rebirth swallowed the previous infrastructure, something curious happened, or perhaps didn't happen. The new site didn't really come up. The operations team rushed out a maintenance page and we collected ourselves around our handsome wooden table, expensive chairs and gnawing sense of dread. This was not going well. We didn't have a rollback plan. The random self-selection of engineers at the table decided our only possible option was to continue rolling forward, and we did. An hour later, the old infrastructure was entirely gone, replaced by the Digg version four.

Servers reprovisioning, maintenance page cajoling visitors, the office took on a "last days of rome" atmosphere. The champagne and open bar flowed, the ping pong table was fully occupied, and the rest of the company looked on, unsure how to help, and coming to terms that Digg's final hail mary had been fumbled. The framed Forbes cover in the lobby firmly a legacy, and assuredly not a harbinger.

The day stretched on, and folks began to leave, but for the engineers swarming the central table, there was much left to do. We had successfully provisioned the new site, but it was still staggering under load, with most pages failing to load. The primary bottleneck was our Cassandra cluster. Rich and I broke off to a conference room and expanded our use of memcache as a write-through-cache shielding Cassandra; a few hours later much of the site started to load for logged out users.

Logged in users, though, were still seeing error pages when they came to the site. The culprit was the rewrite's crown jewel, called MyNews, which provided social context on which of your friends had interacted with each article, and merged all that activity together into a personalized news feed. Well, that is what was supposed to happen, anyway, at this point what it actually did was throw tasteful "startup blue" error pages.

As the day ended, we changed the default page for users from MyNews to TopNews, the global view which was still loading, which made it possible for users to log in and use the site. The MyNews page would still error out, but it was enough for us to go home, tipsy and defeated, survivors of our relaunch celebration.

Folks trickled into the office early the next day, and we regrouped. MyNews was thoroughly broken, the site was breaking like clockwork every four hours, and behind those core issues, dozens of smaller problems were cropping up as well. We'd learned we could fix the periodic breakage by restarting every single process, we hadn't been able to isolate which ones were the source, so we decided to focus on MyNews first.

Once again, Rich and I sequestered ourselves in a conference room, this time with the goal of rewriting our MyNews implementation from scratch. The current version wrote into Cassandra, and its load was crushing the clusters, breaking the social functionality, and degrading all other functionality around it. We decided to rewrite to store the data in Redis, but there was too much data to store in any server, so we would need to rollout a new implementation, a new sharding strategy, and the tooling to manage that tooling.

And we did!

Over the next two days, we implemented a sharded Redis cluster and migrated over to it successfully. It had some bugs--for the Digg's remaining life, I would clandestinely delete large quantities of data from the MyNews cluster because we couldn't afford to size it correctly to store the necessary data and we couldn't agree what to do about it, so each time I ended up deleting the excess data in secret to keep the site running--but it worked, and our prized rewrite flew out the starting gate to begin limping down the track.

It really was limping though, requiring manual restarts of every process each four hours. It took a month to track this bug down, and by the end only three people were left trying. I became so engrossed in understanding the problem, working with Jorge and Mike on the Operations team, that I don't even know if anyone else came into the office that following month. Not understanding this breakage became an affront, and as most folks dropped off--presumably to start applying for jobs because they had a lick of sense--I was possessed by the obsession to fix it.

And we did!

Our API server was a Python Tornado service, that made API calls into our Python backend tier, known as Bobtail (the frontend was Bobcat), and one of the most frequently accessed endpoint was used to retrieve user by their name or id. Because it supported retrieval by either name or id, it set default values for both parameters as empty lists. This is a super reasonable thing to do! However, Python only initializes default parameters when the function is first evaluated, which means that the same list is used for every call to the function. As a result, if you mutate those values, the mutations span across invocations.

In this case, user ids and names were appended to the default lists each time it was called. Over hours, those lists began to retrieve tens of thousands of users on each request, overwhelming even the memcache clusters. This took so long to catch because we returned the values as a dictionary, and the dictionary always included the necessary values, it just happened to also include tens of thousands of extraneous values too, so it never failed in an obvious way. The bug's impact was amplified because we assumed users wouldn't pass in duplicate ids, and would cheerfully retrieve the same id repeatedly for a single request.

We rolled out that final critical fix, and Digg V4 was fully launched. A week later our final CEO would join. A month later we'd have our third round of layoffs. A year later we would sell the company. But for that moment, we'd won.

I was about to hit my six month anniversary.


Digg V4 is sometimes referenced as an example of a catastrophic launch, with an implied lesson that we shouldn't have launched it. At one point, I used to agree, but these days I think we made the right decision to launch. Our traffic was significantly down, we were losing a bunch of money each month, we had recently raised money and knew we couldn't easily raise more. If we'd had the choice between launching something great and something awful, we'd have preferred to launch something great, but instead we had the choice of taking one last swing or turning in our bat quietly.

I'm glad we took the last swing; proud we survived the rough launch.

On the other hand, I'm still shocked that we were so reckless in the launch itself. I remember the meeting where we decided to go ahead with the launch, with Mike vigorously protesting. To the best of my recollection, I remained silent. I hope that I grew from the experience, because even now I'm uncertain how such a talented group put on that display of fuckery.

Read the whole story
notadoctor
11 days ago
reply
Oakland, CA
Share this story
Delete

US puts fierce squeeze on breastfeeding policy, shocking health officials

2 Comments

Enlarge / They did what? (credit: Getty | Tim Clayton)

In May, a US delegation to the World Health Organization issued stunning trade and military threats in its opposition to a well-established and otherwise uncontroversial resolution encouraging breastfeeding, according to new reporting by The New York Times.

The hundreds of delegates in attendance expected an effortless approval of the resolution by the World Health Assembly, which is the decision-making body of WHO. The resolution simply put forth that mother’s milk is the healthiest option for infants and that countries should work to limit any misleading or inaccurate advertising by makers of breast-milk substitutes. It affirms a long-held position by the WHO and is backed by decades of research.

But more than a dozen participants from several countries—most requesting anonymity out of fear of US retaliation—told the Times that the American officials surprised health experts and fellow delegates alike by fiercely opposing the resolution. At first, the US delegates attempted to simply dilute the pro-breastmilk message, voiding language that called for governments to “protect, promote, and support breastfeeding” and limit promotion of competing baby food products that experts warn can be harmful. But when that failed, the US reportedly put the squeeze on countries backing the resolution by making aggressive trade and military threats—a move that further stunned the assembly.

The Ecuadorian delegation, for instance, was expected to introduce the resolution but was weaned off the idea after the US threatened to impose harmful trade measures and withdraw military assistance—which the US is providing in the northern part of the country to help address violence spilling over the border from Colombia.

Officials from the US, Uruguay, and Mexico said that at least a dozen other countries—many of which are poor countries in Africa and Latin America—dropped the resolution after the US sucked away their interest.

“What happened was tantamount to blackmail, with the US holding the world hostage and trying to overturn nearly 40 years of consensus on the best way to protect infant and young child health,” Patti Rundall, a breastfeeding advocate who attended the assembly, told the Times.

Feeding fears

A spokesperson for the Department of Health and Human Services—the US agency that led the effort to drain the resolution—denied threatening Ecuador but explained the agency’s opposition to the resolution, saying:

The resolution as originally drafted placed unnecessary hurdles for mothers seeking to provide nutrition to their children. We recognize not all women are able to breastfeed for a variety of reasons. These women should have the choice and access to alternatives for the health of their babies and not be stigmatized for the ways in which they are able to do so.

Though high quality, safely prepared substitutes can provide adequate nutrition for infants, emphasis on breastfeeding stretches back through decades of concern from health experts and officials that milk-substitute makers were causing harm with their marketing strategies. Advocates argue that companies such as Nestlé have a long history of making misleading nutritional claims about milk substitutes and donating the substitutes to hospitals in developing countries. Starting infants out on a substitute in a maternity ward can make breastfeeding more difficult for mothers later. And once mothers leave the hospital or clinic after giving birth, they may not have access to sanitary conditions or clean water in which to mix powdered substitutes and/or be able to afford enough substitute to keep their infants nourished. These scenarios open infants up to disease and malnutrition, advocates say.

The Times notes a 2016 series in the Lancet in which researchers estimated that universal breastfeeding could spare the lives of 823,000 children each year and save $302 billion in economic losses. The WHO has long said that breastfeeding is the optimal feeding method for infants and recommends exclusive breastfeeding for the first six months of a child’s life and continued feeding with introduction of other foods up to two years of age.

In the end, the US’s effort to dash the WHO resolution encouraging breastfeeding was largely unsuccessful. Russia ultimately sponsored the resolution and the American delegation did not issue any threats to the country.

Read Comments

Read the whole story
notadoctor
12 days ago
reply
“...the US reportedly put the squeeze on countries backing the resolution by making aggressive trade and military threats... Russia ultimately sponsored the resolution and the American delegation did not issue any threats to the country.”
Oakland, CA
Share this story
Delete
1 public comment
acdha
13 days ago
reply
“pro-life”
Washington, DC

Wet Plate Photography Makes Tattoos Disappear

1 Comment and 2 Shares

Here’s something you may not have known about the 1800s wet plate collodion photography process: it can make certain tattoos disappear in photos. It’s a curious phenomenon that photographer Michael Bradley used for his portrait project Puaki.

“The idea was first sparked when I saw some wet plate collodion images from photographers around the world who had shot people with tattoos,” Bradley tells PetaPixel. “I had been shooting on the wet plate collodion method for a few months and was looking for a long-term project when I saw these images of people with tattoo’s and noticed that some faded away depending on the color of that tattoo.

“I noticed that the green/blue shades looked like they were most likely to disappear, especially on someone with slightly darker skin, and this sparked the idea.”

Bradley decided to focus his camera on the indigenous Māori people of New Zealand, whose traditional tā moko tattoos have been making a resurgence.

Tā moko is different from ordinary tattooing because chisels (called uhi are used to carve the skin and opposed to using needles and puncturing. As a result, the skin is grooved rather than smooth in the tattoo areas.

Bradley realized that when photographs of traditional tā moko were captured back in the 1800s, the tattoos themselves barely showed up at all and where therefore lost to history.

“The wet-plate photographic method used by European settlers served to erase this cultural marker – and as the years went by, this proved true in real life, too,” the project’s statement reads. “The ancient art of tā moko was increasingly suppressed as Māori were assimilated into the colonial world.”

To capture the rebirth of tā moko, Bradley captured two portraits of Māori individuals who have facial tattoos: one with a digital camera to show the tattoos as it looks in real life, and one with the wet plate collodion process to show the same subject without the tattoos.

Bradley’s 48 resulting photos (and the videos that accompany them) are currently being exhibited at the Kōngahu Museum of Waitangi in the Bay of Islands, New Zealand, through September 2nd, 2018. You can also find more of Bradley’s work on his website.

(via Puaki via Fstoppers)


Image credits: Photographs by Michael Bradley and used with permission

Read the whole story
notadoctor
12 days ago
reply
Technology affects how we remember. I examined biased are still biases
Oakland, CA
Share this story
Delete

Photo

1 Comment and 9 Shares


Read the whole story
popular
17 days ago
reply
notadoctor
18 days ago
reply
Oakland, CA
Share this story
Delete
1 public comment
zippy72
17 days ago
reply
This is me...
FourSquare, qv
Next Page of Stories