HNHacker News
TopNewBestAskShowJobs

jkingsbery

2,170 karma · joined January 19, 2010

submissionscomments
jkingsbery··on Is This Prime?
I think it's called "The Ridiculous Fish Theorem."

+1... I lost two games in a row, both times on 91.

jkingsbery··on Generating JSON Directly from Postgres
"Your scientists were so preoccupied with whether or not they could, they didn’t stop to think if they should."

While interesting, the advice offered in this post is generally bad, or at least in complete. OK, so, you "cut out the middle tier..." what now? Are web clients connecting to PostGRES directly? Will PostGRES handle connection authentication and request authorization? Logging that a particular user looked at particular data? Can the client cache this data? If so, for how long? Even taking the author's premise, this is not a good pattern - a simple JSON get request still comes with a bunch of other stuff that this doesn't bother to address.

But the premise is wrong - few applications just spit out JSON like this, they also have to handle updates, and there's business logic involved in that. Data needs to be structured in a way that's reasonable for clients to consume, which isn't always a row of data (maybe it should be collapsing all values in a column, so that in the JSON the client just gets a JSON array of the values).

jkingsbery··on The Case Against the Covid-19 Lab Leak Theory
> WIV officials told the investigators that serum samples for all staff and students in the bat coronavirus group subsequently tested negative for Covid antibodies. We have only their word to go on because the lab hasn’t been independently audited...

Why has the lab not been independently audited? If I was going to make a case against someone making an error, saying "well, that group said they didn't make an error" seems like a bad argument.

> The first human case of Covid could have been infected hundreds of miles away, perhaps closer to the horseshoe bat caves of southern China.

Have we found such a person?

> You can get on the fast train in Wuhan and be in Guandong Province, the home of the original SARS outbreak, in under four hours.

If someone had been in Guandong, and taken a train to Wuhan, it seems to me that the early cases would not have been so concentrated in Wuhan, for two reasons: 1) if Wuhan is really all that accessible by different means of transportation, you'd expect people would just as easily take it from Wuhan to other places, and 2) presumably, the fast train doesn't just stop in Wuhan (Google says it's 1000 km away), meaning an infected person on the train could easily have infected someone else on the train who got off in between. Honest question: is that not how the spread of disease would work?

> There is no evidence that the Wuhan Institute of Virology, or anyone else, ever had any strain that similar to Covid-19.

Again, that's what we're trying to figure out. You can't take the thing we're trying to figure out as evidence one way or the other, that's Begging the Question (https://en.wikipedia.org/wiki/Begging_the_question)

> Given that Covid (or its direct ancestor) must exist in nature, it’s more likely that it got out naturally (like SARS and MERS) than that it took an undetectable detour through a secure biolab.

I don't understand this jump. Most people I've read arguing that the lab leak theory is possible agree that there must be some base virus in nature. The question is, what is the more likely path from that natural virus to the COVID outbreak, through a lab or through some other animal (or animals)? I understand the argument that we didn't find the transmission chain for SARS or MERS or Ebola for years, but why does asserting it started in nature lead to the conclusion it therefore had a natural path? This again seems like the author is trying to prove a conclusion using the conclusion.

> The fact that China is being secretive about Covid-19 isn’t evidence for any particular theory.

When SARS came out, I don't remember the Chinese government blaming Italian frozen food or athletes at an international athletics competition for SARS. I don't remember them promising people that it's not transmittable between humans, or that it's not a big deal. I think simply saying "China is being secretive" is not looking at the whole picture.

> All theories of the origins of Covid-19 should be investigated, including lab origin theories. We should go wherever the science takes us.

This seems right to me. We can't make bricks without clay though.

jkingsbery··on YouTube’s rules are used to silence human rights activists
I personally don't know what to do about it. But it seems like it's worth establishing what at least the truth is before we do something (either by deciding to, or by deciding not to do anything).

I wish it were so that everyone knew that "Communism is a tortuous system," but it unfortunately does not seem to be the case anymore. People who were born and never lived in a world with the Soviet Union or the Berlin Wall are now voting age, and they don't have an aversion to communism. Even before the Soviet Union fell, a lot of people in the west were sympathetic to the Soviet Union. Fortunately we had people like Solzhenitsyn who documented how bad it was. Atajurt and the Uigher witnesses are giving us what may become our generation's equivalent of "In the First Circle."

jkingsbery··on Ask HN: What software do you use to gather requirements?
What a lot of teams in my company do is have a less formal part (more like brainstorming) done with Quip (https://quip.com/) before having the more formal part in Amazon WorkDocs (https://aws.amazon.com/workdocs/... Disclaimer, I work for Amazon). Workdocs is a pretty good tool for versioning, commenting on and sharing Word documents, but it's not great for multiple people working on a document at the same time. Quip is great for multiple people working on a document at the same time, but doesn't really have any versioning mechanism.

For personal stuff, I use emacs org mode in a git repo. Org mode lets you define a hierarchical structure to the topic, hide the hierarchy that's not relevant to the part that you're focused on now, track what things you've done, and when relevant provide supporting information in spreadsheets embedded right there. Worth checking out.

jkingsbery··on YouTube’s rules are used to silence human rights activists
> Atajurt has collected thousands of video testimonies from family members of Turkic Muslims who have disappeared in Xinjiang. Witnesses show their identification to prove they are real people.

Given that these people are giving their testimony to Atajurt (and not the Chinese government), these are people trying to share their story so the rest of the world knows how bad things are. As the article says, this knowledge is either shared directly (the videos themselves are shared), or indirectly ("the information in the videos is then used by other organizations..., which documents [sic] where detentions are occurring, which communities are most affected, and who has disappeared."). YouTube is therefore preventing one of the most oppressed groups in the world from explaining how they are oppressed.

That sounds pretty bad to me.

jkingsbery··on Europe's Software Problem
If people are willing to create a ton of spam for a t-shirt, I would imagine that a lot more spam would be created if contributions were rewarded with more money. Or at least, it's not immediately clear what mechanisms would exist that would prevent spam pay-for-contributions that also wouldn't work for t-shirt contributions.
jkingsbery··on Software Developer Shortage Is Coming
> These graduates enter either a managerial track, such as Project and Program Manager or Staff Software Engineer, or they enter a technical track, such as Senior Software Engineer and Software Development Engineer II and above.

At least at my company, Principal Software Engineer is considered a technical/individual track role, and my assumption was that Principal/Staff Software Engineer at most companies were. Is that not the case?

jkingsbery··on Boring Problems Need Attention Too
This is an excellent answer. The original post, insinuating that engineers are demonstrating "engineering mediocrity" or not "correct," is lazy for not understanding why this problem is ill-defined.
jkingsbery··on Europe's Software Problem
Right. Last I read, Hacktoberfest created a huge amount of spam for Open Source contributions, and all that was on the line there was a t-shirt. You seem to be suggesting that real money should be involved - do you think the outcome will be different than Hacktoberfest?
jkingsbery··on Some Americans are breaking out of political echo chambers
I've done something similar. If you look at what, say, the National Review thinks is important on a given day, and compare that to what the NYTimes is reporting on, it's pretty clear that we're not merely disagreeing about a particular set of facts, we're living on different planets.
jkingsbery··on Some Americans are breaking out of political echo chambers
> Until you can passionately make arguments for both sides, you don’t understand the issue.

Russ Roberts, host of the podcast Econ Talk, is really good at this. Independent of the guests political background, he does a good job of offering debate using a charitable interpretation of other people's arguments (even ones he doesn't disagree with). He also is pretty good at trying to find common ground with people he disagrees with, and when he realizes he's made a statement that would demonstrate his bias, he's pretty quick to call out his own bias.

jkingsbery··on Ask HN: Would Spreadsheets with Lisps be much more powerful?
If you're interested in this topic, you should look into emacs (a text editor that uses a Lisp-like language for customization) and it's accompanying org-mode (something that helps you take notes in emacs, and also has a built in spreadsheet), which allows for using emacs lisp for defining functions: https://orgmode.org/worg/org-tutorials/org-spreadsheet-lisp-...

Good luck!

jkingsbery··on The Queen’s Latin or Who Were the Romans? Part I: Beginnings and Legends
> Many students are more than a little surprised to find that the actual contents of Latin literature are often rather less elevated than they might have expected

Our college Latin teacher was trying to get us to translate Catallus more faithfully, especially after one student gave a particularly polite translation. It was a moment in my career as a student I'll never forget - not only using pretty graphic language in class, and having a professor respond "Right, good."

jkingsbery··on French set to replace English as EU’s ‘working language’
In case there are any other French-reading Latin-philes out there: https://www.lefigaro.fr/vox/culture/et-si-la-langue-officiel...
jkingsbery··on The pedantic checklist for changing your data model in a web application
For whatever it's worth, a pretty good book that covers this in a lot of detail is https://databaserefactoring.com/. One of the authors had several notes on his website about this prior to putting them into a book: http://www.ambysoft.com/books/agileDatabaseTechniques.html
jkingsbery··on National Geographic now recognizes The Southern Ocean
My very eager mother just served us nine pickles.
jkingsbery··on Keep Your Identity Small (2009)
I want to agree with a lot of this, but I think there is a lot that is more complicated than this essay allows for.

Consider in the footnote:

> There may be some things it's a net win to include in your identity. For example, being a scientist... A scientist isn't committed to believing in natural selection in the same way a bibilical literalist is committed to rejecting it.

The comparison of "scientist" to "biblical literalist" is, when you think about it, sort of an odd one. Creationists get a lot of press, but strict creationists make up a minority of religious people (depending on how the question is asked, something like a third of Evangelicals believe in Creationism according to https://blogs.scientificamerican.com/observations/how-many-c..., and even less of other Christian religious groups). It's kind of like comparing Thomas Aquinas to the pseudo-scientist Aliens guy from History channel. After all something like a third of all US adults believe that some UFOs have been alien space craft (https://news.gallup.com/poll/350096/americans-believe-ufos.a...... meaning many more Americans believe that we've been visited by aliens than believe in Young Earth Creationism). At the same time, Aquinas's Summa was structured around surfacing objections to religious hypotheses of the day.

But whether the "scientist" label really is as Graham says is debatable. For one example: I wrote a paper in college for a class in Artificial Intelligence, where we were assigned to pick a problem and write about how people solved it. I wrote about some AI approaches taken to learning the suffixes of English words. As I dug, I found that there were two really well entrench camps between Connectionist (people who favored approaches using Neural Networks) and Symbolist (people who favored approaches like decision trees). There was a several-years back and forth between the two camps, and they were pretty clearly not "following the evidence wherever it leads." Whenever the authors were comparing their own algorithms to the other side's, they used the best version of their own but a mediocre version of the other's algorithm. Perhaps though this examples serves to bolster Graham's larger hypothesis, that people don't like to consider things that go against their identity, as the scientists in question clearly identified as Connectionist and Symbolist camps.

I don't know how pervasive it is among scientists, but given the frequency it comes up in the EconTalk podcast as the host talks with academics, it seems like it is pretty common for scientists, being humans, to fall into the same trap of identifying with their own work.

On the other hand, many (not all, but many) of the theologians I've read actually do a pretty good job of following the evidence where it leads. Or, in many cases, where it led them, as many of the most compelling theologians are converts. CS Lewis is a pretty famous example - and his book Surprised by Joy is a great account of simply following the evidence. Thomas Merton is another good modern example, as is Peter Kreeft. Of course, St. Augustine's Confessions is still the archetype of this conversion story. So I don't think it's as simple as "religion = labels, science = pursuit of truth" in the way Graham portrays it.

> If people can't think clearly about anything that has become part of their identity...

Graham does make note in the opening that people often make arguments for which they have not much evidence, but there is a gap between "people can't think clearly" and "people don't think clearly." There are plenty of examples of people thinking clearly about things where are part of their identity, as there are plenty of examples of people including something in their identity because they've thought clearly about something. One of my favorite quotes from GK Chesterton is "the object of opening the mind, as of opening the mouth, is to shut it again on something solid." Yes, some people use labels as security blankets. But plenty of people apply labels only after a long-fought intellectual struggle. Chesterton is himself an example of this, as he describes in his own writing how he embraced the label "orthodox" after a circuitous route of trying to discover truth for himself.

jkingsbery··on Ask HN: What is the most bloated website you use
Every team I was a part of between 2007 and 2014 used Jira, and I pretty much haven't used it since. It was notoriously slow back then.
jkingsbery··on Tech Interviews Are Stupid
As others have mentioned, this is more a sign of bad interviewers than a bad format. Of the companies I've worked at, at least one has allowed googling during the interview, and most de-emphasize all the stuff you have to look up in order to focus on the problem-break-down part of the question.

I think there's too much imitation without understanding in interviewing. Candidates hear questions, and then go on to repeat them when it's their turn to be interviewers, without really understanding what's behind the question. These new interviewers then approach the interview as if they were administering a coding test, whereas what they ought to be doing is facilitating a conversation using the question as a starting prompt.

The other problem that I see sometimes with interviewing is interviewers asking obscure questions. Really good candidates aren't solving the problem on the spot - they've come across problems like the one the interviewer asks, and are using approaches they've already validated in their life experience. When an interviewer asks an obscure question, they are not helping to facilitate a conversation as to whether the candidate can handle the typical things that would come his or her way.

jkingsbery··on Wuhan lab staff sought hospital care before Covid-19 outbreak disclosed
I can't say I've ever read OANN or NewsMax, even though I am conservative. I have read the New York Times before though, and as the linked article and several others by the same journalist indicate, evidently the Times is selective about showing differing views.
jkingsbery··on If the Lab-Leak Theory Is Right, What’s Next?
As the Geraghty article I cited mentions, the Wuhan Institute of Virology leak theory has at best circumstantial evidence, and very well might not be true. And in any case, the claim is not that it was intentional. While not impossible, there is currently zero evidence today that the virus came from anywhere else. There is zero evidence of any cases anywhere before those in China (indeed, before those in Wuhan). China is not known for having lax border crossings, making smuggling in vials unlikely. Wuhan is not close to a border, if someone came to Wuhan already infected, there would be a trail of infected people from the border to Wuhan, or other people would have been infected on a plane if a human carrier brought it in. If someone brought it in during the World Military Games as you seem to think in other posts, the timeline doesn't make sense, since the first cases of COVID weren't for at least a month after, in November (and besides the possible cases by the people who work at WIV, most cases started in December). Your hypothesis does not fit the data.

So I ask you: what evidence, even circumstantial, do you have for making the claim you do?

jkingsbery··on If the Lab-Leak Theory Is Right, What’s Next?
There probably were conspiracy theorists doing just that. But there were also articles like the one by Jim Geraghty (https://www.nationalreview.com/2020/04/coronavirus-china-tra...). Compare what he wrote to what is in the recent Atlantic article.

Geraghty:

> And none of the above is definitive proof that COVID-19 originated from a bat at either the Wuhan Center for Disease Control & Prevention or the Wuhan Institute of Virology. Definitive proof would require much broader access to information about what happened in those facilities in the time period before the epidemic in the city.

> But it is a remarkable coincidence that the Wuhan Institute of Virology was researching Ebola and SARS-associated coronaviruses in bats before the pandemic outbreak, and that in the month when Wuhan doctors were treating the first patients of COVID-19, the institute announced in a hiring notice that “a large number of new bat and rodent new viruses have been discovered and identified.” And the fact that the Chinese government spent six weeks insisting that COVID-19 could not be spread from person to person means that its denials about Wuhan laboratories cannot be accepted without independent verification.

Engber:

> Arguments in favor of the “lab-leak hypothesis” remain grounded, as they ever were, in the mere and highly suspicious fact that a coronavirus likely borne by bats, likely from a cave in southwest China, emerged 18 months ago, quite suddenly, in a city very far from southwest China—where researchers had assembled an archive of cave-bat-borne coronaviruses.

As Geraghty recounts yesterday (https://www.nationalreview.com/the-morning-jolt/even-xavier-...), many in the media, rather than assessing the facts themselves, just resulted to name calling against a conservative-leaning but measured writer. I also haven't seen Geraghty (and again, other more measured writers) claim to have definitive proof of the lab-leak theory (as some conspiracy theorists might claim), only that it is a possibility that shouldn't be dismissed out of hand. It is illogical, irresponsible and unscientific to say "well, this theory was picked up by racists, therefore we should dismiss it" rather than evaluating a possibility on it's own merits.

jkingsbery··on Wuhan lab staff sought hospital care before Covid-19 outbreak disclosed
For a contrasting point of view, here's an article from a conservative, religious writer who follows how the media covers religion stories from 2014. In this article (and is a common theme in his writing), he discusses "Kellerism" a term named after Bill Keller, a former New York Times editor who said in 2011 how the New York Times does not seek balance in a whole bunch of areas. The author, Terry Mattingly, has over the years documented many cases in which CNN, NYT and other similar outlets don't cover stories that intersect with religion or religously affiliated people or communities fairly (seemingly, in many cases, because the authors covering the stories lack the basic knowledge to even raise questions allowing for balance).

https://www.getreligion.org/getreligion/2014/8/2/this-keller...

jkingsbery··on Wuhan lab staff sought hospital care in November 2019
I would agree that, yes, many of the things you raised are things we'd look at regardless, but in any "5 Why's"-type investigation, you have to get to the root cause. Whether this came from a lab will probably have a huge influence on the future of Gain of Function research, both how it is conducted and whether we as a global society think it is a good idea. If on the other hand COVID was a natural occurring disease, but we have been unable to figure out the path from its origin in animals to eventually infecting humans, then we will probably need better techniques for tracking such transmissions. I think everyone knows that bats carry diseases, but most of what I read has indicated that in the natural scenario there was probably an intermediary animal, and knowing what animal that is surely would change how we approach that animal and its habitat in the wild.
jkingsbery··on Nonnative English speakers share their gripes about speaking English
The first time I heard/read "prepone" and "do the needful," it was maybe jarring because I had never heard either of those before, but in both cases it was pretty obvious what they meant.
jkingsbery··on Ask HN: How do I learn more about entrepreneurship/startups?
Others have mentioned "learn by doing," which I think is absolutely necessary. At the same time though, it doesn't help if you start off doing it wrong, bringing assumptions about how a company works that are based on big-tech.

Having spent a decent chunk of my career at both start-ups and at a Big Tech company, there are lots of things that make sense to do at one that don't make sense at the other. As you make the move yourself (or, if you hire others, as you onboard people who are making that move), it's important to keep in mind how the risks are different between large companies and small companies. That probably doesn't answer your question directly looking for books or resources, but I think it's important to keep in mind as you consume books or resources.

jkingsbery··on Ask HN: How do I learn more about entrepreneurship/startups?
Others mentioned MicroConf. The organizer of MicroConf, Rob Walling, wrote a book where he gives very similar advice: https://www.amazon.com/Start-Small-Stay-Developers-Launching..., both around starting with a market and focusing less on technology and more on sales. He is writing for a particular niche of smaller, non-VC-backed startups, but the general approach still seems similar.
jkingsbery··on Robinhood’s big gamble
Did anyone else on reading the headline picture Cary Elwes asking "An archery contest?"
jkingsbery··on Ask HN: How to negotiate continuing to work remotely?
Employment lawyers are helpful with, well, employment law, which companies need to comply with. They don't help with everything though. Employment lawyers don't understand what I want and what trade-offs I'd be willing to make to have those things happen, nor do they, as employment lawyers, have insights into the specific trade-offs my employer "I want to have a different working arrangement" isn't a legal question.

I recognize my good fortune. However, on a site called "Hacker News," I'm going to assume that most people here work in technology and have roughly similar leverage to what I have or have had in the past. If your employer doesn't treat you well and you work in software development, IT or a related field, you should leave because there are employers out there that will treat you decently. None of us are entitled to one specific job.

Unions make sense in some situations. Evidently people in the technology sector have been able to work this specific question out without them.

← PreviousPage 6 of 21Next →