Hasnain says:

Powerful stuff.

“There were many lessons to be learned from the Iraq war, but this, for me, was the most central: We don’t know what we don’t know, and, even worse, we don’t always know what we think we know. Policymakers are easily fooled by people with seemingly relevant experience or credentials who will tell them what they want to hear or what they already believe. The flow of money, interests, enmities and factions is opaque to outsiders and even to insiders. We do not understand other countries well enough to remake them according to our ideals. We don’t even understand our own country well enough to achieve our ideals.”

Hasnain says:

Reading this chapter was a pretty compelling advert for the book.

“The heuristic here would be to use education in reverse: hire, conditional on equal set of skills, the person with the least label-oriented education. It means that the person had to succeed in spite of the credentialization of his competitors and overcome more serious hurdles. In addition, people who didn’t go to Harvard are easier to deal with in real life.”

Hasnain says:

“The digital divide is only one of many factors driving a shift in who is having children. For most of the 20th century, women with the highest level of education—that is, those with the best career prospects—have had the fewest children. But this inverse relationship between education and female fertility is weakening, and some demographers suspect that it will flatten out or even reverse in the coming decades. In some Nordic countries, it already has. To some extent, this shift simply reflects rising education levels; although it was unusual for women to attend college a century ago, it’s the norm now in high-income countries. But the shift is also spurred by rising economic inequality, in which the digital divide plays a part. “The world seems to be moving toward a situation in which affording to have children is for those who are privileged,” Billari said.”

Hasnain says:

Great read on a bunch of debugging tools.

“These are rough notes for how I debugged it, in case it's useful for someone searching on this topic. I spend many hours documenting advanced debugging stories for books, talks, and blog posts, but many things I never have time to share. I'm experimenting with this "rough notes" format as a way to quickly share things. No editing, spell checking, or comments. Mostly screenshots. Dead ends included.

Note that I don't know anything about Slack internals, and there may be better ways to solve this.”

Hasnain says:

“Caches are based on assumptions. Fundamentally, a cache assumes that there's either some amount of temporal or spatial locality of access (i.e. if Alice is sending work now, she'll probably be sending more work soon, so it's efficient to keep Alice's stuff on deck), or there key distribution isn't uniform (i.e. Bob sends work every second, Alice sends work every day, so it's efficient to keep Bob's stuff on deck and fetch Alice's when we need it). These assumptions don't tend to be rigorous, or enforced in any way. They may change in ways that are invisible to most approaches to monitoring.”

Hasnain says:

Yikes.

“So you can imagine our surprise when we were able to gain complete unrestricted access to the accounts and databases of several thousand Microsoft Azure customers, including many Fortune 500 companies. “

Hasnain says:

Great read.

"A big one, which a lot of people just can't quite believe, is that it enhances motivation. People who are more self-compassionate, when they fail, they're less afraid of failure. There was a study where helping people be more self-compassionate about failure [on a test], later on when they had a chance to study for a second test, they actually studied longer than people who were not told to be self-compassionate. Because, basically, it creates an environment where it's safe to fail, so self-compassionate people are often more likely to try again. They also have more self-confidence, because they aren't cutting themselves down all the time."

Hasnain says:

Amazing read on measurement.

“That's one particular example, but I find that it's generally true that, in areas where no one is publishing measurements/benchmarks of products, the products are generally sub-optimal, often in ways that are relatively straightforward to fix once measured. Here are a few examples”

Hasnain says:

Sage advice. His blogs are always great.

“One final note: there is an important distinction between “working hard” and “maximising the number of hours during which one works”. In particular, forcing oneself to work even when one is tired, unmotivated, unprepared, or distracted with other tasks can end up being counterproductive to one’s long-term work productivity, and there is a saturation point beyond which pushing oneself to work even longer will actually reduce the total amount of work you get done in the long run (due to the additional fatigue, loss of motivation, or increasingly urgent need to attend to non-work tasks that this can cause). Generally speaking, it is better to try to arrange a few hours of high-quality working time, when one is motivated, energetic, prepared, and free from distraction, than to try to cram into one’s schedule a large number of hours of low-quality working time when one or more of the above four factors are not present.”

Hasnain says:

TIL symbol versioning.

“The alpine-glibc project attempts to package the GNU C library (glibc) in such a way that it can be used on Alpine transparently. However, it is conceptually flawed, because it uses system libraries where available, which have been compiled against the musl C library. Combining code built for musl with code built for glibc is like trying to run Windows programs on OS/2: both understand .EXE files to some extent, but they are otherwise very different.”