sjtindell

Hired 10,000 people in Q2 this year alone? What percentage of those people actually do anything of value besides cash the company’s checks? That’s actually staggering to me. Cloud division I can see growing. What else?

This doesn’t sound right at all. The classic story is “he/she took me to the cleaners”, fighting over every possible asset, being forced to sell a house or property you don’t want to because the spouse demands half. I don’t know the frequency but it happens. My own parents fought hard over the house in particular and my mother ended up having to buy my father out at a huge loss to avoid selling it.

In my mind it is the same problem everywhere. Supply is too low and is kept low for many reasons. There is insane demand from around the world to live in Manhattan.

I would love a Crazy Eddie’s Discount Domains website built using 90’s html where the domains are sold like timeshares. Do I even own this domain? Not clear, but at $0.90 cents a year who can complain. A counter for page visits and Eddie’s cocaine budget.

I have read a lot of financial crime books and can’t recommend Billion Dollar Whale enough. What’s so astounding is the immediacy with which he went on one of the biggest spending sprees ever. A case study in how to disprove people who say, “you could never spend that much money in a lifetime”. The fact he used stolen money to finance The Wolf of Wall Street is the cherry detail on top.

To me, using women to sell a product reeks of a bygone era and a certain mentality we’re working to get away from. I think that alone is a perfectly acceptable reason to think a company sucks and not use their product. Voting for behaviors with your dollars is important.

True, we’re trying to produce bots that reliably do things (make people laugh) that humans can’t even do reliably. People who can feel out a room and use the right joke, or the right reassurance or whatever, are not even very common.

I am a fan of this talk. Working on the sysadmin/devops/sre side of enterprises, I originally found the transition from manual installs to things like puppet and then to docker containers and surrounding tooling refreshing. Building your own services, monitoring error budgets, and all the other principles like infrastructure as code seemed to be “the way”. I have recently become severely disillusioned. On the development side, one need only spend time in the JavaScript ecosystem to get a taste of some of the worst of the baggage we are building for ourselves there. You could spend more time learning toolchains than writing code. On the infrastructure side, it seems we never bother to deploy even a dead simple web service without a full build out of a multi-region Kubernetes setup with ci/cd services, secrets services, auth services, logging, monitoring, pipelines for the data stores and image builds, networking, etc.

At a certain scale these things seem necessary. But I’ve had the most success actually getting things done with small, competent teams where I replaced many different tools with as few as possible. For example I recently turned a complex spinnaker setup into a set of bash scripts that perform helm installs. It makes debugging, knowledge share, and development of the “system” so much easier. Why was the choice made originally, what features does it lack, there are tradeoffs certainly. But I feel we as an industry are losing the thread somewhere.

I often see people holding up StackOverflow and their few servers model as a bastion of sanity. I see them say, “I could run your whole company’s SaaS stack on a single box with systemd”. While I find it hyperbolic, more and more I am considering this not just a casual topic but a vital one.

Sometimes I find people asking - what about disaster recovery, high availability, scalability? When what we need is to slow down and ask, does it work? Is it usable? Or we’re so focused on isolation, deploying security tooling, the systems get complex, and no one understands them to the point they seem much less secure.

Simplicity seems key. Sticking to as few tools as possible, at lower levels, and only growing or adding as really well thought out requirements arise. This isn’t exactly what Johnathan Blow was talking about (abstraction and losing knowledge) but feels strongly related to me.

Personally I think this is an artifact of a time when there were other ways for you to make money.

Republicans are in favor of the DEA and drug scheduling though.

Google maps is hard to replace for me, being out with friends and someone says “here’s my address” and we just map it, too easy.

There is no centralized way to drive that change on the needed scale. Most humans don’t even view what we’re doing in civilization as shared project, let alone agree that this is required to protect it. I think they fundamentally lack the abstract thinking required. They won’t change until they feel the pain themselves. A poor fit for an abstract, not quite “here yet” problem. In the United States we can’t even get them to agree it’s a problem at all. But I think once we are collectively feeling the pain, solutions will be swift and surprising.

I don’t see this at all. They didn’t make less than you because of some moral failing by society at large. They made less because there is such a glut of artists that the bottom pay rate is literally less than zero, hence “do it for the exposure”. There is more art created every day today than ever before. The old school art forms you think are so crucial just aren’t.