A “sum type” is a union. A “product type” is a struct. C has had these since before 1970.

If someone uses these terms as though they were some new amazing inventions, they probably do not have much programming experience yet. Usually when kids say “sum type” they mean “automatic syntactic sugar that unwraps a union based on a tag,” but note that this has very little to do with the actual type or how it is represented!

(And ML had this feature since the 1980s, and was not the first to do it.)

I just got back from a meditation retreat run by these guys:



I can recommend that most people would benefit from one of these retreats.

I will explain below ... [1/n]jhourney.io They are teaching how to access extreme states of consciousness called the jhanas, which were once thought to be incredible difficult to attain; but lately it has started to appear that they are learnable by almost anyone.

@realGeorgeHotz Obviously the culture is going to be deeply broken, as when these companies consist of 40,000 engineers, 39,500 of those engineers are not the people who build new things, they are what was hired for, people to do small amounts of work to fill things out a little bit and to @realGeorgeHotz help bulk out the company so its growth rate can be epsilon faster because the company was racing to go public or get bought. The culture of programming has changed because it now mostly consists of people who showed up to the party that was going and that they would not

The people who are very upset about Twitter layoffs do not see the underlying context, that the next 5 years in Silicon Valley is going to be this times N million employees; Twitter is just going a little more preemptive:

When companies are trying to exhibit hyper-growth, the incentive is to bulk up because not doing so might limit your growth rate, and you are taking bulk from your competitors; you're willing to burn lots of present-day cash in order to push that exponential slightly higher.

Note to the 1990s: When it's time to extend the Microsoft PE format beyond 16-bit relocation counts, do not hire the summer intern to write this code. (Or the documentation).

Site of the original Apple Store in San Francisco, a thriving spot during the city’s recent heyday: The commercial space right next to it:

When I was up-and-coming in the games industry, any time you gave me a choice between a cushy job that paid well, or a challenging job that paid less, working on something meaningful or boundary-pushing, I chose the latter, even if it came with unpleasant circumstances. This meant I was poor a lot of the time, but it was the primary way that I improved my design and programming skills.

Re-doing things, that didn't come out well enough the first time, is one of the most effective methods of making good things. If you're working on something new and different, you won't know in advance what is going to be best. You try some things, and learn what will be good. This feels a little different when distributed across a team, but the basic idea is the same as when it's just one person working. All the time I remove things I have done before, and replace them with something better, in both programming and design.

@meglio @StevenBHutton @cmuratori We don't use branches. I try even to discourage programmers from running de facto branches, where they are making a big change that they don't check in for a long time, because that tends to slow things down and introduce problems. Almost all major changes can be made @meglio @StevenBHutton @cmuratori incrementally, and this doesn't cost extra in terms of planning etc, because you get contact with the reality of the rest of the system sooner, and spot problems sooner.
Not the same as branching but related, the #1 way that builds break is when someone has a bunch of changes

The "excuse factory" Casey discusses in this video is something I see all the time, not just about performance, but also functionality, or how many people it takes to program something. The person making the excuses ...

https://twitter.com/cmuratori/status/1411438335931785218

... is pretty much always reflexively starting from the position that what they are doing must be right, and let's come up with rationalizations why it's right. No first-principles thinking is involved (How many instructions can the machine run per second? What is the size of ...

palladiummag.com/2021/05/14/we-… This is an interesting article but I disagree with it on several points. San Francisco was never the capital of “tech” — that was always the South Bay. SF became trendy recently because the south ran out of office space and 2nd-dot-com-boom twentysomethings wanted nightlife.

We're now hiring someone to help work on the compiler and surrounding tools like program visualization and debugging. Help save civilization from the vicious feedback loop of bad software, and accompanying skill deterioration, into which we are getting stuck. I'm looking for a systems-programmer kind of person for this role. Game and graphics experience is helpful but not required. Real experience with complex software written in systems programming languages like C or C++ is required...