feifan

The hard thing about plain-text is that you quickly run into domains/problems that are too complicated to easily represent as plain-text. When that happens, you run into problems: the text representation becomes really verbose (e.g. XML) or exhaustingly terse/unergonomic (e.g. LaTeX, IMO), and you can only build anemic tooling.

I'd love to see the opposite — bigger keyboards, with more keys dedicated to domain-specific functions (something like this[0]). I think that would mean less cognitive overhead in general as you do a particular job.

[^0]: https://twitter.com/mwichary/status/1537631208414191616?s=21...

Back when I had two monitors, I got the same cables for both and I indiscriminately plugged them into my computer (I had a laptop, so plugging/unplugging happened ~daily). So the mapping between computer port and monitor definitely wasn't going to stay the same.

$30B is still A LOT in absolute terms

TIL nitrogen contributes to fertilization

Their other website[0] mentions nitrogen from the air.

[0]: https://solarfoods.com/science/

I used to feel like I "had to blog" and sort of forced myself to come up with topics and thoughts on those topics. I didn't enjoy the exercise. Lately though, I've used my blog to publish essays describing a view of the world that I'm trying to make happen through a personal project, as well as updates on how that project is going. It's been really useful to be able to share a link explaining my thoughts with new people, and it's always exciting when someone, unprompted, shares a link to one of my posts, or starts a conversation with me because of them.

I don't think you need to build a brand — you say that you're already getting organic search traffic, so you can probably just focus on performance marketing (i.e. figuring out what other things people are already searching for and bidding on those keywords). Building a brand is a roundabout way of doing that (get people to know about your brand, and then hope that they search for your brand to find your product).

I am not a marketer though, so take this comment with a big grain of salt.

(Assuming you would want it to be your full-time job if it could get there): how close is it to making enough for it to be your full-time job? If it's profitable, one way to look at it is that you have an unbounded timeline for learning/figuring out the marketing to grow it. You don't necessarily need to raise a VC round for it.

IMO it's perfectly fine to start helping on a bounded part of a bigger project, but then decline to help on further asks/scope-creep.

Everyone's saying to ask for money … which is fine, but also don't forget to be human and make/keep friends. If someone's asking for a little bit of help and you're feeling up for it, by all means help them (and don't feel like you have to bill for the help to avoid being used). If you're not feeling up for it, you can decline ("sorry, I'm busy/I can't fit this into my head right now"). If it's a bigger ask (e.g. traveling to them for a longer engagement), then yes absolutely agree on compensation ahead of time.

I could see myself using this in Lightroom where most features can be triggered by a keyboard shortcut but then manipulated (i.e. dragging sliders) with the mouse. It would be faster to have one hand on a full keyboard and the other on my mouse, rather than moving my mouse hand back and forth or using the mouse to invoke all the features.

It expired a few years ago, hence the startups chasing the space.

Presumably, people who understand messaging

I've been thinking about/prototyping some programming interfaces in this direction. I'd be interested in getting your thoughts on this! https://feifan.blog/posts/tanagram-creating-a-database-based...

I also see their IPO showing up in Robinhood's IPO Access list (i.e. available for purchase at the IPO price)

IIRC, renewing a Tesla registration is $6xx/year whereas renewing a gas car's registration is $1xx/year

I think that's right, but it's worth noting that "until the blood sugar drops" is doing a lot of heavy lifting there. That varies a lot by person. E.g. I wear a CGM, as does a friend. We can eat the same thing, and my friend's CGM reading drops back to his baseline in less than an hour while mine stays elevated for several hours.

I eat low-carb foods and intermittently fast to compensate, but if I ate a "mainstream heatlhy" diet (3 meals a day, grain bowls, etc), my blood sugar would be moderately elevated for most of the time.

Oh I know, I've done basically the same thing and dropped/kept off ~20 pounds. My point is that physiologically it's hard, and by reducing it to just-a-number or just-willpower, it's ignoring very real aspects of life for many people (myself included when I did that).

In technical terms, CICO is a very leaky abstraction.

CICO is probably valid but it's easy to misinterpret. For example, in the presence of insulin, fat and muscle cells take in calories, regardless of the energy state of the rest of the body, resulting in hunger and probably more calories-in. While insulin is elevated, the calories stored wIthin fat cells is inaccessible to the rest of the body … would those calories be considered "in" or "out" according to CICO?

My pet theory is that the thinner 2016 chassis (and/or hotter chips around that time) crossed some thermal/entropy threshold and the heat is causing a lot of the software flakiness … although I have no way to prove that

Nice! I got my MP to replace my 2016 as my personal computer after the latter died, but that was a few months before the W5700X was available. I kinda wish I waited to get that; oh well.

It is a good size for being a footrest! IMO the aluminum and sharp edges don't make it a particularly comfortable footrest though haha

Empirically, the 2016–2020 MacBook Pros seem to be relatively unreliable. My 2016 would frequently have wake-from-sleep problems, occasional kernel panics, and fried its logic board with a pop after about 3 years. My 2019 thermal-throttles every day, kernel panics a few times a month, and has a battery that just gives up under load.

Meanwhile my 2019 Mac Pro (tower) is probably the most rock-solid reliable computer I've ever used. Months of uptime, never has sleep/wake problems, and only one kernel panic in the 1.5 years I've had it.

I've used iCloud Keychain for all my personal stuff since it existed. It's pretty bare-bones, but the built-in integration into Apple's software makes it a no-brainer for me. It's worked great!

To be fair, I think that's more attributable to the secular trend around car prices recently (car prices across the board are up because of chip shortages and buying demand from rental companies) rather than anything specific to Tesla.

But, re "there is so much misinformation about Tesla it is unreal" — I agree, but that doesn't mean you can just categorically ignore _all_ information about Tesla.

> The battery has a life of ~120k miles …

This appears to be untrue. High-mileage Teslas are showing about 10% range degradation after ~250kms on the odometer: https://electrek.co/2020/06/06/tesla-battery-degradation-rep...

Notably, this isn't about Tesla specifically — the broader point is that the technology exists for mass-market car batteries to support many hundreds of thousands of miles/kms within a reasonable lifespan of the battery.

You're making a stock-vs-flows comparison: a battery is manufactured ~once for the lifetime of a car and then produces ~zero emissions per marginal mile driven. Combustion engines produce emissions for every mile driven (and require some emissions to produce as well).

IIRC, the average price of a car last year in the US was around $41k. There are many options whose MSRP is under that price, e.g. https://www.myev.com/research/buyers-sellers-advice/comparin...

IMO it depends on what part of the ruby feel you most enjoy. Sorbet makes it harder to work with the runtime-magic parts of Ruby (and in some cases we rely on tooling-generated code instead of runtime dynamic stuff), but it doesn’t affect the code you’d write within methods all that much.

(Disclaimer: I work at Stripe and use Sorbet)

IMO the Sorbet additions do impact readability, but it dramatically improves understandability. In a codebase past a certain size/number of contributors, you're always going to have ambiguity about what sort of data you're working with. For example, in my part of the codebase we have models representing an entity that is identified by a number in external systems and represented as a whole data object in our code. If I have `foo.entity`, it's unclear just from the "real" code whether that'll return a data object or just the numerical identifier. Having the Sorbet annotations on that method (and available on hover in the IDE) make it a lot easier to navigate the codebase and resolve these ambiguities.