Huh I noticed this section in the linked NACS post:

> As NACS is now recognized in a SAE recommended practice (RP) under SAE J3400, we have removed the technical specifications and CAD from our website.

So something that was previously freely available now requires a $300 payment to access.

I'm sad to see that.

I don’t think the SAE has the authority to remove anything from the public domain?

It likely is still freely shareable for existing copies.

See also a novel w/o letter E: https://en.wikipedia.org/wiki/Gadsby_%28novel%29

So: https://en.wikipedia.org/wiki/Greenspun%27s_tenth_rule ?

Exactly this. See also the profusion of filter expression languages, eg here's Algolia's: https://www.algolia.com/doc/guides/managing-results/refine-r...

Or here's one built into Symfony: https://symfony.com/doc/current/components/expression_langua...

Copy on selection is ubiquitous in all terminal apps. Shift+insert can be a reliable "paste" shortcut in most places.

Shift+ins is CUA, an actual standard. Which makes far more sense than parroting Microsoft’s decision to use CTRL as a modifier, aping Mac OSs keyboard shortcuts in Windows 3.1.

That’s fine for paste, but copy is the real problem. Also, shift and insert are very very far apart from each other and are not one-handable if you’re holding the mouse with the other hand.

Copy is Ctrl+Ins, cut is Shift+Del. Looking at my ANSI keyboard, all are achievable with one hand. Though, when these were designed, GUI’s on IBM based machines weren’t really a thing and two-handed keyboard shortcuts were common-place.

With a right handed mouse it’s hard to do Ctrl+Ins with my left hand. But it’s possible, I’ll admit. It’s just pretty suboptimal given how ergonomic Cmd+C is on a Mac keyboard (thumb on Cmd is way easier to reach for than pinky on ctrl, and reaching for ctrl+Ins is harder than either.)

Yeah, mice weren’t really a thing on the early IBM AT’s and clones when CUA was being developed. Mice were around and the GUI was very rudimentary (Windows 2.0), certainly when compared to System 4 and 5. In an ideal world, the Meta key, like those found on Sun keyboard, would have been brought over to AT keyboards to cover this.

TIL https://en.wikipedia.org/wiki/IBM_Common_User_Access

I can’t stand X11-style copy-on-select/middle-click-paste, and disable that everywhere if that’s what you’re referring to. I habitually highlight text as I read it and hate it when it blows away my existing copy buffer. I want explicit copying, I just don’t want it to collide with SIGTERM.

At least Google's offices have nice perks: free gym, healthy and nutritious meals, healthy and tasty snacks. Personally those perks are enough to negate the terrible commute.

I was told Amazon's offices has none of these.

It's sad https://pkg.go.dev/embed was not mentioned in a post about web development in Go :-)

Having a true single binary bundling your static resources is so convenient.

Massively underrated. It's actually used to build the pkg.go.dev website itself.

https://github.com/golang/pkgsite

it does mention it:

there’s just 1 static binary, all I need to do to deploy it is copy the binary. If there are static files I can just embed them in the binary with embed.

It is mentioned now.

Which time golang read file, build time or run time ?

embed package allows to embed assets directly into the binary, so the files are read once during build time and then you can access them as e.g. a byte slice, a string, or a special FS object that acts like an in-memory file system

Build time

Build time. It literally embeeds the file in the binary.

Sounds like you can generate a folder of files from an input excel using FUSE.

Wait what do you mean? They can have an HTTPS server and MITM, but how can they get a certificate for the DoH server I use?

So, like, you are assuming someone using a resolver that ignores the certificate chain of trust, as an evidence that DoH is not useful?

Do your program language _show_ you the certificate information when you use an http library to connect to an HTTPS service?

Sure the other end of the DNS query may not be encrypted, but I can easily decide which government to trust, and run my DoH server there.

> your browser does not show certificate information for the DoH server.

It doesn't show it, but I expect it would put up an error message if the DoH server's cert is invalid.

Call me snarky but instead of just multiply/devide by 1.6, you rather prefer remembering the Fibonacci numbers up to, I don't know, 200 to be useful?

In practice, at least in my experience, only up to 13. (Though, I have the sequence memorized up to 21 because of a sign outside Chattanooga[0], which is what gave me my "hey, wait" moment about this.)

A combination of both the fundamental theorem of algebra and Zeckendorf's theorem has allowed me to fill in the rest so far. For example, 25 mi = 5 * 5 mi, which yields 5 * 8 km = 40 km. As it turns out, that is how far Cleveland, TN, lies from Chattanooga.

0: https://usma.org/metric-signs/tennessee

>A combination of both the fundamental theorem of algebra and Zeckendorf's theorem

How do you apply the fundamental theorem of algebra here?

I will go out on a limb and assume they meant the fundamental theorem of arithmetic, that all numbers can be (uniquely) factored into primes.

The list isn't that long. 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, and 144.

I memorized this almost by accident. I was doing hand rolled spaced repetition system for conditioning myself to fix some bad habits, and they came up often enough that it was memorized.

Yes, but remembering the multiples of 10 is vastly easier: they are 10, 20, 30, 40, etc. Remembering the multiples of 16 is quite a lot easier: 16, 32, 48, 64 etc. You probably already know them.

Now to convert from miles to km replace a multiple of 10 by a multiple of 16. 70 mph becomes 112 km/h.

To convert in the opposite direction do the opposite. 130km/h is 128 km/h + 2 km/h = 80mph + 1 mph (rounding down since you don't want to have to justify this calculation at the side of the road, to a gendarme, in a foreign language).

1.6 km = 1 mile is just as accurate as using 1.618.., the golden ratio. (Enough for driving, not enough for space travel.) And using the Fibonacci method is less accurate than the golden ratio since small Fibonacci numbers are only approximately the golden ratio apart.

The only possible justifications for the Fibonacci method are:

1. You want people to know that you know what the Fibonacci sequence is.

2. You enjoy overengineering.

3. You're one of quite a few people who believe, for whatever reason, that the golden ratio appears all over the place like in measurements of people's belly buttons, the Great Pyramids, and so on, and that this has some spiritual or mathematical significance.

Just spend a few sprints in a sufficiently dysfunctional agile team and you’ll learn all sorts of higher Fibonacci numbers.

Technically they could also also be functional teams with a long history of gradual point drift... Though most organizations shake teams up before then.

If a ding that gets mentioned in planning is 1 point and they regularly try to fit 144 point projects in a sprint without splitting it up, that's likely to be rather dysfunctional.

I mean it is possible that gradual point drift would get there. But sufficiently improbable that I know what I'm betting on.

For some reason we are using 20 instead of 21. I didn't even bother asking why

That is because fibonacci is used to indicate the ‘guess’timation part. 5 Instead of ‘twice as much work as a 2’ (4). The idea is when you get higher numbers the opposite happens: 21 sounds like a very specific number and not like a rough guess. 20 does.

So it becomes 1 2 3 5 8 13 20 40 inf ?

By that logic shouldn’t people just do 1 2 3 5 10 15 20?

It's art. It's possibly not very useful (who knows, though?), but some will find beauty, joy, entertainment, surprise, amazement, incredulity, anger, disdain, indignation, fear, nostalgia in this. Others will feel the deepest and purest indifference.

I'm even lazier, 1.6 might be hard to do without a Calc, I just grab 50% of the original number, 10%, and then sum them to the original

In the same vein, converting pounds to kilos is "divide by 2, subtract 10%".

no way just discovered this lmao i now only need a similar thing for fahrenheit to c/viceversa, i'm sure it exists. however, i'm so lazy i will not search for that information now

For Celsius to Fahrenheit, just double it and add 30. It gets you pretty close for the temperature range that humans live in.

That's how I do many kinds of percentages... like 72% is 50% and then 10% twice and then a couple of 1%s

Here’s a practical solution based on this:

Take your kilometres e.g 60, divide by 5 -> 12, now times that by 3 -> 36.

Take your miles e.g 80, divide by 5 -> 16, times 8 -> 80 + 48 -> 128

So any conversion reasonable conversion can be done in your head with your 3, 5, and 8 times tables

Yea, and since it's the golden ratio, the division is roughly the same as multiply by 60%. So you can just remember to take 60% and add it to the original or treat it as the answer depending on if you're going to or from the smaller measurement.

That's what I was thinking. I good example of over-engineering though.

Some people think differently. Some methods click with some people, while making others think it's making it overly complicated. The new math drove people crazy. The concept of doing extra math by rounding a number up to a whole number, and then subtracting the diff seems like a lot of work, but is amazing when used frequently to do "in your head" type of work. That extra match can lead to an answer faster than traditional math

How often you drive 200mph??

I think it's more common to convert distances.

I consider any distance reachable with one tank of gas commonly used (so up to 400 miles).

It would be great if the results can be filtered by the CPU generation. I'm specifically interested in replacing my 8th Gen mini PC and won't want to buy anything older than 8th Gen.

Seconded. You can do this on eBay by choosing specific CPU models, but it would be really helpful to filter more broadly. "Intel 8th gen" for sure, but it could also be useful to be able to combine that with, say, "Intel -T series" or "Intel i5".

I've added keyword filters so you can include/exclude keywords.