throwaway9870

Not custom copies, but a locked down cache of packages. For Gentoo you can do this by locking the portage tree you use and keeping a copy of the distfiles from the first run, for Python it was a requirements.txt file with a cache of the tar files from PyPi, for go it was including 3rd party code in repo. I don't know what the team did for npm.

It was really nice doing a full image rebuild and knowing the only thing that changed it was you explicitly changed.

If there are too many packages to pin, then I assume there are too many package to audit changes in for each image build you do? If that is true, how do you have any confidence that the image is going to function correct?

That makes sense. I think the real issue isn't Docker vs Nix, it is that some package managers are almost impossible to use to build reproducible images. I worked with debootstrap 10+ years ago trying to script reproducible builds and found it exceeding hard. Gentoo made it almost trivial (Google used it for ChromeOS so perhaps they felt similar). I will look into Nix.

It appears that with the proper package manager support, Docker would be fine?

I come from a hardware background and seem to be a lot more paranoid than most software folks. I would struggle to trust a build where so much is not pinned.

I guess that would not fit my definition of "fully".

Do people really build images this way? It sounds completely insane to pull packages like that randomly from the Internet.

I can't imagine using a package system for building images that doesn't allow a full local cache of exactly the packages I want installed.

> If you run docker build twice with the same Dockerfile you might get 2 images that behave in different ways. For example, a third-party package could silently be updated and cause breakage. Aggressively pinning dependency versions helps, but doesn't completely prevent this issue.

If "aggressive" means fully, then why doesn't that fix the issue?

That is exactly what I want to do for the same reason. Any chance you would be willing to share your JS code?

Where did I say it should not be revisited?

> Do you have an actual citation for this?

I would start reading here:

https://www.epa.gov/regulations-emissions-vehicles-and-engin...

No. Modern engines with Tier IV compliance are very expensive and it is hard for even large companies to develop, tool and manufacture a wide variety of engines required by diverse product lines.

> If Americans shouldn’t have a say in California and Californians shouldn’t have a say in Palo Alto

"A say" in what subject? Currency used? Sales tax paid? Laws regarding civil rights? Laws regarding non-compete clauses in employment contracts?

We have a complicated legal and governing system that has worked out which subjects are covered at which levels of government. This is pretty much the same with all developed countries. Debating it in black and white terms as you present is not a very reasonable way to look at the issue.

When the exchange rates go crazy because nobody wants dollars. That being said, the USD is the key currency of the world so we have been able to avoid that issue. If the US causes the world to look away from the USD, it is game over for the US lifestyle. Screwing the rest of the world by printing USD is a bit like the Norway oil fund - it is an amazing source of wealth that will not last forever and should be spent wisely. The US is doing anything but that.

Yes, look up tier 4 diesel regulations. They have had a significant impact on the industry.

This is a common enough technique that linux standardized it to some degree: https://man7.org/linux/man-pages/man2/eventfd.2.html

I don't think anything you describe has to do with semi-conductor issues. MBAs didn't outsource semiconductor manufacturing to China to save money. First, most isn't done in China anyway, second, the price of Fabs became prohibitive and the US govt lacked any strategic plan for semiconductors. Companies had no choice but to go fabless. When IBM, AMD and others started divesting themselves of their fabs, it would have been a great time for a national conversation on the future of that industry. Finally, the cost overhead of the US is just huge, so it isn't just trying to save money it is about being in business or being out of business. You just can't compete in almost any industry if you don't take advantage of cost savings available in other areas of the world.

It is going to get worse with the new stimulus bill as companies now must compete against the government for resources more than ever. You need manual labor? So do all the contractors doing govt jobs. You need engineers? So do all the companies with billion dollar energy contracts. Nothing wrong with this long term, it is going to be the sudden jolt that causes a lot of issues.

The conviction of your opinion seems out of line with your understanding of numbers behind it.

Many years ago, I worked on a serial link (think SATA, PCIe, USB 3, etc.) that had a small embedded processor to self test the link, perform analog component calibration, link equalizer calibration, and even make an eye diagram for link characterization. We could patch it with a few instructions over the JTAG port. There are way more processors in ICs than you would ever imagine.

Then that is a great job for other unmotived people. For motivated people, the poor interview is a great signal to look for a different company.

My take is that too many people are telling others how to interview. It is their money, their company, their job, their life. Let them do it the way they want as long as they follow the legal requirements.

If you think you have a better way, tell us about it without passing judgement on others. Tell us why it works, how great it is, give us some qualitative and quantitative data. Most of all, enjoy your competitive advantage. But for god's sake, quit telling other people they are doing it wrong.

It is embarrassing for the country that a CEO from a company had to discover and report this rather than someone from the govt doing their job. When I see people write "late stage capitalism", I actually wonder if we aren't in "late stage US govt". They had it too easy for too long and have forgotten how to do their jobs.

I have a pasture 20ft from my barn and no way would I want that as close to my house as my turf is. Mice, ground hogs, ticks, a gazillion insects, ground hornets, all sorts of weeds including poison ivy, wild animal poop you can't see because the vegetation hides it, so you step in it. You can't walk through it in summer in shorts unless it is mowed like grass. Forget walking barefoot.

What you describe is not replacing turf grass with pasture, which is what the parent described that I responded to.

You don't want that near your house unless you like pests and rodents.

Turf lawns started out being grazed, but very quickly transitioned into being maintained manually by humans. It took a while before became common because care was so time consuming and obviously expensive.

A turf grass lawn is incredibly practical. It aids with the control of bugs and rodents, it is easy to walk and run in without twisting an ankle because of a hidden hole, it great for sports like soccer, bad mitten, corn hole etc. You don't lose everything you drop, you can walk through it without getting ticks on yourself (try walking through a pasture in the NE during the summer). You can throw a towel down and sunbath. Infants can learn how to walk on one and play with minimal risks. They are great for water fights with your kids in the heat of the summer.

Turf lawns are incredibly useful.

Back to the original point - the problem with shutters today on most houses is not they exist, but they are such a low-quality effort. Superfluous items can have a place in design, but at least make an effort and show an ounce of care. Here is a video showing a minor detail that can make a big difference:

https://www.youtube.com/watch?v=KI3bC0mRp2o&t=301s

Great example. Compare the traction of your machine to an 8N. That engine will outlast 4-5 briggs engines easy (10K vs 2K hrs). They have an efficient gear drive as opposed to hydrostatic. Lots less maintenance and no need for gallons of hydro fluid. They have a robust 3pt hitch with an excellent PTO drive. They will drag your tractor up and down the field all day long. I really mean to be constructive, sorry if I am not sugar coating it.

Go look at an old farmall or ford tractor and understand how they solved these problems 70+ years ago. Be careful thinking you are smarter than those engineers.

They are terribly inefficient and who is going to be servicing all these hydraulics in the 3rd world? Every farming town in the USA has someone who can make custom hydro lines quickly. How many of those shops in Africa?

The problem is that this solution isn't even close to being realistic. 28HP engines are not created equal and this engine isn't anything like the old 28HP engines from tractors of yesteryear. The tires are wrong, hydraulic drive is wrong, no PTO yet, etc. Bolting tube together on a moving machine like this, WTF. The tube works free over time as it collapses which will cause play and egging of the holes. This is completely wrong. They need to go back and study the tractors built in the late 30s and early 40s and copy those. This is nothing but a mess. Are any of these people mechanical engineers who have built Ag equipment?

Would it not be easier if Go just provided a stack trace attached to the Error? How to cleanly do this in Go I don't know, I do however do this in embedded C++ and it works well. I agree without context, errors can be hard to track down when they come from functions that are called by many other functions.

I am not a fan of manually wrapping errors because it seems inferior to stack trace.

Also, I hate that Errors in Go are mostly just random strings, super hard at a high level to do anything intelligent.

> Teachers, social workers, public defenders, armed service people.

I don't agree with that in all cases. I have a friend who is a teacher and admits teaching is a great job for him (2nd grade) and he doesn't know what he would do otherwise. Probably some type of physical work. He is a great guy, but just isn't that sharp. He is punctual and dependable, but not real good at thinking on his feet. He makes about $82k teaching (20+ years experience in a reasonably affordable area) and I can't think of another job that would pay him that well, give him the summer off, great healthcare and a pension.

I know a few more teachers and perhaps it is the group of people I hang out with, but they all are closer to him than to some person making a sacrifice for the good of society.

First, there is no such thing as digital. It is just an abstraction to facilitate our understanding and design of circuits. Any number of issues can cause that abstraction to fail, even deep within a modern chip, let alone at the board level. On-chip cross talk, RC drop, and ground bounce are simple analog issues that have to be contended with on digital designs. Every chip has a power supply and the smaller geometry chips have less tolerance to supply spikes. Same with I/O including voltage range tolerances and ESD protection. All that gets harder in smaller geometries. Also factor in circuit lifetimes (including digital). Due to a variety of mechanisms, small geometry chips just are not as reliable and some of these mechanisms are very hard to accelerate so qualification is a nightmare.

A smart phone is not a "way more noisy" environment than a car. It has different issues, and frankly a far more constrained and predictable environment. A load dump event in a car can take the 12V supply over 100V. I have never heard of that in cell phone. Long wiring harnesses are a nightmare and require careful design. It is hard to compare a cell phone and car without a clear definition of "noisy", but as someone who has designed systems similar to both, I think the car is a harder set of problems. Just not as constrained noise sources or components.