eurg

> What were the most challenging technical problems?

While technically the customer asked for the solution to problem A, they wanted us to solve problem B, but we failed to invoke our inner patio11, and as a result delivered what they said instead of what they meant. A frustrating lesson for me. At that point in my career I should have known better already.

ETA: sibling comment about Cortex-M0 is spot on. You really start counting bytes. Some kiB of RAM sound a lot, but when you have to juggle several tasks, and don't have to option to write from scratch but instead reuse existing old code you've lying around, stuff gets dirty pretty quickly. Other team member was doing the hard work, I was just providing tooling for them.

Where did Robert Haas say this? Quick search didn't surface much, except for a blog post from July by Hans-Jürgen Schönig: https://www.cybertec-postgresql.com/en/postgresql-zheap-curr... Doesn't sound like they discontinued the project.

Once did a prototype that had to work with a really memory poor device. Compression and speed is meh, but still worth it (small serial interfaces are no racing dogs either).

Also very simple to understand and port and play around with, and to compare with other algorithms. More or less took this project as an excuse to learn some compression basics. Can recommend

The hobby repair shop can only be liable for a very small number of bikes, those they worked on. You cannot restrict the number of users of free source software, and you cannot restrict the user's risk profile. Like "good enough for an offline arcade game, but nothing else". Analogies have their limits.

But yes, lawmakers will decide, and given that they for instance try to de facto prohibit aftermarket OpenWRT installs, I have a guess how they would decide.

Mansplaining? It happens to men too, not just by.

A bit generic, though.

Some additions:

I want to second the clarifications by adrianN and shawnz -- there is a difference between not actively caring about suffering, and not even seeing the point in anyone caring about it.

It seems my "I in fact believe you should see a psychologist" worked exactly as well rootusrootus implied it would. Many seem to think I want to label the previous poster as a psychopath, and do so in hatred. That was not my intention. I admit, I fail at communication, regularly, even more so the more I try to refrain from it.

Although suggested otherwise, my message did *not* contain a diagnosis. Given this extremely limited context, I doubt anyone could give a proper remote diagnosis. And I especially did not attach a psychopathy label to the previous poster. I personally know of at least three mental/neurological atypical (and not voluntarily chosen) states that cause people to (partially) adopt nihilism, and I'm sure mental health professionals have seen quite a few more.

I know that rule of thumb, but it's only a rule of thumb. I haven't yet seen a way to suggest seeing a doctor via the written part of the web that won't immediately get interpreted as name calling by many people.

If the reduction of suffering is not something that you internally see as valuable, you might want to check with a psychologist (I mean that plain matter-of-fact-ly). While I can construct philosophical arguments and know what nihilism is, outside of thought experiments this is highly unusual for any individual capable of any empathy at all.

Caching is a bit dangerous, and most shells don't support asynchronous prompt updates. And cross-shell support seems to be a prime feature. Spending probably a lot of effort just for improving zsh experience might just not have been important enough yet.

The power of powerlevel10k is that it can draw the plain prompt immediately, and respond to keyboard immediately, and then backfill any advanced prompt information (and properly shifting the already added text to the right and wrapping it correctly). I know that zsh has support for this, and bash does not (powershell neither); my prime reason why I use zsh. Fish, I don't know how it handles prompts.

This is very powerful in large git repos etc. You're never waiting for the prompt. You just type immediately, w/immediate response. So, you only wait for the status information to show up when you need it. So you also don't pay an excessive penalty if you prefer overly chatty prompts.

AFAICT the lazy-loading and async-support tickets above talk about internal implementation details, timeout support (so cancelling stuff like a too slow status report and showing a simplified prompt). I may misunderstand that, I only gave it a quick glance.

Maybe starship already supports this for zsh, or not, but at least they don't promote it front and center.

Last time I checked it was slower than https://github.com/romkatv/powerlevel10k, because it doesn't decouple the rendering of the prompt from the background processes collecting status information. Ergo no instant prompt. Also, some other cool tricks w.r.t. prompt vs. backscroll.

Has that changed?

Thanks for the explanation. I don't really understand why this service owner wants me rather using my censored ISP provided DNS, or Google's, but I'm sure they have a good reason.

Now this is unexpected, and amazing.

This is one of the most frictionless donation buttons ever. I love it.

Patreon, Paypal, SEPA transfer, all those are a hassle, comparatively.

This donation thing used by Signal works exactly as it should be. Enter numbers, hit enter, done. No "please cookie us", no 20 times transfer to other domains, no account creation, and they also don't require stuff like MasterCard 3D secure (which IMNSHO really is useless for donations).

Zero hassle, 100% great, and with a nice UX.

It's easy to quote a low price if you have zero stock to sell. The alternative was not masks available at low prices, the alternative was no masks.

I've to admit I'm surprised that the battery management is not on a separate IC. While quite a few battery management decisions are better done by software, hard-limits on current/temp/discharge really are not.

Am I misinterpreting this?

Malleable backups are not secure when they need to be restored, something that could be necessary if by some chance a server of yours became dead suddenly.

Lenovos etc. don't care, but I've seen this first hand with some Acer and Asus laptops. It has nothing to do w/peak power consumption, that is a wholly different problem, of a different class of devices.

(beware, I'm a dog)

Misusing those primitives can kill all guarantees, of course, the amount of kill dependent on the amount of misuse. Using non random keys will kill even more.

It's about ease of correct implementation, use, and what kind of degradation happens if a side-channel leaks a specific item:

It seems that w/AES (within the AES-GCM construction) can leak a key that is useful against all other uses of the original AES key that was used to derive the authentication key.

ChaPoly uses a construction where all crypto operations for a message exchange use a key which if leaked, cannot be reused to attack so many other message exchanges as well, because it incorporates both main key and nonce in the derivation.

HT to Filippo (https://twitter.com/FiloSottile/status/1260598284395126786). Another submission 10 hours ago didn't catch (https://news.ycombinator.com/item?id=23165258).

Too bad, because the blog post illustrates some limitations with the construction that are interesting to know, like the problem with using the same authentication key for every key instead of key/nonce pair, or the fact that AES-GCM is not "message committing" - a TIL for me.

It's also very pragmatic in the end: "Don’t throw AES-GCM out just because of my opinions. It’s very likely the best option you have."

I doubt that any part of law enforcement or worse parties will agree with you on the zero weight part.

it redirects via a domain on common blocklists

Counterpoint: After a promising two weeks, it did nothing for me, back to square one. OTOH a colleague of mine recommended this after having personal success.

It is frustrating that there seems to be only trial and error in all of this.

As an aside, sometimes local data-providers prohibit use of transit data by Google, with the result that you have to use an inferior route planning tool.

See some take on it at https://www.lesswrong.com/posts/gptXmhJxFiEwuPN98/meetup-not...

That has not been my experience. In a C# shop, if you want to do something not-C#, best is to switch employers. I don't say that placing such high value just into what tool you use is smart; but changing tools in a company is often a total no-go.

Newspapers and literature are quite distinct. Good authors do not require the smokescreen that typical journalists use. Whatever the language.

Below someone quoted this: "it will be assumed that all foie gras came from duck or geese that have been force-fed unless “documentary” evidence is provided to the contrary."

So, non-force-feed remains legal.

As linked somewhere else, but for video affine people, and directly relevant, see this old Ted talk: https://www.ted.com/talks/dan_barber_s_surprising_foie_gras_...

There is: https://www.ted.com/talks/dan_barber_s_surprising_foie_gras_...

He talks about a farmer who is not force feeding the animals, but instead prepares the grassland such that those animals voluntarily eat themselves into a fat liver. Taste seems to be superb, for those who eat animals for fun.

Neither one interprets the sequence, but cat will print the file unfiltered to stdout, and stdout is processed by your terminal, and then the terminal will interpret it. less filters before printing to the terminal.