Erm, don't the guys making homemade Nixie tubes have this figured out?
18,936 karma · joined February 8, 2014
For a professional tool, animations are anathema. They interfere with muscle memory.
Someone who uses a program continuously can be clicking or typing before a dialog box or button is even in the right position.
My wife drives me MAD with this. She has already clicked the cancel button on a popup before I can even read the first word in the dialog box. This is fine when she is working as the dialog box is just a dumbass notification from some idiot UI designer. This is NOT fine when she has asked me to help debug a problem. I have to force her hand off the mouse so that I can read the damn error message before she clicks it away.
The reason you don't use prefix adders in 1980 is that you can't possibly route them because you don't have enough metal. So instead, you use chunks of Manchester carry chain because the "tapping internal nodes" that everybody cites allows you to route nodes in diffusion and polysilicon instead of having to use metal.
Of course, THAT only works because you have 5V (or more) and can connect lots of transistors in series and still have them work. As your voltage falls you can't connect as many transistors in series, so you switch to architectures that prefer active gates over passthroughs and long chains.
So, as your available metal layers, supply voltage, transistor speed, threshold voltages, capacitive load and power dissipation all shift over the engineering landscape, your "optimization" shifts with it.
Anthropic doesn't care. The goal right now is simply to avoid any and all bad PR on the way to the cashout IPO.
And paternalism will generate far less bad PR than somebody using AI on something that does real damage and makes headline news.
And this is another problem--your "justice system" should NOT be where you place your mentally-ill--they belong in a (possibly secure) medical facility and not with rank-and-file criminals. This is yet one more issue with the US system.
Wikipedia article makes no obvious mention of correlation between sentencing and deterrence. Linked article, in fact, demonstrates that alternative programs provided almost all the improvement.
I do NOT concede that as the KEY purpose. And when you call "protecting society from criminals" the key purpose of the justice system you wind up with the horribly broken mess that is the US justice system.
Yes, A purpose of the justice system is to remove from society those who cannot be trusted.
But another purpose of the justice system should be to rehabilitate those who can be. And the US justice system is HORRIFIC at that. If anything, the US justice system is a net negative on rehabilitation. The way the US system throws everybody together does more to let old criminals teach new ones their tricks than any improvement from any rehabilitation program can counteract.
Most other forms of crime, especially violent, are almost completely insensitive to the harshness of any possible punishment sentence.
If a brain is sufficiently broken that it no longer has the limiter against harming another human being, prison sentences won't do anything to fix that.
https://www.dailyartmagazine.com/henri-riviere-master-printm...
The Shadow Theatre at "Le Chat Noir" was fairly famous, no?
Everybody loves capitalism--until they are on the other side of the arrow.
I use buttermilk powder for quite a few recipes so that I can control the liquid content independently of the acid and fat content.
Sure it did. And Silicon Valley is just the current nexus.
In the mid 1800s, Pittsburgh was the startup technological nexus for chemistry and engineering around steel. In the early 1900s, Detroit was the startup technological nexus for car and machine manufacturing. In the 1960s, Silicon Valley became the startup technological nexus due to Fairchild and the Traitorous Eight.
The "trick" to creating a startup technological nexus is for the price of a startup to be in the range that experienced, generally young (roughly 25-35), mid-level technical professionals can fund.
So, the optimists all swim in the cash while your contrary position fails to keep pace with the bull market; and then the bear market hits and you all get obliterated equally.
My memory seems to think that wasn't true.
However, there were a ton of manufacturing and manual labor jobs that were capable of supporting a small family and they didn't need a college degree.
So, you had most young people earning positive money for four years at a very biologically fertile age rather than going into soul crushing debt at that age.
Then I got Mary pregnant
And, man, that was all she wrote
And for my nineteenth birthday I got a union card and a wedding coat
The timeliness of the repair is more important than anything at harvest time. Farmers have to be able to repair their own equipment or risk suffering an egregiously expensive loss.
Those are not the only choices.
Garbage collection is at one end of the spectrum--fully manually managed is at the other end of the spectrum. There is also another axis of acquiring/releasing allocation by object or acquiring/releasing object by pool. And, if you have it, there is the axis of allocate only at startup and never free until end of thread/application vs. allocate only at a frame of of time and then destroy them all at the next frame vs. allocate whenever and wherever.
RAII encourages the usage of lots of tiny individual objects allocated whenever and wherever all with their own lifetime cycles and makes understanding the memory usage of your application very difficult (this was the whole reason Rust was made--C and C++ made managing memory in Firefox ridiculously diffused and impossible to corral).
And, I'll be blunt, I think that Rust/Zig/C3 etc. are not the right direction in spite of the fact that I use Zig a lot. I think that the garbage collected languages cede far too much in terms of performance to the compiled ones and GC languages (like say OCaml) should be being used for systems programming more often.
For example, I think we would all be in much better shape against AI vulnerability scanners if more systems programs were in GC-type languages.
Although I think that there is FINALLY an actual spec for pointer warp. However, very few compositors support it.
Certainly it matters much less in the modern era.
However, certain fundamental decisions of a language can be dealbreakers.
Requiring declarations on your functions and giving those declarations sigils so that they can be parsed quickly is an important syntax decision. Almost every modern programming language has converged to this idea.
Or take, for example, Lua. For me, personally, the 1-based-ness of Lua is simply a dealbreaker no matter how good anything else about it is.
For the "Lisps", I LOATHE the fact that you traverse lists and vectors in completely different ways--you can't just drop any container-ish thing into something that iterates/collects it. This is something that both Clojure and Racket seem to agree on--you have something that acts like a "collection" and you can walk across it the same way regardless of the specific type of collection it is. Of course, that is why a bunch of Lisp purists loathe Clojure and Racket while I like those languages. Shrug.
I find RAII (Resource acquisition is initialization) to be the source of all things evil if it infests a programming language. The popularity of C++ and Rust speaks to the quite large number of people who think my opinion is bullshit.
So, yeah, base syntax matters far less than it used to. But the engineering decisions that went into making that syntax correspondingly are far more important.
I suspect it was a later incarnation that I used.
> It was a fun place to be.
It was. I was in the hardware side during the Hostile Giveaway(tm) to Intel and Compaq.
Good luck with your cat.
Following your comments and trying to trail things out, the only thing I can find is a reference to AltaVista LiveTopics.
https://dannysullivan.com/why-search-sucks-you-wont-fix-it-t...
I suspect what I used was that "GRAPH" button in the AltaVista Refine picture.
We might want to archive that image. It seems like the only evidence that this stuff ever existed.
And I still miss the AltaVista illustrated diagram (Java Applet) that would allow you to drill down and specialize the search results. No modern search has ever matched that, again.
@cImport is moving to the build system in an upcoming version (0.17, I think?). It's inconvenient, but it's unfortunately necessary.
The issue with @cImport is that when and the order the imports occur are almost impossible to define if you have multiple files and libraries. By moving it to the build system, the order now becomes deterministic and you can refer to it properly.
Your memory is faulty. AltaVista was always super fast--it never had the advertising bloat that the other ones had until the very end.
The problem AltaVista had was that it didn't scale when the Internet went exponential--so AltaVista would give you good search results until you asked current, topical questions. AltaVista relied on running a single, super-expensive stonking huge Alpha machine while Google ran on lots of commodity servers that spidered constantly.
Take one look at the mess of incompatible detachable cords that come with LED lighting fixtures from China to see what results.
They could have used bog standard IEC 320-C5 cables, but they didn't. Why? Because they wanted to cheap out on the conductor AWG and IEC 320-C5 is quite specific about the conductor gauges and current capacities.
So, the LED lighting manufacturers produced a bunch of cables that are in almost exactly the same shape and size factor as IEC 320-C5 while being just enough incompatible to not invoke the certifying authorities--all to save a couple pennies in copper.
Agreed, LLMs are particularly good at this kind of task.
For example: my windowing system on Linux would intermittently freeze. Diagnosing it was a pain--so I bounced the logs off the LLM. It gave me a couple of hypotheses and the commands to enable the correct logs for when it happened again. After the third time it happened, it pinpointed that a particular USB hub was causing the issue. I removed the devices downstream from that hub and haven't had an issue since.
We saw this happen in Houston. Many of the worst public schools suddenly "improved". It's a miracle! Oh, they did this by encouraging the lowest performers to drop out. Whoops.
And this is before we start talking about all the high GPA students who now all magically need IEPs (Individualized Education Plans) because it gets them an extra 50% of time on their tests. So, now you have your best students loudly (because these parents are active) soaking up lots of resources genuinely meant for your worst students.
That completely ignores the social and political aspects.
You need to understand that the people who have the authority to do so do not want to document bad teachers, ever. Documenting bad teachers makes political waves and principals and superintendents never want to make waves because that impedes their ability to both do their job as well as get their next job.
Even if a teacher is very bad, they may be well-liked or be an important part of the community. If you attempt to remove that teacher, they may rally support from the community that can be extremely noisy and inconvenient.
It's not relevant. Don't try to convert git users--you don't need to. They're fine doing what they do, and the git backend store is acceptable. People who understand how broken git is will gravitate to jj with the lightest of prodding--give them a light poke and then don't harass them further.
The main difference is that I can teach jj (like I could teach hg) to normal people. And, because jj is based on the git store, they can operate in a Github world.
So, just teach jj to people who haven't yet broken their brain with git. Running on the git backend already gains most of the network effect, so don't proselytize the git people.