theresistor

> Isn't that what everyone is doing, or are we Frenchmen the exception? > For reference if the author reads this, we write the latin x exactly like the cyrillic х, i.e. reverse c, bottom-left to top-right diagonal, normal c.

I was taught script in the US and Italy as a child, and never learned it like this.

I had this argument with people working on VR headsets, where a physical parameter was designed to cover the 5th to 95th percentile. I had to point out that flat-out excluding 10% of the population is a pretty crappy starting point...

In most amateur imaging you can trade time for quality. By stacking enough images, satellite trails get averaged out of the final image.

Very high end amateurs get pissy about it because they paid a lot of money for high end equipment to minimize imaging times, but for the rest of us it’s not a huge impediment as we already needed lots of subframes to get high SNR anyways.

You’d be amazed what even a small smart telescope can see even in Bortle 9 urban skies.

I have one of the mentioned SeeStar smart telescopes, and have been quite successful in taking photos of both galaxies and nebulas from the rooftop of my apartment building in the middle of a city.

Sure, it takes time. Sure, it would be faster/better under darker skies. But it’s not impossible!

iTelescope does exactly what you’re outlining. There used to be a few others but I think they have mostly died out. I assume the economics don’t really work out.

Also also, AMD’s play has always been to produce HW that offers good performance/$, with the downside of having much weaker SW offerings to go with it.

Consoles are always pressured to minimize upfront purchase costs, and they generally replace the vendor-provider SW stack with their own anyways.

> It might be that regalloc needs to be taught to rematerialize

It knows how to rematerialize, and has for a long time, but the backend is generally more local/has less visibility than the optimizer. This causes it to struggle to consistently undo bad decisions LICM may have made.

I say and hear it all the time in the US...

Really? I read the same sentence (as an American) and immediately thought that they must be referring to British English. Certainly nobody says brilliant as an affirmation here.

And "no problem" and "not bad" are both common colloquial statements in American English.

As far as I know Apple will still sell you individual tracks (DRM free, I think?), though it’s a bit hidden.

A very recent example: https://github.com/ocaml/ocaml/pull/14369

This is a pretty standard topic, and not really a compiler optimization. It's usually called a unity build.

[0] https://en.wikipedia.org/wiki/Unity_build

My grandfather also worked on it, as a technician in Los Alamos.

He had previously been working for a scientific supplies company in Chicago that was (unbeknownst to him) providing supplies to the Manhattan Project. Apparently his boss was aware of it, and when my grandfather's draft was called a letter from his boss convinced the draft board to assign him to Los Alamos instead. He was eventually able to get my grandmother, a secretary and typist, a job as a secretary in Los Alamos as well so that she could join him. She teased him the rest of their lives, because as the secretary to someone more important than a lowly technician, she had technically had a higher security clearance than he ever did!

The Atomic Heritage Foundation collects records about people who were affiliated with the Manhattan Project, as well as oral histories. Perhaps they have more information about your grandfather's work? See here: https://ahf.nuclearmuseum.org/ahf/bios/

Kid me absolutely ran The Bug on an old PowerMac G3: https://macthemes.garden/themes/8191e1471dc9-the-bug/

> I tend to agree but, playing devil's advocate, is this true for other roles? Does a movie director need to know how to build sets? How to sew costumes? How to use Blender/Maya/Houdini?

I don't know that much about movie making, but my understanding is that there would be managers and/or leads within each specialty, who are (among other things) managing the interaction between their specialty and the director / producers.

That seems pretty comparable to what's being discussed here.

My father was a Classics professor, and one of the most popular classes he taught was "Latin Terminology for Medicine and Law".

All of the major C/C++ compilers are deterministic functions of their inputs. This is a pretty major property, and any regression is rapidly detected and fixed.

Optimization "timeouts" aren't done in wall time, but in something that can be deterministically counted: number of instructions, number of recurrences up a call stack or expression tree, etc.

> Do they just target the lowest common denominator of operations? Or do they somehow adapt to the operations supported by the user's CPU?

Mostly the former. Some highly optimized bits of software do the latter—they are built with multiple code paths optimized for different hardware capabilities, and select which one to use at runtime.

> Do dynamic languages (Javascript, Python, PHP...) get a speed boost because they can compile just in time and use all the features of the user's CPU?

Hypothetically yes, but in practice no for the languages you mentioned because they don't map well to things like SIMD. Some JIT-based numerical computing systems as well as JIT-based ML compilers do reap those benefits.

The textual IR is not backwards compatible, but the bitcode format has been best-effort auto-upgradeable for as long as I've been involved (2006). The policy is documented here: https://llvm.org/docs/DeveloperPolicy.html#ir-backwards-comp...

If you know that you need to offload matmuls, then building matmul hardware is more area efficient than adding an entire extra CPU. Various intermediate points exist along that spectrum, e.g. Cell's SPUs.

> Also, people often mistake the reason for an NPU is "speed". That's not correct. The whole point of the NPU is rather to focus on low power consumption.

It's also often about offload. Depending on the use case, the CPU and GPU may be busy with other tasks, so the NPU is free bandwidth that can be used without stealing from the others. Consider AI-powered photo filters: the GPU is probably busy rendering the preview, and the CPU is busy drawing UI and handling user inputs.

You can, but what does it help? Modern OSes are architected assuming an always-online, the-world-ends threat model. Thus causes them to be heavily locked down, eliminating a lot of the customizability and hackability that older systems had.

And that’s not to mention applications. It used to be common for GUI applications to be scriptable and to support plugins!

I miss having a world where the threat model of having your computer compromised wasn't "your entire life ends". Where a device wasn't assumed to be online all the time, and where every device didn't contain the keys to your bank account.

My favorite memory of Halted is finding an Atari 2600 in a box with an Atari company store sticker on the box, and someone’s DIY controllers with it.