pstoll

> People don’t expect any return on investment from physical theories - noone bashed string theory for failing to make useful predictions.

Uhhh… heck yeah they did. It’s the primary and most widely challenged aspect of the theory. This does not help your paper be more credible.

Few observation / opinions from a parent who sent kids to a Russian math school in an area with “good public schools” and kids are now older (HS & college). YMMV.

The teachers by and large care about and enjoy math. That is good. It’s rare to see this in public schools in the US before high school.

Unless your kid loves math, don’t make them do it.

Kids under 6th grade don’t need to grind and do lots of homework. Aka chill.

Best sardonic joke - can’t wait until someone starts a Russian School of English.

Ah - I see. Visuals are in the preprint PDF, not the website.

I coudn't find any visual examples on the site - did i miss them?

Was hoping to see how the various code examples actually render.

How many people die each year from non-GFCI outlets? Data to back it up?

I know I should run my own ZFS cluster at home. It would make me a better person, it’s the righteous right thing to do.

Maybe in my younger days I would have been purer of spirit. Alas, time, kids, other hobbies & commitments have stood in the way.

I bought a NAS (synology). Life is full of compromises; this is one of mine.

We have gone through a similar journey on a project at work - starting by using Z3 then moving to a simpler in-house built boolean constraint solver. Will see if it is something that could be open sourced.

Sure - add the Apple A chips; Qualcomm chips for communications, etc…

We talk power, not servers, at work because it’s the limiting resource at any scale. “How many MWatts do you need next year”. Its not that you can’t do these workloads on CPUs or GPUs - you can. But you can do it at 10x-30x less power on custom silicon.

That’s why Amazon did Graviton3; Google did TPUs and VPUs. Ditto on personal computer with Apple & M1/M2.

“Pure” does definitely connote that you’ll be able to read all the code in the given language.

That’s exactly what “pure” means - “pure rust”, “pure go”, etc. IMO you can’t say the heart of all the work is a c++ lib and call it a “pure JS” anything.

More accurately / correctly / usefully would be calling it “JS wrapper over a c++ library cross compiled to JS”. Or maybe “All JS at runtime” or some other qualifier.

Want to see how to cross-compile a non-trivial c++ lib? Check out here!

Want to see a great JS wrapper library where we had to make cross-language & cross memory-management API decisions? Check this out!

But - want to read some awesome high performance image processing algorithms in JS? Not this.

And also some good articles by the CockroachDB folks on vectorization.

https://www.cockroachlabs.com/blog/vectorized-hash-joiner/

https://www.cockroachlabs.com/blog/vectorizing-the-merge-joi...

This seems similar, but perhaps less ambitious, than the very well done Nand-To-Tetris (https://www.nand2tetris.org/) - anyone know both?

Nand2Tetris has open source simulators, great docs, lectures / help via coursera. I’ve used it with kids.

Admitted bias as EE who used circuits & assembly back in the day in addition to CS, but I do think knowing the underlying “machine” of compute helps inform good systems thinking / trade offs. Plus it’s fun for people who are engaged by this type of logic puzzles / problem solving, which I think is a large set of folks who like software.

And for completeness, here’s another company that I also used at $previous_job that provides a “semantic model” offering. This write up also helps describe where it fits in.

(This one has just enough content vs marketing for me not to feel embarrassed posting here on HN for people who want to find out more. And IMO the BI landscape is littered with pablum from an engineers POV, often obscuring the nature of the technical problems to be solved in the space - which are very cool.)

https://www.atscale.com/blog/what-is-a-universal-semantic-la...

Quick scan of dbt metrics - looks like it lets you define a flat, single view of a metric with a predefined set of “filters” (Ie sql predicates). Also it only allows a metric to come from a single table; “joins” are a TODO.

It is missing the runtime engine to “interpret” the semantic model and so arbitrary but correct joins among tables and generate the appropriate SQl to realize those arbitrary joins.

If anyone know dbt better and also knows semantic models, please correct.

Well said!

We need more of this problem space exposed to engineers and not just for “analysts”.

I’ll share a couple other articles from a company that does a nice job explaining the technical problems in what is traditionally “business analytics”.

The space is OLAP and you may have scoffed at the idea of “OLAP cubes”, but man were they useful. In the way that excel powers a ton of business processes, cubes powered a lot of analytics. Underlying tech is cool but they are showing their age: https://www.holistics.io/blog/the-rise-and-fall-of-the-olap-...

Another write up of this idea of a semantic layer above raw sql statement: https://www.holistics.io/blog/holistics-data-modeling-explai...

So this “semantic layer” leverages the latest tech to deliver the same business insights faster, better, more flexibly. Ie once you define this semantic layer over your data (ie how all your sql tables are connected), the semantic engine knows how to query up and down your data model, writing the SQL queries for you, on the fly. You can ask and answer new questions without writing new queries. And with modern columnar query engines (eg big query, spark, presto, etc), perf is usually pretty good.

Cool. Like open source Looker.

We adopted Looker at $previous_job. Then they got bought by Google, which was great for us as we were becoming a big GCP customer. I strongly encouraged google / looker team to at least open source their LookMl (looker modeling language - equivalent to MQL). They couldn’t figure it out.

This type of metric definition is so empowering for businesses. Not enough engineers grok why this is useful.

This is a really accessible write up.

Although I agree with the comment elsewhere of asking for a “examples in Rust” comment somewhere to save everyone the 15 sec of squinting “is this … C++, or…” who don’t instantly recognize it.

Also funny that I just shared a Twitter thread on this topic to a coworker.

I was sharing the POV of this prof with the “wow, it’s amazing to think all the hard-won gains I ever did with deep feature engineering and tuning is now useless!”

Sharing in case others find it helpful.

https://twitter.com/moyix/status/1469401502422818823?s=10&t=...

Which references this thread:

https://twitter.com/karpathy/status/1468370605229547522?s=10...

Interesting - so more like “verity but verify”. Very cool.

tldr: “scientific” study with N=1 (author / researcher) draws useless conclusions, but because it contains “cats pics” gains some attention.

Basic - naming a capture group. Awk / posix - \1 Perl - $1. Python - (?P<name>).

I’d find it hard to ignore the specifics. I mean if you only ever use one tool and were never exposed to other regex engines I guess.

You must not have suffered enough, I mean used regexs widely enough.

He meant “syntax” in the sense that different regex engines have different syntax and capabilities - can I do a negative look ahead assertion in engines Z, how do I do a zero width lookaround in pcre, gnu, python, posix, etc.

Depending how far down the rabbit hole you want to go, start here:

https://swtch.com/~rsc/regexp/regexp1.html

The article lists alternatives but doesn’t mention Cython. That’s a miss. I’ve done both. If numba works, it’s great. I found it a bear to install. Cython is more work, did what I needed - allowed me to incrementally type Python to get C code performance for general numeric code where memory allocations mattered (large arrays).

Making false statements on an visa application form is grounds for pretty much any obvious consequences that follow.

It’s infantile but I’m tickled by the nickname I heard for him:

No-Vax Just-a-Dick

He said he had COVID a second time not long before going to Australia (Dec). Which is truly fortuitous timing - if he didn’t have COVID around then, he couldn’t get into Australia via an exemption for the recently infected.

And then video surfaced of him giving an interview while he said he had COVID. He choose to apologize for giving an interview while having COVID.

And then he managed to fail to disclose on his entry application that he had traveled to various other countries before coming to Australia.

This is someone who was looking to flaunt the rules and it didn’t work out for him.

Sure, perhaps I was inarticulate - he’s a world leading medical researcher who heads up a world leading medical research institution; he is one of the 10 most cited medical researchers in the world. Yadda yadda…

Interesting point I heard about why we don’t (yet) have a vaccine for HIV where we do for COVID - the human body can fight COVID. Ie we mount an immune response and some people survive and recover on their own. This is largely not true with HIV - the human body hasalmost no ability to mount its own response.

Making a vaccine is more straightforward (ie we expect to be able to) when we know there is a response by our own immune systems - hence we expected to be able to make a COVID vaccine (although the time scale is still an amazing accomplishment - thanks 20+ years of mRNA research!). And we are still working on a vaccine for HIV.

No one gets a free pass, for sure, but he’s a pretty qualified thinker of things medical and health. So coming here and calling him out on it should probably warrant a bit more than a “uhn uhn”.

(Edited - asimpletune above said it more eloquently than I did!)

> as vaccination rates increase the ratio of vaccinated:unvaccinated hospitalized patients is approaching and surpassing parity.

Parity as in 1:1? Because I’m pretty sure raw counts and case rates are still nowhere near parity. You mentioned data by age ranges as well.

This commentary on a new study is pretty good: https://twitter.com/zeynep/status/1482501951065661440?s=21

Pointing to this recent study: https://www.alberta.ca/stats/covid-19-alberta-statistics.htm...

I think activating “crisis standards of care” is a bit more substantial and worrying than any previous claims of being simply “overwhelmed” which could be subjective. This is what we are seeing here in the US.

https://www.google.com/search?q=intitle%3Ahospitals+intitle%...