michael_j_ward

Who is "we"?

> Dealing with every possible source of non-determinism, re-writing services to be testable/sans-IO [2], etc. takes a lot of engineering effort.

Are there public examples of what such a re-write looks like?

Also, are you working at a rust shop that's developing this way?

Final Note, TigerBeetle is another product that was written this way.

From the abstract, it sounds more like a toolkit for "Interactive / Hypermedia" papers. The paper itself is still dead.

I was hoping more for "Living" as in "active, uncertain, will grow over time".

A toolkit for expressing the research from beginning to the end - the state of the world as you understand it, highlighting the key uncertainties and experiments, and mechanisms for viewing the history.

As a motivating example for the type of "living research paper" that I'm thinking of, think of long-running software design decisions a la the implementation of `async/await` in rust.

And in the context of discussing "whole word" vs "phonetic" systems of learning the English language, it seems like outright deception.

My kids love that book, and my oldest had me read it to his preschool class earlier this year.

Here is a much more creative reading by Ludacris [0]

[0] https://www.youtube.com/watch?v=PFtHeo7oMSU

Pointing out a few outliers in such a large system like the English Language and using them to justify reclassifying the entire thing is braindead.

It's an alphabetic writing system where the letters largely correspond to sounds, in contrast to logographic writing system where the symbol corresponds to the entire word (like Chinese or Japan).

Yes, there are special rules and outliers that you need to learn in English, but it seems absurd to not classify it as phonetic because it's not purely phonetic. This is doubly so when discussing phonetic vs whole-word learning systems, as is the topic with "Sold a Story".

I highly recommend famed computer scientist Richard Hamming's "You and Your Research" [0] which covers "creating luck" from a researchers perspective.

[0]: https://www.cs.virginia.edu/~robins/YouAndYourResearch.html

If you don't mind sharing, what grade-level, decade, location were you taught this?

The key part about the pizza example is the demonstration of mathematical thinking completely disparate from what I experienced at a Catholic high-school in Chicago during the 2000s.

TIL what an "SMT Pick and Place" [0][1] machine is.

Summary: $1000 and with open source hardware and tools, I can make custom computing hardware for your projects.

Seems pretty cool.

[0] https://www.youtube.com/watch?v=oFJrbRKUXdc [1] https://en.wikipedia.org/wiki/Pick-and-place_machine

Do I have this Napkin math correct?

400 tons produced per year (original article)

$1000 per ton (generous [0], current index at $770)

$400k per year revenue (calculated)

$25M Euro investment (original article)

~63 years payback on *revenue* (being generous again with EURSD=$1)

Do I have that right?

[0] https://www.cmegroup.com/markets/metals/ferrous/hrc-steel.ht...

Despite seeing myself almost precisely described in that video and the original article, and despite having been diagnosed with ADHD as a college student...

It's hard for me to believe that there exists persons that won't / don't identify with that description, and thus the cause of my executive dysfunction isn't ADHD, but just a failure to cultivate habits of discipline.

Having suffered through learning highly configurable software tools independently enough times, I agree completely with "digesting one or few commands" at a time works better for me too.

Spitballing:

We could probably make a general solution for organically producing "learning guide" if power users 1) share their custom configurations and share the frequencies of commands they use.

Start with the general configuration, knowing nothing about the commands the tool offers. The learning guide is just a queue of "here's the most common config edit that you don't have" and "here's the most common command that you don't know about."

We are already share configs with eachother via the dotfiles pattern. We somewhat share command patterns via coding streams. Seems doable.

(In general, making it easier to learn new tech and for devs to share / refresh the learning material for a new tech seem underdeveloped.)

Is Ilograph open source?

heh - you're obviously correct and now I'm kicking myself for forgetting the #1 difference.

FTT does not offer any sort of interest like Terra did. The benefits of holding are strictly discounted use of the FTX platform [0]

[0] https://help.ftx.com/hc/en-us/articles/360052410392-FTT-Stak...

What the argument misses / why FTT is different than Terra

1) Alameda Research owns FTX, one of the largest and arguably most important crypto exchanges.

2) FTX offers fee discounts to FTT-stakers and additional discounts if you pay in FTT. [0]

3) Trading volume on FTX thus creates an organic demand cycle for FTT. The large firms will buy, stake, and then continuously refresh their supply.

4) The vast majority of the volume at FTT will be in margined accounts at FTX. I am uncertain if the volume analysis would capture FTT movements in (3).

Now, there's clearly financial alchemy going (giving away real economic value to boost an asset that you can then get leverage on) but that'd be better for Matt Levine or someone to flesh out.

[0] https://help.ftx.com/hc/en-us/articles/360024479432-Fees

Just want to say congratulations to the team!

2 years and 9,500+ commits is a hell of a feat.

https://github.com/influxdata/influxdb_iox

> (1) I can have multiple domains for the same device, say gitea.mytsnetwork.com and netxcloud.mytsnetwork.com can go to the same device

I tried setting up caddy on a machine and then using caddy to reverse-proxy requests to each service i.e. `grafana.my-machine.tail-hex.ts.net` and `controller.my-machine.tail-hex.ts`

Obviously, `caddy` has no problem with the reverse proxy bit, but I did fail at being able to point multiple routes or subnet routes at the same machine via tailscle / magic-dns.

I'm sharing because it feels like something I should be able to do, and feel dumb not being able to figure it out.

I think you underestimate the selling power of "Green" options, particularly among those with dollars to spend.

I bet lots of schools would pay up for "sustainable" laptops, actually.

I'm very interested in the contours of this relationship with Google.

- What kind of commitments did each party make to each other?

- Did Google request anything of Framework? What requests did Framework agree to? Which did they deny?

- What differentiates this product from the normal offering?

Yes- there is a life-year cost to bringing death forward.

But there has now been no excess mortality in Sweden compared to it's neighbors Finland and Norway (and all 3 blow the US and Europe out of the water.)

https://ourworldindata.org/grapher/excess-deaths-cumulative-...

Sweden was the well publicized "control group."

It was not selected post-hoc, it was selected in 2020.

The result was more death early and less death later, while suffering no learning loss or adverse effects on the children and little loss in quality of life.

Here's an article from 2020 disparaging the high costs [0].

If you find any major news giving an update now that Finland and Norway have caught up in excess death, please share.

[0] https://www.science.org/content/article/it-s-been-so-so-surr...

if litestream ever enables logical replication, I think you could do `SQLite logical replication --> embedded materialize-db --> back to SQLite`

For others like me that wondered "Wait is proof by induction not induction?"

> Although its name may suggest otherwise, mathematical induction should not be confused with inductive reasoning as used in philosophy (see Problem of induction). The mathematical method examines infinitely many cases to prove a general statement, but does so by a finite chain of deductive reasoning involving the variable n, which can take infinitely many values.

https://en.wikipedia.org/wiki/Mathematical_induction

If one does not need petabytes of scale, but am otherwise interested in the data lineage / observability / workflow being sold by databricks, what would you suggest?

Some evaluation Criteria:

- Ease of maintenance and operation is almost paramount.

- It's fine if the solution never lives anywhere but 1 single virtual server that scales vertically (data might grow to a couple TB, but not PETA BYTES)

- Similarly, 20 9's is not a criteria. If the machine fails and it takes an hour till someone goes and re-deploys, that's fine.

- Declarative, reproducible deployment with an easy upgrade story would be great

- Ideal if the deployment can be run locally for quick developmnet

This may interest you

https://www.edgedb.com/docs/edgeql/

Sounds similar to xtdb

https://github.com/xtdb/xtdb

Given ~67k deaths under 50 years old total from COVID [0] and over 100k overdose deaths in 2021 alone [1], I'll wager heavily that is not true for the under 50 crowd.

[0] https://www.cdc.gov/nchs/nvss/vsrr/covid_weekly/index.htm#Se...

[1] https://www.cdc.gov/nchs/pressroom/nchs_press_releases/2021/...

> 1 in 350 Americans died of COVID in the past two years

Putting 1/700 a year death from covid into context of our overall mortality rate: 1,027.0 deaths per 100,000 population [0]

I know more under 40 that have died from overdose in the past two years than I know that have died from covid (which is 0)

[0] https://www.cdc.gov/nchs/fastats/deaths.htm