teej

There are lots of human-in-the-loop systems with algorithms involved. People deeply care about those systems.

To be more precise, they “don’t care” that the system produces false positives. Did you consider they made this decision with full knowledge of the trade offs?

If you’re going to swap Facebook for Meta you must also swap Google for Alphabet

What resources would you point to for those who want to learn modern product/event based analytics?

Our bodies act as generation ships for our micro biome to explore the stars.

When it comes to luxury cars and clothes, you force others to see that you’ve spent for status. That doesn’t exist for NFTs. There’s no public forum I can go to where others are forced to observe my status signals.

High end art is more about investing, money laundering, and tax havens then status.

> In fact, someone can "sell" an NFT to themself in a wash trade between two wallets they control. The last trade price of the NFT is now recorded as the transaction price, but it's possible that money never actually changed hands.

I heard rumors this is what happened with the Beeple nft

One of the mistakes Mixpanel made was to position themselves as a “better” Google Analytics. That meant a generous free tier without the benefit to search that Google gets.

Amplitude, from the moment I was aware of it, was more about productizing the Facebook/Zynga style product analytics approach.

I left Zynga for an early startup in 2011. At that time, I tried to use Mixpanel for acquisition and retention analysis - it fell woefully short. I wasn’t able to use any of the built-in reporting.

Meanwhile, I have been a mega fan of Amplitude from the first time I ever used it. It was built for the “product data” use case first, not as a Google analytics replacement. That positioning made it easier for them to demand premium pricing.

2025 is not that far off though. It’s a hell of a lot closer than 2050. Easy to ignore political promises that come into effect after the politician will be dead.

The pomegranate <> grenade connection all of a sudden made “grenadine” make sene.

They stake a lot of money on being right.

Or that the meaning of these phrases has changed over time.

> if you are foolish enough to pay markup on items that you could order yourself directly from AliExpress that's your own problem

How do you expect the average consumer to automatically know if a product they’re looking at is being resold from AliExpress?

A new role has arisen in the last few years that captures much of this responsibility - Analytics Engineer.

This article by Claire Carroll describes the role and motivation for it https://www.getdbt.com/what-is-analytics-engineering/

Cyber attacks nowadays demand their ransom payable in crypto.

This is the weirdest part of the article by far. And absolutely no explanation.

Lecture 1 of that series is surprising and hilarious for a class about databases.

Fortnite as of 6 months ago had 25 million daily active players.

San Francisco had some of the heaviest lockdowns in the country, yet 100 miles east in Modesto the mayor actively defied the statewide lockdowns. California is a big state.

For last 20 years in the US, we have been recovering from decades of misinformation about the “dangers” of fat. The avocado is one item that benefits from a return to nutritional sanity.

That shift has laid the foundation for items like avocado toast or Chipotle’s guacamole to help popularize avocados.

There are lots of other factors, I’m sure. But it helps that they’re tasty.

Check out Facebook’s Gorilla paper (https://www.vldb.org/pvldb/vol8/p1816-teller.pdf) which features some clever compression techniques for time series data generated at regular intervals.

Qatar is the highest in the world at 49 tons per capita per year. So you’re off by a lot.

That doesn’t make 416 tons a lot by any stretch, but we are still early in the technology cycle for carbon removal/sequestration.

I don’t know if it’s just that complex 3D topology is hard or what, but those blue-yellow knot diagrams have never helped me understand a knot.

The fact that I can use git for data if I carefully avoid all the footguns is exactly why I don’t use git for data.

The article briefly mentions RPL. It’s a programming language for Hewlett-Packard RPN calculators. It’s an interesting curiosity if you’re in to niche programming languages.

https://en.wikipedia.org/wiki/RPL_(programming_language)

Thanks for the in depth explanation! I look forward to exploring it more.

They call this issue out in the docs specifically when the fat column is a sort/distkey. rsbb in the Intermix slack has studied this in the most detail I know of outside AWS but they haven’t published their book yet.

Cool! So if I select two columns with a predicate in the first, does it scan all stripes for the second column?

One of the gotchas of columnar storage (coming from Redshift) is that you lose all of the compression benefits if you have just one column that’s fat or hard to compress.

In Redshift columns are stored in blocks. You want to fit roughly the same number of column values per block across all your columns. But if you have one column where a small number of values can fit in a block, the rest of the columns end up leaving most of the block space unused. The result is wasted disk space and poor query performance.

This Postgres extension has similar-sounding storage ideas with stripes, but it’s not clear to me if it suffers from the same issue.

My first test to vet this would be a table with 50 columns of ints and one column of md5 hashes stored as varchar.

They provide a platform that lets users share playlists.

None of this money goes to Spotify.