mind-blight

It seems pretty clearly spelled out. If you have enough traffic that an events table is slowing down your postgres instance, you can easily set this up as a service to offload the events table. The author says <1 million MAUs, and you probably don't need this.

It's built on tech known for handling very large amounts of traffic, which answers the how after the what.

My understanding is that just vegetable oils are generally bad for cooking because of how they retain heat. Even if they have a high smoke point, it's easier for them to get to it.

Have you doing that she has the same issue with coconut oil? Specifically, the saturated ones that come in a jar?

I'm also curious how it would affect CSAM proliferation. That's one of the biggest angles of attack on encryption (see how many times the UK has tried to ban it). If their techniques mitigated that, then it could take done if the wind out of the sales of folks trying to weaken encryption

I mean, the quick shift makes sense with vaccine deployments and the medical community figuring out how to deal with it all happening within a few months. We went from not being equipped to deal with the disease to being equipped in a pretty short window.

If governments were using covid as an excuse to control their population, then I would have expected them to hang on to the rules for as long as possible. Instead, we saw rules change as the context changed. That's generally not what totalitarian takeovers look like

Yeah, I was there from 2013-2016. I got to use a few of the products that would eventually turn into Foundary. I've actually hung out with the author a couple of times (I was on the AirBus deployment as well as in the building right next to the healthcare team for a while, though we didn't actually interact on any deployments).

Going for operation use cases was a huge win. Once novel data existed in the system (rather than just transforms of existing data), it became a lot harder to rip out. That could be as simple as having someone merge records so you know that two companies are actually the same.

Foundry was a really interesting case because it was basically an enterprise ETL platform before those became very popular + a team of people who helped you get data into it. One of the genius things about the business model was that it operated like a consultancy, but built contracts like a product company. That allowed them to charge based on the value provided rather than hours worked, then pull the best lessons from the deployment back into the main product.

You should really take into account the advent of COVID vaccines, the evolution of the disease (it appears that less lethal variants), and how human immunology works (people create antibodies if they survive the first round of a disease).

There's a tendency amongst folks who have strong opinions on covid measures to create false dichotomies and ignore how context changed over time. Lockdowns appear to have been a good idea during some of the disease (i.e. before we knew how to treat it, and before vaccines became readily available), and became less important as the context changed.

I've started receiving emails that say 'reply "unsubscribe" to stop receiving emails' rather than have an unsubscribe link. This just started happening a few months ago, so I think this is a workaround that someone figured out.

I've started blocking all of them and sending straight to spam.

I've looked into it a bit. I'm getting to the point where I either want to use a full ORM, or I want to write raw SQL and have it checked against the database.

Query builders are nice, but they sometimes end up in this middle ground where they have the worst parts of an ORM (needing to learn a new syntax and not getting access to all the DB features) and the worst parts of raw SQL (Unpacking relationships into objects w/ pointers is error prone boilerplate that sucks to write).

I love SqlAlchemy since it's still enough of an ORM to solve the main problem I want solved. Super curious to hear if you've had good experiences with Keysley and how it deals with transforming relationships into objects

It probably is, but you can do that workflow with DuckDB too. It's just really flexible since it can all be done in SQL. Check out the DuckDB postgres connection (https://duckdb.org/docs/extensions/postgres.html)

To import from a Salesforce Bulk Export CSV (for example) into a postgres table is a few lines of SQL

``` INSERT INTO pgdb.accounts ( id, created_at, updated_at ) SELECT "Id", "CreatedDate"::TIMESTAMP, "UpdatedDate"::TIMESTAMP FROM read_csv('/path/to/file.csv') WHERE "Id" NOT IN (SELECT id FROM pgdb.accounts) ```

The path can be in a cloud storage provider as well, which is really nice. You can do updates which join from the postgres db or the other CSV files (or a MySQL database) as well. The data transforms (casting to timestamp, uuid, etc.) have been super handy along with all the other SQL niceties that you get

Honestly, I hated Prisma for the longest time. Like, hated it. I tried to rip it out of projects multiple times. Why would you build a node library in Rust? (as one of many problems) But, I had a mental shift recently that helped me appreciate it more: I use Prisma only for features that fit neatly into an simple ORM (i.e. building a web page based on a bunch of joins). Anything else, I use raw SQL.

They released TypedSql (https://www.prisma.io/docs/orm/prisma-client/using-raw-sql/t...) which is heavily inspired by PgTyped. That lets me write raw Postgres SQL that's converted into TypeScript. The other things I do:

- If I want derived data, write views that encapsulate the transform. Prisma supports reading from views - If I need something more complex, use DuckDB + python for analysis and write to the appropriate table. - If I need to cache complex queries, just use a materialized view and read it as a prisma object

It's not perfect, but that let's me use prisma for what it's good at (Managing an ORM and deeply nested queries), then fall back into raw SQL for everything else.

Going straight to SQL has been a breath of fresh air, but, let's be honest, dealing with deeply nested joins really sucks when all you want to do is build a page that shows a company, all of it's people, and all of those people's relationships. ORMs are pretty handy for that last case, and I use SQL for everything else

I've been using duckdb to import data into postgres (especially CSVs and JSON) and it has been really effective.

Duckdb can run SQL across the different data formats and insert or update directly into postgres. I run duckdb with python and Prefect for batch jobs, but you can use whatever language or scheduler you perfer.

I can't recommend this setup enough. The only weird things I've run into is a really complex join across multiple postgres tables and parquet files had a bug reading a postgres column type. I simplified the query (which was a good idea anyways) and it hums away

I'm a huge remix fan. Remix, Shadcn, postgres, and Prisma have been great for rapid development

I'm not sure where the derision is coming from. It seems like they made an assumption about something they have experience with, got a bunch of feedback, changed their behavior, then took the time to write about their journey for the rest of us.

The author is engaging with everyone in good faith. This kind of openness and willingness to change would be applauded.

The same thing happens in the US. There's a court case where someone put a EULA inside of a CD packaging and made opening the CD sufficient to trigger the EULA. They won, making that an acceptable easy to force a license on someone

Most variables on JavaScript are (essentially) pointers. A common mistake people make in the language is keeping references to objects in a global map, which prevents them from being garbage collected (often a bad caching implementation).

You can use things like WeakMap or WeakRef as one solution, but there are usually better options

I one thing that would help at the federal level is a dosage guide, similar to how alcohol works. The drink equivalent is a useful starting point for beer, wine, and spirits. There isn't a "normal" dose, so everyone just asks their friends, and the people they get advice from are generally regular stoners with a much higher tolerance. I've had to argue with someone to convince them that 10mg from an edible is a very high dose for most people. I fortunately convinced the guy who wanted to try for the first time to take 2.5 and just set how that works for the evening

It is, though apparently folks tend to titrate to the same blood number with smoking since it hits so fast. Obviously not perfect if you don't know to wait a couple of minutes, but it seems like experienced smokers get the same amount of THC from old strains with about 5-8% as they do from newer stains when 25-30%

Most of the research I'm aware of shows the the link isn't causal. There aren't higher cases of schizophrenia in countries with higher cannabis use than without. Specifically, US teens use marijuana at much higher levels than many European and Asian teens, but rates of schizophrenia appear to be the same.

ETA: Andrew Huberman did a podcast with Dr. Matthew Hill (a researcher in Canada) that covers a lot of the most recent days and studies

It looks like a crypto scam. I just watched the whole thing, and it runs in a loop, mostly talking about crypto and signing up for a crypto wallet

So are gravity and friction. I don't know how well tested or accepted it is, but being just a theory doesn't tell you much about how true it is without more info

I think graphile worker is Node only. This project is for Python.

This is the exact use case I'm running into right now. I've been looking at BullMQ some it has good typescript support, and is working towards a 1.0 for python. But, I have tried it out in a production stack yet

One caveat: at will employment doesn't allow a company to fire you for "any" reason. It allows you to be fired for "no" reason.

So, if a company says "we're firing everyone wearing red shirts", then you can potentially contest that. If they fire you without a reason, then it's legal. If you can prove there was actually a reason, and that the reason was illegal, then you can sue them for wrongful termination

I mean, that's fine if you discuss it up front with the company before heading on site. You severely limit your options, but I know a few companies that have agreed to payment.

Though, it's worth noting that the company is already spending a lot of money interviewing you. Especially if they're paying for flights, lodging, and meals for an on site.

It's worth figuring out what your goals in demanding compensation for being interviewed are. If it's making sure the company isn't "taking advantage" if your time, then I think that's already covered (interviews are super expensive expensive for companies) so long as 4 of those 5 hours are with actual humans. If it's due to financial hardship, most companies will be sympathetic and work with you. If it's to satisfy a sense of moral righteousness, then I think there are better options than sending and invoice after the fact (which comes across as passive aggressive and would ding you on your communication skills assessment)

I've been on both sides of plenty of multi hour assessments. Unpaid training is absolutely not the reason. As an employer, I don't want to spend time training anyone unless they're actually going to be hired.

It's really hard to figure out if someone is going to be a good fit for a job, so the idea is to get as many different perspectives as possible. Hence why you often interview with different people on different roles as part of the process

I'm not sure if there was a specific article your trying to link me to - that's just a list of all FDA cannabis regulation articles.

I'll admit that I didn't read much of the article. My knowledge comes mostly from working in hemp from 2017-2019. I was asking back then went people weren't doing this since, based on my reading of the law, it seemed perfectly legal (if not very sketchy). I'm not sure if the regulations have been updated since then, but it seems like 'no' based on current behavior

Totally depends on the product. The federal farm bill states that hemp products need to be under 0.3% THC to be considered hemp. How high you get depends on the milligrams. So, edibles (gummies, etc) can remain under a 0.3% THC by volume in the product while still having more than 5-10 mg per edible.

You're not wrong about people being able to make your life hell, but the edibles absolutely fall under the letter of the law. (The one exception being the concentrate used to name the edible with have a large percent THC until it becomes diluted again)

Long running transactions can easily lock up your database. I'd definitely avoid those. You're better off writing status records to the DB and using those to determine whether something is running, failing, etc.

I'm really curious how you folks compare to something like Apache Airflow. They do a similar durable execution w/ DAGs on top of postgres and redis. They're Python-only (one definite difference). I'm curious what other comparisons you see

ETA: I really like the idea of this being entirely built on Postgres. That makes infrastructure a lot easier to manage

I'm usually a +1 for the empathy train and second chances, but it seems like the guy legitimately committed fraud and hasn't taken responsibility for his it his company's actions.

There's a big difference between an honest mistake and a dishonest one. Given the harm he's seemingly caused people through the dishonesty, it doesn't make sense to give him the opportunity to continue lying and harming people

Edit for typo