ic_fly2

I can recommend his introduction to energy models. It is a bit older but explains the idea very well.

CIA favours lab leak theory after a new head is picked. A person that has always been a vocal supporter of the lab leak theory.

No facts changed, just a Trump appointee starting work.

Have used dell business laptops for many years now (Vostro 13 and friends)

Used to be amazing, first one lasted a very long time (8years), my third one is currently giving up the ghost. And while Linux used to work great on them the uefi bios is so locked down that I couldn’t run it on the one i currently own.

On a related note, I’m out shopping for a new laptop and am considering a framework…

The whole appeal of Django is that you can just swap in a different templating engine / syntax if you prefer.

And you can add a DB and let there be managed by some other system.

Same goes for extending manage.py

I’ve worked on Django projects big and small and generally the speed are which we could deliver even on strange edge cases impressed.

The comments here are somewhat confused, understandably as the article is too. Cash is really expensive for shops, many digital payment methods cost a fraction. (Not in the US perhaps, but that whole payment system is antique)

The interesting question is how anonymous digital payments will work with stable coins. While in theory perhaps traceable, in practice not doable at scale.

We might be going back to a world with more cash like payments.

We run a decent Keras model on production.

I don’t need a custom loss function, so keras is just fine.

From the article it sounds like Waymo run on Keras. Last I checked Waymo was doing better than the PyTorch powered Uber effort.

Reminds me of the contrast between a small Bayesian Modeling conference that was a discourse chat and a big one of a snake themed programming language. They had a custom game where ypu could walk around, a custom video player and all sorts of fancy stuff.

Obviously the discourse conference worked great and the other one was an unmitigated disaster. The tech didn’t work, the talks were bad and the interaction was 0. Only a few influencers hyping themselves.

You make a very valid point.

I’d ignore unpriced other externalities, that is a large rabbit hole. (Do people increase consumption of other goods thanks to the savings of this ticket…)

I’d rather look at the project, subsidies again and the cost and realise that at this scale there are few individual projects. So this is a very large knob to dial down the emissions.

Secondly, the subsidies go into the German economy. The rail system is very domestic economy heavy, so at a time where the German economy is going into a recession, spending 3bn to subsidise transport and reduce people’s cost of living and pay for German workers and investments is not such a terrible idea.

So while the price tag is high, the money isn’t turned into trees or buried like in pure abatement projects and at this scale it is a nice leaver to pull.

As someone running neo4j in production I can just warn that the DBs are a pain and need a lot more care and love than Postgres or Oracle DBs. Even much larger instances. Maybe their cloud offerings are better, but they are quite expensive.

1) depends, I find ML much more interesting than software engineering.

2) Not me

3) With difficulty, but through courses like fast.ai

If you’re not into maths and problem solving this is probably the wrong path. The main value add you bring is being able to transform a business problem into solvable maths.

Read Introduction to statistical learning and look at the fast.ai courses. I also recommend the paper attention is all you need. If you find all of those things interesting and not too mathsy then I’d say go for the switch. Else, maybe look at options in data engineering or as a software engineer in a ML team.

I’m all for upskilling.

On your example, when I was an intern in a factory, I was banned from the pneumatic tools with a counter grip. Because if you use them incorrectly your finger/ hand / arm is a flesh pancake.

My solution suggestion here is definitely to empower more people to work on this by teaching them basics of data base design and enough python to write a dag in airflow.

We’re in the process of shifting an entire business from db stored state driven by large (Java and .net) apps to an AMQP based system.

I can’t really say I’m enjoying it. But it does help with scale

With all the data issues strong quality and normalisation I often get the impression that enabling more people with non CS backgrounds to do this work is not necessarily a good thing.

In other words, if writing python and sql is the skill requirement that stops you from making an etl pipeline, maybe do something else.

It should be noted that the previous formula was from before 1983 when it was changed and it was changed because it overstated inflation. So a better headline should be, academic sour about being wrong all the time uses outdated formula to calculate higher inflation number. People are angry about inflation now because they compare prices to before the pandemic, so cumulative inflation over a longer period than one year. Also given most mortgages are 30year fixed, the change in 1983 was the correct thing to do, though it could be updated to use internet data on rents more.

Section 230 is at core of problem. And the lobbying efforts the profits can buy.

Legally wsj nyt twitter vogue and instagram are completely different despite clearly being the same product.

It’s insane these platforms aren’t the same.

If you don’t believe this compatible free speech on the internet, post content you don’t think belongs on the internet here and see how quickly the community gets rid of you.

It’s true that Django and especially DRF feel a bit overkill. But for any app that does actual business the extra bits have often saved my bacon.

Sorry to hear this. This will end in drama.

Work out what drama is acceptable to you and work towards that. This will get ugly.

Your partner clearly makes quick penis driven decisions for good own benefit. This can’t work.

Call me old fashjioned but what is wrong with django and htmx?

Works beautifully and fast, send only rendered code to the front. Have db admin for when you scale

My understanding is that people prefer power bi to tableau. So that says everything.

This is amazing, I was maintaining my own custom solution that did this.

I have a small page that modifies my get requests to google by adding -site:… for a bunch of most annoying content farms for stuff I search often (docs)

Within the normal flight envelope this holds. But getting into a spin in a glider in ms fs shoes you the limits of lookup tables / their implementation as the behaviours is completely wrong. ( At least ask the way up to fsx)

As a non American outside of the US I’m sort of gleeful. Nice to know that this shit is done to everyone.

Maybe some us based tech giants could do something about it or just not go and fuck over any non US based service

France has strong support for French (language) music.

The end result is that a lot of it is pretty decent. (Sure there is complete trash, they don’t have better tastes than others)

Some rappers write books that win literature prizes. Plenty of French language hip hop and rap has deep lyrics that go way beyond what most English language stuff achieves.

No, because their neighbours would send them power. And so on. A larger grid is much stabler. It would also extend outside of the area badly affected by the snowstorm.

Simple proof, no where else in the US or any other developed country does such a situation arise. And even most developing countries have better connected grids. When southern Australia didn’t, it paid for that in rolling blackouts.

It seems really stupid to loose 200 billion USD in an avoidable event.

And I don’t know the word for the state of mind for risking losing that again.

You can but a lot of snowplows, interconnectors and backup plants for that money.

Reminds me of London, which of never prepared for snow as it never happens. Happened twice in the short time I lived there. Each time costing a lot more than a bit of preparation.

Like to add some context.

The 2021 event cost north of 200billion USD and killed people.

A similar event, even half the magnitude, is likely, and very bad.

It is also an active policy decision by Texas. This would be easily fixed by integrating into the national grid and applying some supervision to the utilities.

Texas, it’s people through it’s politicians, chose this situation.

Yeah, happens quite a bit. Also just being able to dig and figure out why something is doing a weird thing happens all the time. I’d say about once a month in my team of three people, and we’re not even doing fancy stuff.

Most recent examples I can remember: polars data conversion errors, batch sizing in azure service bus sdk, some weird validation thing in pydantic and the performance monitoring in sentry not working as expected. The pydantic bug got a patch / note in the docs that was actually already merged.

You’re not going to need this for a js library that doesn’t have a saas component.

I like it.

I totally get the problem projects like mongo, redis and sentry face. AWS, azure etc just eat their lunch and don’t help anybody. That will kill open source in the long run.

This still gives me access to inspect the code, contribute fixes I might want, self host if I feel like it or pay someone for the SaaS if I don’t, knowing the money goes to someone who contributes to the project.

And if the original authors bork the product and after two years it becomes clear they won’t fix their ways, I can fork the old product code I liked and develop it into the product I want. Heck if I’m right and my idea is better than the original I’ll eat their lunch too.