Just mock the repository which calls the db, and to test the queries themselves, use the real db as a docker container, or integration test with an existing db

Bruh stop with the middle class bs, there’s only working class and the owner class.

I know it's not helping, and off-topic, but I cannot resist noting that the ecological niche in Hacker News comment threads, previously occupied by doctrinaire libertarians, have now been almost entirely re-stocked with doctrinaire Marxists. Nature is healing!

The middle class is the designation for the part of the population the owning class has duped into thinking they are temporarily embarrassed millionaires.

I’m not sure if you’ve seen housing prices lately but most of the middle class are technically millionaires now, just of the “land rich, cash poor” variety.

Back that "most" up with facts. I'm not sure how much we can trust Zestimates.

6-figure AI scientists are middle class workers increasing the roi on investments made by billionaires into gpu capital.

It’s to replace qa testers for video games

They’re expensive and ruin our bottom lines.

Because you can’t scale out just the cache part of Postgres, one machine can only have so much memory

If you have a second machine, why not just put a Postgres read replica on it? Letting the WAL deal with replica consistency is much simpler than making the client responsible for keeping an external cache in sync, and you get the benefit of keeping everything in Postgres.

Either I pay a performance penalty waiting for my cache entry to be synced to the replica, or I risk reading stale data from the replica, no?

The new Macs stay really cool

My thoughts, dreams, plots and my schemes….

Price control is very half assed, what needs to be done is nationalization of the supply chain for food.

> nationalization of the supply chain for food

Historically, that doesn't work out so well.

Alan wake 2 pretty much requires it too

They were not indiscriminate in the murders. The leaders are always given a choice of paying tribute before hand. And they do stay true to their word if tributes were paid. It’s one of the major reasons why the mongol empire was so successful.

The idea that being forced to pay tribute or die is not murder is laughable. The idea that the Mongol empire was successful is also laughable. It could not even last 100 years before falling to infighting. And how do you even measure success? They killed 11% of the entire world population. Why would you want to remember anyone like that in a positive light?

+1 to this, reactive programming is a major pain in my ass. With virtual threads there’s no longer a need for them.

In general yes, but FWIW measurements point out that reactive vert.x seems to handle saturation much better. (1)

Also, vert.x are already exploring how to adapt to the new threads model. (2)

1 - https://gist.github.com/vietj/fe9f886d489853ab07111ba5715b13...

2 - https://vertx.io/blog/vertx-virtual-threads-incubator/

Just curious -- have you personally confirmed this?

Not trying to cast doubt... just noting that I have exactly the same feelings (hopes?) but I have yet to personally confirm it, in large part because it's not yet fully GA so we haven't jumped in with our production workloads.