Communism Jokes

@RongSpeller @Landeur @alifnomad

@Landeur @alifnomad Calm down gentlemen. Elon Musk has neither USA nor British DNA.

This will sound incomprehensible to many people, but this could have been Britain. We could easily have become leaders in AI and space if we wished to. Instead we built an international welfare state.

@AdamHoltererer building vertically is much more expensive

Dumb question, but why not data centers, but vertical?

@sean_j_roberts @Sceat_ try TDD with your agents. In my experience, it produces much better results. Instead of mentally tracing what the agent is doing in the production code, I can rely on the test layer to verify that the behavior is correct. It makes it much easier to iterate with confidence.

@Sceat_ I don't do TDD and I don't tell my agents to do it either. If you write tests second it's good practice to make sure you can make them fail

I watched Opus 5 delete a line of code to prove the test it wrote actually tested what it thought it was testing earlier today I've worked with a lot of developers who never did that

@brankopetric00 that's why you must know what you are doing.

ChatGPT is quietly making DevOps engineers worse, and almost nobody is talking about it. Here's the problem. You ask it "should I do X?" and it says "Great question. Yes, X is a solid approach." You ask it "should I do the opposite of X?" and it says "Great question. Yes, that's a solid approach too." It agrees with everything. It's a yes-man with a keyboard. And a yes-man is the worst possible teacher, because growth comes from friction, not applause. I've shared free DevOps content for 2 years, and the engineers who actually level up are the ones who get pushed back on. The ones who have to defend their thinking to someone who won't just nod along. Real senior engineers don't cheerlead. They ask "why not the simpler option?" They ask "what breaks at 10x traffic?" They ask "did you check for a forward-only migration before you rolled back?" That pressure is the education. So I built an AI that argues instead of agrees. It takes the senior engineer's seat, challenges your call, and a strict grader scores whether your reasoning actually held up. The opposite of a yes-man.

@minallkamel @mherlihy0816

@mherlihy0816 probably someone is still typing it

does anyone know the German word?

@Tazmania_ct y como te los esconden en la historia que enseñan en Cuba. El tipo era un penco de lecheria

El meme del día sin dudas es este 🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣🤣

@IamAroke Yes, and my point about ORMs is that they often prevent you from targeting queries precisely. You usually end up either lazy-loading relations and creating a large number of queries, or eagerly loading an entire relation tree and fetching far more data than you actually need

@communism_jokes many of the Systems having this issues serve a lot users and are very critical

The real bottleneck in most backend systems is still the database, not the application code.

@IamAroke You can still scale databases horizontally, you just need to architect your system with replication lag in mind.

@communism_jokes Nah, I’ll say it’s inability to be scaled horizontally as users increase

@kai_fell @NotNordgaren So by their definition C is also memory safe but with a global unsafe block

@yokoseineko @kai_fell @NotNordgaren

this program compiles without errors and segfaults at runtime I wonder why!

@wonderDeveloper @Tazmania_ct

@Tazmania_ct Eso no es por pinar del rio por allá ? El abajo del sol y lis encajada hasta la garganta. Que clase tipo mas singao.

ajajajajajaja

@AiKathG @open_parens yes, but the JVM is already extremely well optimized for all that crap. GraalVM can also brings you the "native" feeling on top. Just look at that syntax.

@communism_jokes @open_parens Despite being different approaches (runtime vs. compile-time), reflection in Java/C# is no way better

C++ should simply declare bankruptcy at this point and start again.

@unclebobmartin I've never written this many tests before, and it's honestly incredible. AI is automating almost every part of my day-to-day work. I'm finding and fixing production bugs in 20-30 minutes instead of spending 1 or 2 days on them. Tracing and debugging issues has never been easier.

Morning bathrobe rant: 20X.

@chainmonkey2 @IamAroke i guess your domain is too flat then

@IamAroke No. It’s because you have a relational DB

@StokeWillie @ChShersh no, you can't.

@ChShersh Sure, the syntax is verbose, but the funny thing is that you can read this code — or at least understand what it does — without any knowledge of C++ reflection at all.

There are cathedrals everywhere for those with the eyes to see

@SakshiSugandhi let me guess, you don't have the degree, right?

Why do colleges spend 4 years on theory when companies only care if you can build and ship?

@MoureDev Eso es márquetin barato

Lo de los medios de “información” hoy en día es una locura. En cuanto vi este titular, me puse a buscar las fuentes, porque chocaba directamente con el discurso habitual de Dario Amodei. Y esto es lo que encontré: La frase existe, sí, pero Amodei no estaba expresando su propia opinión. Estaba relatando lo que dijeron algunas empresas que probaron Mythos de forma anticipada. Él no afirma eso en ningún momento. Sin embargo, la noticia elimina ese contexto en el titular y, además, asegura en el cuerpo que Amodei lo “admite”, presentándolo como si fuera una declaración propia. El titular es engañoso: modifica lo ocurrido y genera más miedo que información (todo puede comprobarse en la entrevista original de Bloomberg). En lugar de facilitarnos el acceso a información fiable, cada vez tenemos que dedicar más tiempo y esfuerzo a verificar lo que publican... Cada día hay menos diferencia entre un tuitero random y un medio profesional.