Özgür Tanrıverdi

unpopular verdict: if you treat it as a slightly improved C and only selectively use the std goodies, C++ isn’t too bad.

looks like for me late 2026 will be full of jblow creations: i’ll either be cracking this game’s mysteries or coding in jai once it’s finally released.

Cashing out crypto soon? Don’t trigger an IRS headache or bank freezes: • Cash-out risks • Tax mistakes • Compliant off-ramps • Large exits Join our LIVE AMA today March 23 | 2PM ET Clinton Donnelly (@CryptoTaxAudit ) + Hugo Leijtens (@CenseAG) Learn how to off-ramp safely👇 streamyard.com/watch/VMFtaQcp…

as a software engineer: its surprising to me how frequently we are forced to explain "state reduction" even though how fundamental it is. i wanted to have a structralized reference to it for this very specific purpose: otrv.dev/state-reductio…

the amount of times i forgot to decode the input hex in cryptopals is embarrassingly high

its funny how in the agentic coding world grep and sed are the fastest growing “editors”

AI can make work faster, but a fear is that relying on it may make it harder to learn new skills on the job. We ran an experiment with software engineers to learn more. Coding with AI led to a decrease in mastery—but this depended on how people used it. anthropic.com/research/AI-as…

@benfurkankilic guvenli mesafe kesinlikle en sagliklisi 👌

AI öncesi de bu kadar hype var mıydı hatırlayamıyorum. Sürekli gezegenin en önemli tool’u o an çıkmış gibi davranılmasından biraz sıkıldım artık. Benim yöntemim: Çıkan tool’ları deniyor ve kendi hayatıma entegre etmeye çalışıyorum. Çalışıyorsa ne ala, iyi değilse güle güle.

@brankopetric00 spot on. decoupling is not always needed or valuable

Event-Driven Architecture is great until you need to trace a transaction through 4 queues, 3 topics, and a lambda. Sometimes a synchronous REST call is fine. Keep it boring.

when it comes to using LLM’s in software, one thing almost everyone seems to agree on is that we still need a human in the loop. this is why due to amdahl’s law, it cannot improve productivity as drastically as people claim unless it gets rid of this limitation.

@jorandirkgreef the portion of the system that can be made parallel

DistSys Interview Challenge Many factors may limit a system’s ability to scale. But which limit dominates a system’s ability to scale?

@matteocollina @MichaelArnaldi @nodejs @fastifyjs while i agree what i cant get out of my head is: due to amdahls law, if a human is in the loop then these tools will never live up to the productivity increase claims people make

@MichaelArnaldi @nodejs @fastifyjs The human in the loop isn't a limitation to overcome. It's a feature to protect. Full post: adventures.nodeland.dev/archive/the-hu…

.@MichaelArnaldi wrote that software development is dead. He's right about a lot of things. But he's missing something critical. I wrote a response 👇 🧵

@T_Zahil if you dont want to change your entire workflow but rather fit LLMs into it incrementally, CLIs give you that option

I'll be honest: I don't understand the hype around Claude Code and all those AIs in CLI It's SO much easier to have an interface (Cursor, Conductor...) to understand what's going on, to review the changes, etc

2026 is gonna be the year of composable software options like arch linux and @Neovim LLMs are quickly closing the UX gap that made such software niche

Abstractions scale code. Bad abstractions hide constraints and break decision-making. otrv.dev/when-abstracti…

A few things I've noticed as all devs write code with AI. When you write foundational / architectural code of a new project by hand, you "feel" the code pushing back if your abstraction isn't right. You feel when something is harder than it should be. The code is telling you it's not in the right shape. Good engineers are sensitive to this. When you're using an LLM, you keep pushing right through this in a way that feels like you're making progress, and it may even be directionally correct in a sense, but the underlying foundation of it all is actually bad in a way that either kills progress of the LLM later as it buckles under the complexity it has created or destroys your ability to maintain the code long term. Related to this, I see a general restlessness with just sitting and thinking about a problem for a while. As I've been working on a new library here at Laravel, there have been days where it feels like I mainly just stare at my screen thinking about something. When Claude Code is at your fingertips, it's tempting to just start yapping into the terminal and watching code come out the other end. Again, directionally correct in some ways, but often doesn't land on the elegant solution that is waiting to be discovered.

i member

@thdxr they recognize and leverage the human factor which is very rare for engineers in general

everyone talks about the bun team's technical ability - which is definitely a real and rare thing the crazier thing is how good they are at doing everything it takes to get something adopted so many of these decisions are counter intuitive to the majority of programmers

@_trish_xD interesting. to me the compiler errors is one of the weakest parts of Zig for now

trying zig for the first time compiler didn’t just throw errors, it explained why i was wrong like a mentor

ready to jump on the train

MCP is a very bad protocol built around a very valid problem