Alex Martossy

@alexaka1

I try not to be tismy. 💻 If you poop with the door closed, then you DO care about your privacy & should care about online privacy too!

Hungary

github.com/alexaka1

Joined January 2010

322 Following

132 Followers

Replies Reposts Media

Alex Martossy
@alexaka1
Oct 6
.NET is NGMI. They have embraced agentic engineering, my issue got a fixing PR by Copilot within an hour (!). But I have to wait for a full year (!) before the bug is fixed in a stable release 🤡 @dotnet github.com/dotnet/runtime…
Alex Martossy
@alexaka1
Aug 15
Replying to @DamianEdwards and @EricRichards22
My long standing issue is .net is the worst of all worlds. Monthly patches, but features breaking don't get fixed for at least 2 months after breaking. New features are only added to preview releases. Most bugfixes are only fixed on Next. LTS is useless if the tool itself breaks
Alex Martossy
@alexaka1
Oct 2
This model can doomloop at 20x the rate as other frontier models
Joana Carrasqueira
@jocarrasqueira
Oct 2
1M output tokens in a single run! 🔥 Most frontier models cap output at around 64K tokens. @GoogleDeepMind once again increases the bar as Argon goes to 1M, as an industry first. Gemini 4 Argon just changed what "one prompt" can mean.
Alex Martossy
@alexaka1
Oct 2
The current state of ChatGPT app is proof that OpenAI is no longer using it internally. Travesty
Alex Martossy
@alexaka1
Oct 1
They named it after noble gases because their models are completely inert.
Google
@Google
Sep 30
Today we’re introducing Gemini 4 Argon. It delivers frontier performance in complex workflows across real-world software engineering, knowledge work, and cybersecurity defense with an industry-leading 1M token output limit.
Alex Martossy
@alexaka1
Sep 30
Codex has completely fucked up worktrees in the app. Great job OpenAI. Environment files are no longer respected, so there is NO WAY to set up worktrees correctly on creation. Amazing job truly.