Mădălin Ignișca

@GraemeVIP @claudeai I'm using both Claude and Codex on their maximum plans, and most weeks I don't hit the limit. Rarely maybe 15-25 euros extra in a week when I push a bit more (I enabled extra paid credits both). And thanks to their integration, I use both in Claude Code ;-)

Allow me to fix your post, "You can now run multiple Claude sessions side by side for around 5 minutes before we lock you out and demand another 200USD off you". I can burn through the 5h limit in about 8 minutes. That's absurd in anyone's book. I can absolutely hammer Codex for hours (same tasks) and it doesn't run out. Both on the same plan. How on Earth is anyone getting work done with Claude? GPT5.4 is a more advanced model than Opus 4.6 right now anyway - especially since they nerfed it for the Mythos/4.7 release.

We've redesigned Claude Code on desktop. You can now run multiple Claude sessions side by side from one window, with a new sidebar to manage them all.

I think that it is OK that @claudeai has issues sometimes. It forces me to take a break, walk the dog that was staring at me for 2h now... 500 errors should be acceptable every 50 minutes for 10 minutes for the physical help of users.

@bcherny for the last few days, Claude Code CLI asks occasionally if I allow session transcript sharing. Personally, as I use it for side personal projects, and I'm open for it, is there a config variable I can make it YES permanent?

@addyosmani @googlechrome I never understood this idea of vertical tabs on other browsers. why is it better? laptops have a 16:9 or 16:10 ratio. vertical bars make sense on 21:9 or even wider screens which probably are 1 in 100000 cases used.

Vertical Tabs are officially in @googlechrome! Right-click any window -> "Show Tabs Vertically". Boom.

@Hetzner_Online I think the cloud pricing page is not making sense for any visitors anymore. The previous version was very easy to understand, understanding pricing per locations. Now it is a long table without an understanding what is the difference between cpx31 and cpx32

Conclusion? #GeminiAI provided for more or less same amount of spending way more value and help for me to work compared to #codex Maybe @OpenAIDevs @OpenAI should review their limits compared to the competition as it's the worse in the AI big league.

Codex? One trial, and I exhausted allowance of 5h... And I'm not even putting it to hard work, it only triggers reviews for 2 parallel repositories with a commit every 40-50 minutes...

So OpenAI's 5h limits are so small compared to Gemini that literally it makes me wonder if it worths to pay the subscription. Disclosure: _read the comments for disclosure_ #ai #vibecoding #BuildingInPublic #ClaudeCode #codex #GeminiAI

Escape corporate tech! AI empowers you to break free & build your dream of helping small businesses thrive. madalin.me/ai-your-escape… #ai #vibecoding #BuildingInPublic

Unlock AI's full potential! Discover how one developer went from manual coding to managing an AI team, revolutionizing app development. madalin.me/from-manual-co…

@amorriscode Let me test remote sessions, with voice from the phone... It will be something like "Hey Claude. Check the current open PR, look for all Gemini and Codex suggestions, explain them to me, and go ahead and do which ones are good to do, and ignore the bollocks".

sometimes you need somebody to talk to we just rolled out voice mode for Claude Code on desktop and Cowork 📢

@adrianasobota_ @GergelyOrosz I have some personal projects built in public. Care to review one of them? I'm very open to feedback. Comment back or send me a DM.

When it comes to AI agents / AI tooling + coding, I hear an awful lot of talk about: Efficiency Iteration speed / PR output rate / lines of codes produced I hear zero mentions about: Quality Customer obsession This will bite back, and it probably already is...

@trq212 It's so good, that I did not even observed I'm not instructing it for saving "memory" the way I figured out previously.

now that it's been 2 weeks, what have you liked most about auto-memory? what's not working? if you have specific stories of when it went well or poorly, we would love to hear!

@noahzweben I've started using the feature like 20-25% of the time. Even have thoughts to switch to a larger iPhone (still use my 12 mini)

claude --remote-control claude --remote-control <name> to spawn an interactive session with remote-control enabled!

COMPLETE SECURITY SCANNING TYPES: #GITHUB & #GITLAB A merged, deduplicated list of all security scanning capabilities available across both platforms, with recommended open source tools for each scan type. madalin.me/complete-secur…

@noahzweben I was using 2.1.70 just a bit later you posted...

Remote Control is now available to Team and Enterprise users! Please update to 2.1.69. We are fully launched. Enjoy the fresh air <3 Default off but admins can enable at claude.ai/admin-settings…

I finally rebooted my blog from scratch! 🚀 Built the entire app exactly how I wanted using @ClaudeAI Code. To kick things off, my first post is a short dive into the Claude Code workflow itself. Check it out here: madalin.me/ai-assisted-pr… #BuildInPublic #AI #ClaudeCode

@jeffrey_way Me too. Opus 4.6 through Claude Code, or the Claude desktop app, or like others through Copilot or Antigravity? If from Claude Code, did u increase it to high level?

@MadalinIgnisca I use Opus for everything.

I've yet to see AI come close to that human thing where we step back and say "Okay, let's rethink this approach entirely." I just watched AI cycle for a half hour trying to solve a weird problem... when the solution was so obviously - "Wait, just don't do that. DO THIS instead."