Kisliy

Your project doesn't live in one file, so why does your editor? I'm building NullSpace: your whole project on one spatial space. Terminals, docs, Git, agents, local servers — all in view, nothing buried. Solo founder, building in public. Private beta 👇 NullSpace.codes/x

@matthewmillerai Chines models not today not tomorrow maybe next year ;)

GPT 5.5 and Claude Opus 4.8 are the only models I use. I can't imagine living in a world where it would genuinely be better to use MiniMax M3 or Kimi K2.6 for coding. Things like speed, output quality, and reliability are very important in todays world. I find that OpenAI and Anthropic are so far ahead it is not even a question as to which model you should use.

@akhil_kestur @NullSpaceCodes Carries everything over, no cold start. It swaps the account under the hood, your conversation + context stay live. Literally one button press.

@Kisliy_K @NullSpaceCodes does a swap carry the conversation state over, or does the new account start cold?

Day 1 / 100 — building NullSpace in public. The Claude Code limit wall kept killing my flow mid-task. So I shipped a spatial account switcher: watch your quota, swap before you hit 0, never lose the thread. 3 on the waitlist. Let's go 👇 NullSpace.codes/x

@JustJerry121 @NullSpaceCodes Exactly, nobody wakes up wanting a "spatial editor," but everyone knows the pain of losing flow. Quota was just my most recent one. The canvas is the answer to a problem you already feel.

@Kisliy_K @NullSpaceCodes Starting with quota pain gives people a concrete reason to care about the spatial editor.

@bridgemindai Well, that's Chinese nothing now nowadays at real

GPT 5.5 and Claude Opus 4.8 are dominating coding benchmarks. Chinese open source models like Kimi K2.6, MiniMax M3, and GLM 5.1 are insanely behind. People keep saying that open source is catching up. That's not true. Claude Mythos and GPT 5.6 are both about to drop. The gap is about to get even wider.

@compileandpush Neither... Nullspace isn't the thing that goes down, it's where you watch everything that needed. Prod logs, Git, agents, terminals side by side, so you catch it before the consequences show up. The editor is the cockpit.

@Kisliy_K The difference between a homelab and a production environment is usually just the consequences when it goes down. Which is this?

Your project doesn't live in one file, so why does your editor? I'm building NullSpace: your whole project on one spatial space. Terminals, docs, Git, agents, local servers — all in view, nothing buried. Solo founder, building in public. Private beta 👇 NullSpace.codes/x

@siya_twt_ IDE nowadays is going on the back, you need to look at Agentic Developer Environment. This is next level

developers, which IDE genuinely has the potential to replace VS Code?

@dark_coderz Wow, I am impressed. Well I hope one day, my 100days challenge will show me same results. Wish me good luck:)

20 days ago: → 0 followers Today: → 1,260+ followers → 5.5M+ impressions → 745+ verified followers → Monetized What's one goal you're chasing right now?

@RajputNikh13836 macOS is literally 10x better than Windows

Windows is literally 10x better than Linux. People just don't want to admit it

@omarvvvr Just shipping your product non stop it will help, and of course work on it well. You don't have time to think about some problems. You only need to ship product)

As a solo founder, what breaks you more? - no users - no revenue

@pcshipp Yea, mate. I just few days ago remove my subscription. Bought claude one, well I love it

Guys, Codex is too bad these days It keeps showing a crash screen, gets stuck on analyzing, and crashes Is anyone else facing this problem, or is it only me?

@bridgemindai I see

@Kisliy_K It hasn't been added yet

LMArena dropped a new Agent Arena benchmark. GPT 5.5 took the #1 spot. It ranks models on how well they actually orchestrate tools for real agentic tasks. This is the metric that matters for agents. Not trivia, not chat. Can it run the tools and finish the job? The agent era has a new leaderboard.

@Kappaemme1926 Canceling Codex and buying Claude code

Codex limits are getting painful. So What’s the best GPT-5.5 setting for coding?

@isiliconx @justbyte_ I don’t know, i have free pro for 1 year, it is unusable at all. Rate limit’s in 1h of work, on 3.5 flash. Also remember please that 3.5 it is not even close to opus models, on each benchmark ;)

@Kisliy_K @justbyte_ Rate limits used to be extremely good

Why do people not use Gemini for coding?

@bindureddy Well price of Mythos is a bit high 🤏 Bit it will cost this money definitely

Mythos will be timed to launch with GPT 5.6 and Gemini 3.5 The only problem is that it costs $70 per 1M output tokens 😲

@C8Luna @samuel_0xK Well, nowadays you need to be fast adapting to reality. As well to the prices and quality pf product. So actually nothing will be changed

@Kisliy_K @samuel_0xK people say that but I bet now that you have been spoiled there will again be sticker shock from codex or claude.

Github copilot isn't usable anymore :( just one single small codereview costs 61 credits???? do you still use github copilot or did you switch after June 1?

@rezoundous You need to try next gen of IDE, it is ADE) much much better experience

Anyone still using IDE?