Bernard Kolobara

@kolobara.com

771 Followers 463 Following

Creator of https://lubeno.dev and https://flawless.dev. 🦀

A screenshot of the console output, with the final few lines being: | Provider | CPU / RAM | Region / CPU | Clone | Install | Typecheck | Workload total | Result | | **luisa-sh-mac-m4-metal-8c** | 8 CPU / 21.49 GiB | eu-central-1, | 3.437s | 9.757s | 20.036s | 33.230s | ✅ |
A tweet from @claudeai saying: Our agreement with @SpaceX means we will use all the compute capacity at their Colossus 1 data center. This will give us over 300 megawatts of additional capacity to deploy within the month. 6:19 PM · May 6, 2026 · 3.4M Views
A screenshot showing multiple browser windows open with a terminal in the middle. The terminal has Helix open in a split panel, inside of Zellij.
A Twitter post by José Valim: At @ElixirConfEU, @chris_mccord and I set up two races between myself and a coding agent. One was to add a new feature to DBConnection (max_lifetime) and another was to investigate a potential type system regression (links below). For DBConnection, I gave the documentation of how the option would behave. Not only was I faster, the coding agent solution had 4x more LOC than mine (it changed files and data structures that were not necessary for the feature). For the type system, the coding agent could not fix the problem at all. It did recognise it was a type system regression, and it did provide an initial fix in roughly the same time as I did, but the coding agent fix was wrong. After I fed it additional tests (from my own solution), it failed to come up with a fix. I use coding agents daily, I often ship their code as is, but they can still slow me down, they are nowhere close to fixing all problems I tackle in a given week, and are often not up to the standards I expect from my own software. 11:40 AM · Apr 27, 2026 1,998 Views
A twitter Bun post saying: We do not currently plan to upstream this, as Zig has a strict ban on LLM-authored contributions.
Physical media is the answer to content overload: it has a thingness that the ephemera of online digital content does not. Its physical nature allows for artistic expression on and in the thing. The act of insertion and removal provides a psychological gate between "starting", "doing" and "finishing". It is the opposite of The Eternal Scroll. It is the hero we need.
Sony Recordable MD
A screenshot of the lubeno repository hosted on lubeno.dev containing the following changes: - Add lazy loading of diffs and UI improvements - Add line expand above and below inside diffs - Switch to refractor + One Light theme for syntax higlighting - Add diff syntax higlighting with Sugar High - Add total lines changed in diff and info about diff target - Add warning and confirmation before repository is transfered.
A screenshot of the https://gitlab.com/gitlab-org/gitlab/-/merge_requests website with a bunch of colourful tags making it look like a mess.
Screenshot of a news article saying: Jack Dorsey's Block lays off nearly half of workforce due to AI "A new way of working which fundamentally changes what it means to build and run a company."
A screenshot of the lubeno.dev repository page.
A screenshot quoting a paragraph from article https://steve-yegge.medium.com/the-future-of-coding-agents-e9451a84207c, saying: "Their stories flew by, but the theme was the same: They are going so fast that other teammates can’t keep up. They have a contributor in Munich who works in PST (our timezone), and he’ll say, “I did X!” And they’ll get mad and say, “Why did you do X, where did you get that information from?” And he’ll be like, it was from 2 hours ago. And they’ll say, “2 hours ago!? That’s ancient!” It might as well have been 2 weeks ago."
A screenshot showing a terminal window running the https://github.com/lunatic-solutions/chat app.
A screenshot of https://lubeno.dev
A screenshot showing a git repository being rendered in a tree like format on lubeno.dev.
A screenshot of a conversation with Claude.ai containing the following text with the words "which is what I'd recommend" underlined: 2. Read both streams concurrently (the threading approach we discussed earlier) 3. Use the fully async approach (tokio::process) which handles this automatically 4. Accept the risk - which is what I'd recommend since git error messages are practically never large enough to cause issues
A text in german saying: Guten tag, Nach den letzten Berechnungen Ihrer Krankenkasse haben wir festgestellt, dass Sie Anspruch auf eine Rückerstattung in der Höhe von CHF 389.00 haben. Wir laden Sie ein, die folgenden Schritte durchzulesen, indem Sie unten klicken
A screenshot of lubeno.dev, a forge for jj. Showing a list of commits.
A screenshot of the Hacker News announcement of the React.js library from May 29, 2013. With every comment being very negative about it.
A photo of the eurovision stage with wireframe mountains in the background.
A screenshot of a browser window that has "localhost:3000" open, showing a very simple UI listening a few git commits for a "test/test" repository.
Tony Stark laboratory with a futuristic UI. From the movie Iron Man.
A screenshot of a terminal running the `jj log` command. It shows 3 empty commits that serve as todos.
A stock chart comparing the Booking platform and AirBnB, where Booking is outperforming AirBnB by 150% percent from 2021 to 2025.
A graph of the declining Porsche stock price over 3 years.
A screenshot from finance.yahoo.com saying "Trump chooses TV doctor Oz to run Medicare, Medicaid".
A FT screenshot showing a graph of declining traffic for twitter in the US and UK, where the UK lost 30% of traffic in approximately 1 year and the US 20%.
A browser window just displaying a big X.
If you have the resources to spend some time exploring, if you are on to interesting threads of novelty that few other people have, and if you have the spirit to tighten your belt, throw out your map, and explore off-road, then your real job is to do so. It is a grave sin to neglect that kind of cosmic duty. But many more people have the means and privilege to quit their tracked careers than ever realize it and act on it. You need far less than you think to live in monk mode and pursue this kind of exploration. What this means in practice is that at some point far before you are or feel ready, you need to quit your job.