frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Zenclora OS

https://zenclora.org/
17•debo_•2h ago

Comments

OsrsNeedsf2P•1h ago
Haven't investigated that much so there might be some big catch I'm missing, but I really wish people would stop making custom OSes and just help out an existing project
nish__•1h ago
What do they need help with?
andy_ppp•1h ago
I wish everyone thought exactly like me too, it’d be great getting exactly what I wanted and basically the world would become utopia over night.
autoexec•46m ago
In the worst cases you get people who just want to say they made an OS by slapping their name on an existing distro after changing the default background image or making a couple tweaks. That's a lot easier to than to contribute something meaningful to an existing project.

More charitably, it's faster and a lot less complicated to modify a distro to your liking than trying to get a major distribution to cater to your whims and philosophy and making your version avilable to others gives them the option of easily installing your new features/modifications as is or using what parts they want. That's the joy of of FOSS. Even when the changes are modest, if they are shared and even a little useful then someone else can incorporate them or build off of them.

waterTanuki•1h ago
Debian isn't known for being bloated so I don't get what the selling point here is. If I want a stripped down barebones system why not just use Arch?
autoexec•1h ago
"Optimized & Light" but also the system requirements say it needs 10 GB for the just the system itself with 25 GB+ recommended. Regular old Debian recommends 10 GB of disk space too (but allows for installations that come in under 1 GB if you really want light)

What kind of bloat did they get rid of to reach the same footprint as the OS they started from?

Retr0id•1h ago
I do not understand the purpose of the `zen` tool, which is a giant bash script with hardcoded install/uninstall steps for each "package": https://github.com/Zenclora/build/blob/697c20a79b5638b4447bf...
josephg•1h ago
Despite their marketing, this script does not spark joy.

Eg, they have a "zen optimize" command, which clears caches, nukes /tmp, turns swap off and on again and runs fstrim. You don't need to do any of this stuff. Clearing cache and swap will make programs run slower. Wiping /tmp can be harmful to running processes.

It all seems like a big LLM vomit.

https://github.com/Zenclora/build/blob/697c20a79b5638b4447bf...

autoexec•1h ago
Generally, clearing caches will make programs run slower once (on first run after the clear), but it could speed things up if there's a massive amount of other stuff cached or if the cached thing you're trying to run got corrupted somehow.
Retr0id•56m ago
There is no situation in which having a populated page cache will slow things down, since those are the pages that linux will drop first to use for whatever else needs them.
groguzt•58m ago
I mean, the comments with numbered steps are pretty much a dead giveaway that its just AI slip.

If I were to guess, I would say Gemini 3/3.1, as it tends to add a lot of numbered steps comments, more so than other models from my experience

pneumic•1h ago
Putting money down on this being primarily LLM coded
autoexec•1h ago
I think it's just supposed to make installing commonly used things seem easier.

I have to admit that (assuming they both work flawlessly) it'd be easier to type "zen install steam" than it is to manually edit a file and run four commands following the directions here https://wiki.debian.org/Steam

owenpalmer•1h ago
It's not an OS, it's a distro
PufPufPuf•7m ago
We're getting vibecoded Linux distros now?

A Model Registry

https://github.com/imsingee/aidy-models
1•singee•2m ago•1 comments

ModelSweep: Open-Source Benchmarking for Local LLMs

https://github.com/leonickson1/ModelSweep
1•leonickson•2m ago•0 comments

Ghosts of Softmax: Zeros of the partition function explain training instability

https://github.com/piyush314/ghosts-of-softmax
1•g4omingron•4m ago•0 comments

Reliance signs $3B deal with Samsung C&T for 15-year

https://www.threads.com/@optifin_advisors/post/DV-LzVqDNwh
1•optifinadvisors•8m ago•0 comments

Learn to build electronics with Oyvind's Circuit Tips

https://www.build-electronic-circuits.com/
1•teleforce•8m ago•0 comments

Is this "the Raku community" in 2026?

https://gist.github.com/2colours/5262cb60987a4a13760b04c1d7cd234b
1•2colours•12m ago•1 comments

Corporate but Winged: Cicikuş v3 Is Now Available

https://huggingface.co/pthinc/Cicikus-v3-1.4B-Opus4.6-Powered
1•pthuser•14m ago•1 comments

Pyodide and Paribury Exchange for in-browser data workflows

1•parburys•18m ago•0 comments

Why some metals turn transparent: A key mechanism for touchscreen conductors

https://phys.org/news/2026-03-metals-transparent-key-mechanism-touchscreen.html
1•westurner•23m ago•1 comments

The Repository of Ill-Advised Ventures

https://www.riav.org/
1•robbieo•24m ago•0 comments

Llmgate – call any LLM via YAML config, 2 dependencies

https://github.com/kesiee/llmgate
1•kesiees•25m ago•0 comments

Control a Waymo by Auto Rerouting

https://waymo.fraud.llc/
2•fearthetelomere•33m ago•0 comments

Zelph is a logic programming system, but not in the traditional sense

https://acrion.github.io/zelph/logic/
1•ta988•33m ago•0 comments

Bill Ackman vs. Carl Icahn Showdown on Herbalife [video]

https://www.youtube.com/watch?v=6QWZbxeJd6g
1•marysminefnuf•33m ago•0 comments

Ad-homineLLM: Are you wrong because you used an LLM?

https://blog.devrupt.io/posts/ad-hominellm/
1•ddtaylor•35m ago•0 comments

Show HN: Simple CBM calculator for shipping and freight

https://cbmcalc.com
2•distartin•41m ago•0 comments

Show HN: DataFlow,Turn raw data into high-quality LLM training datasets

https://github.com/OpenDCAI/DataFlow
1•Junnn•41m ago•0 comments

Flash-KMeans: Fast and Memory-Efficient Exact K-Means

https://arxiv.org/abs/2603.09229
1•matt_d•42m ago•0 comments

Kali Linux MCP Server

https://github.com/cyberillo/kali-mcp-server
1•cyberillo•42m ago•1 comments

Surprising a Model

https://vivekhaldar.com/articles/surprising-a-model/
1•gandalfgeek•43m ago•0 comments

PMetal – (Powdered Metal) LLM Fine-Tuning Framework for Apple Silicon

https://github.com/Epistates/pmetal
1•epistates•44m ago•1 comments

The Beginner's Guide to Dimensionality Reduction (2018)

https://dimensionality-reduction-293e465c2a3443e8941b016d.vercel.app
1•vismit2000•46m ago•0 comments

Manifesto on Symbiosis: A New Paradigm for Civilization Part IV

1•acehs•51m ago•0 comments

Loom – a zero-dependency C++20 blog engine (one binary, one command)

https://github.com/1ay1/loom
1•tfeayush•51m ago•0 comments

Coding Agents for Data Analysis

https://simonwillison.net/2026/Mar/16/coding-agents-for-data-analysis/
1•samuel246•52m ago•0 comments

Manifesto on Symbiosis: A New Paradigm for Civilization Part III

2•acehs•53m ago•0 comments

Always 'Copy Clean Link' When Possible on Firefox, with UserChrome.css

https://joshua.hu/firefox-always-copy-clean-link-url-userchrome-css
3•mmsc•57m ago•0 comments

A Day in the Life of an Enshittificator [video]

https://www.youtube.com/watch?v=T4Upf_B9RLQ
1•celsoazevedo•1h ago•0 comments

TypeScript's rise in the AI era: Insights from Anders Hejlsberg (2025)

https://github.blog/developer-skills/programming-languages-and-frameworks/typescripts-rise-in-the...
2•pramodbiligiri•1h ago•0 comments

Donut Lab Battery Test 4 Changes Everything [video]

https://www.youtube.com/watch?v=5cpYeT4VmSY
1•fencedload•1h ago•0 comments