frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

New protein therapy shows promise as antidote for carbon monoxide poisoning

https://www.medschool.umaryland.edu/news/2025/new-protein-therapy-shows-promise-as-first-ever-antidote-for-carbon-monoxide-poisoning.html
118•breve•3h ago•26 comments

NSF and Nvidia award Ai2 $152M to support building an open AI ecosystem

https://allenai.org/blog/nsf-nvidia
75•_delirium•2h ago•33 comments

Why LLMs Can't Build Software

https://zed.dev/blog/why-llms-cant-build-software
90•srid•2h ago•43 comments

Statement Regarding Misleading Media Reports

https://www.kodak.com/en/company/blog-post/statement-regarding-misleading-media-reports/
23•whicks•37m ago•3 comments

What's the strongest AI model you can train on a laptop in five minutes?

https://www.seangoedecke.com/model-on-a-mbp/
287•ingve•2d ago•103 comments

Launch HN: Cyberdesk (YC S25) – Automate Windows legacy desktop apps

7•mahmoud-almadi•21m ago•1 comments

Is chain-of-thought AI reasoning a mirage?

https://www.seangoedecke.com/real-reasoning/
25•ingve•1h ago•15 comments

Arch shares its wiki strategy with Debian

https://lwn.net/SubscriberLink/1032604/73596e0c3ed1945a/
233•lemper•6h ago•82 comments

Jujutsu and Radicle

https://radicle.xyz/2025/08/14/jujutsu-with-radicle
31•vinnyhaps•1h ago•6 comments

Org-social is a decentralized social network that runs on an Org Mode

https://github.com/tanrax/org-social
115•todsacerdoti•4h ago•21 comments

Brilliant illustrations bring this 1976 Soviet edition of 'The Hobbit' to life (2015)

https://mashable.com/archive/soviet-hobbit
125•us-merul•3d ago•43 comments

Blood Oxygen Monitoring Returning to Apple Watch in the US

https://www.apple.com/newsroom/2025/08/an-update-on-blood-oxygen-for-apple-watch-in-the-us/
29•thm•2h ago•4 comments

Mbodi AI (YC X25) Is Hiring a Founding Research Engineer (Robotics)

https://www.ycombinator.com/companies/mbodi-ai/jobs/ftTsxcl-founding-research-engineer
1•chitianhao•3h ago

Passion over Profits

https://dillonshook.com/passion-over-profits/
33•dillonshook•2h ago•22 comments

SIMD Binary Heap Operations

http://0x80.pl/notesen/2025-01-18-simd-heap.html
20•ryandotsmith•2d ago•1 comments

Meta accessed women's health data from Flo app without consent, says court

https://www.malwarebytes.com/blog/news/2025/08/meta-accessed-womens-health-data-from-flo-app-without-consent-says-court
216•amarcheschi•4h ago•123 comments

Ask HN: How do you tune your personality to get better at interviews?

11•tombert•31m ago•16 comments

Linux Address Space Isolation Revived After Lowering 70% Performance Hit to 13%

https://www.phoronix.com/news/Linux-ASI-Lower-Overhead
100•teleforce•3h ago•25 comments

Funding Open Source like public infrastructure

https://dri.es/funding-open-source-like-public-infrastructure
169•pabs3•12h ago•81 comments

A new poverty line shifted the World Bank's poverty data. What changed and why?

https://ourworldindata.org/new-international-poverty-line-3-dollars-per-day
34•alphabetatango•3d ago•23 comments

Zenobia Pay – A mission to build an alternative to high-fee card networks

https://zenobiapay.com/blog/open-source-payments
201•pranay01•13h ago•211 comments

Great Myths #16: The Conflict Thesis

https://historyforatheists.com/2025/08/the-great-myths-16-the-conflict-between-science-and-religion/
7•stone-on-stone•2d ago•1 comments

Meta's flirty AI chatbot invited a retiree to New York

https://www.reuters.com/investigates/special-report/meta-ai-chatbot-death/
28•edent•53m ago•13 comments

Show HN: Yet another memory system for LLMs

https://github.com/trvon/yams
128•blackmanta•12h ago•33 comments

PYX: The next step in Python packaging

https://astral.sh/blog/introducing-pyx
698•the_mitsuhiko•21h ago•424 comments

"None of These Books Are Obscene": Judge Strikes Down Much of FL's Book Ban Bill

https://bookriot.com/penguin-random-house-florida-lawsuit/
191•healsdata•2h ago•176 comments

OCaml as my primary language

https://xvw.lol/en/articles/why-ocaml.html
352•nukifw•21h ago•251 comments

What Medieval People Got Right About Learning (2019)

https://www.scotthyoung.com/blog/2019/06/07/apprenticeships/
130•ripe•15h ago•77 comments

iPhone DevOps

https://clearsky.dev/blog/iphone-devops-ssh/
118•ustad•6h ago•91 comments

Kodak says it might have to cease operations

https://www.cnn.com/2025/08/12/business/kodak-survival-warning
297•mastry•2d ago•204 comments
Open in hackernews

Show HN: Yet another memory system for LLMs

https://github.com/trvon/yams
128•blackmanta•12h ago
Built this for my LLM workflows - needed searchable, persistent memory that wouldn't blow up storage costs. I also wanted to use it locally for my research. It's a content-addressed storage system with block-level deduplication (saves 30-40% on typical codebases). I have integrated the CLI tool into most of my workflows in Zed, Claude Code, and Cursor, and I provide the prompt I'm currently using in the repo.

The project is in C++ and the build system is rough around the edges but is tested on macOS and Ubuntu 24.04.

Comments

winterrx•11h ago
The domain listed on the GitHub repo redirects too many times.
blackmanta•11h ago
That should be fixed now. It was a misconfiguration of CloudFlare SSL with GitHub Pages.
mempko•10h ago
Wicked cool. Useful for single users. Any plans to build support for multiple users? Would be useful for an LLM project that requires per user sandboxing.
marcofiocco•10h ago
What about versioning of files?
blackmanta•10h ago
The tool has built-in versioning. Each file gets a unique SHA-256 hash on storage (automatic versioning), you can update metadata to track version info, and use collections/snapshots to group versions together. I have been using the metadata to track progress and link code snippets.
yawerali•10h ago
Hader
JSR_FDED•9h ago
Thanks, I learned a lot from this.
sitkack•9h ago
How would you use the built in functionality to enable graph functionality? Metadata or another document used as the link or collection of links?
blackmanta•1h ago
The graph functionality is exposed through the retrieval functionality. I may improve this later but the idea was to maximize getting the best results when looking for stored data.
retreatguru•9h ago
How do you use this in your workflow? Please give some examples because it’s not clear to me what this is for.
ActorNightly•9h ago
>MCP server (requires Boost)

I see stuff like this, and I really have to wonder if people just write software with bloat for the sake of using a particular library.

pessimizer•8h ago
Boost is a nearly 30 year old open source library that provides stuff for C++ that most standard libraries for other languages already have out of the box. You seem to think that it is hipster bullshit rather than almost a dinosaur itself.
SJC_Hacker•8h ago
Blame the committee for refusing to include basic functionality like regular expressions , networking and threads as part of the STL
menaerus•7h ago
The reason for depending on Boost in this repo is just few search characters away - he needs HTTP/WebSocket implementation and Boost.Beast provides it. The actual bloat here in this repo is conan.
noodletheworld•7h ago
? Are you complaining about MCP or boost?

It’s an optional component.

What do you want the OP to do?

MCP may not be strictly necessary but it’s straight in line with the intent of the library.

Are you going to take shots at llama.cpp for having an http server and a template library next?

Come on. This uses conan, it has a decent cmake file. The code is ok.

This is pretty good work. Dont be a dick. (Yeah, ill eat the down votes, it deserves to be said)

airstrike•3h ago
This feels like a shallow dismissal, which is frowned upon per the HN guidelines
yard2010•8h ago
I'm puzzled - where are the header files?
paffdragon•6h ago
You mean these? https://github.com/trvon/yams/tree/main/include/yams
vira28•8h ago
How does this compare to Letta?
rkunnamp•6h ago
Thank you for sharing this. Sorry for a possible noob question. How are embedding generated? Does it use a hosted embedding model? (I was trying to understand how is semantic search implemented)
sync•5h ago
It, uh... generates mock embeddings? https://github.com/trvon/yams/blob/c89798d6d2de89caacdbe50d2...

(seems like there's some vague future plans for models like all-MiniLM-L6-v2, all-mpnet-base-v2)

pbronez•2h ago
Hmm I wonder how much that effects the compression benefits of block level duplication. The mock embeddings choose vector elements from a normal distribution, so it’s far from uniform
huqedato•4h ago
In my RAG I use qdrant w/ Redis. Very successfully. I don't really see the use of "another memory system for LLM", perhaps I'm missing something.
jerpint•4h ago
I also developed yet another memory system !

https://github.com/jerpint/context-llemur

Although I developed it explicitly without search, and catered it to the latest agents which are all really good at searching and reading files. Instead you and LLMs cater your context to be easily searchable (folders and files). It’s meant for dev workflows (i.e a projects context, a user context)

I made a video showing how easy it is to pull in context to whatever IDE/desktop app/CLI tool you use

https://m.youtube.com/watch?v=DgqlUpnC3uw

elpocko•3h ago
>block-level deduplication (saves 30-40% on typical codebases)

How is savings of 40% on a typical codebase possible with block-level deduplication? What kind of blocks are you talking about? Blocks as in the filesystem?

blackmanta•1h ago
I am working to improve the CLI tools to make getting this information easier but I have stored the yam repo in yams with multiple snapshots and metadata tags and I am seeing about 32% storage savings.
elpocko•44m ago
Cool. I have no idea what "stored the yam repo in yams" means. What do you mean by "block-level deduplication"? What is a block?
blackmanta•35m ago
I stored the codebase for yams in the tool. The "blocks" are content-defined blocks/chunks, not filesystem blocks. They're variable-size chunks (typically 4-64KB) created using Rabin fingerprinting to find natural content boundaries. This enables deduplication across files that share similar content.
A4ET8a8uTh0_v2•3h ago
I like it and I will be perusing your code for what could be used in my 'not yet working' variant.
skyzouwdev•2h ago
That sounds like a practical take on LLM memory — especially the block-level deduplication part.

Most “memory” layers I’ve seen for AI are either overly complex or end up ballooning storage costs over time, so a content-addressed approach makes a lot of sense.

Also curious — have you benchmarked retrieval speed compared to more traditional vector DB setups? That could be a big selling point for devs running local research workflow

blackmanta•57m ago
I have not, but that is something I plan to do when I have time.
izabera•2h ago
not trying to be a hater but how is 100mb/s high performance in 2025? that's as performant as a 20 years old hdd
blackmanta•1h ago
The system is honestly tuned for storage efficiency not speed but these configurations are tunable and you can use the benchmarks as a reference for tuning. https://github.com/trvon/yams/blob/main/docs/benchmarks/perf...