frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Reputation Scores for GitHub Accounts

https://shkspr.mobi/blog/2026/02/reputation-scores-for-github-accounts/
1•edent•1m ago•0 comments

A BSOD for All Seasons – Send Bad News via a Kernel Panic

https://bsod-fas.pages.dev/
1•keepamovin•4m ago•0 comments

Show HN: I got tired of copy-pasting between Claude windows, so I built Orcha

https://orcha.nl
1•buildingwdavid•4m ago•0 comments

Omarchy First Impressions

https://brianlovin.com/writing/omarchy-first-impressions-CEEstJk
1•tosh•9m ago•0 comments

Reinforcement Learning from Human Feedback

https://arxiv.org/abs/2504.12501
2•onurkanbkrc•10m ago•0 comments

Show HN: Versor – The "Unbending" Paradigm for Geometric Deep Learning

https://github.com/Concode0/Versor
1•concode0•11m ago•1 comments

Show HN: HypothesisHub – An open API where AI agents collaborate on medical res

https://medresearch-ai.org/hypotheses-hub/
1•panossk•14m ago•0 comments

Big Tech vs. OpenClaw

https://www.jakequist.com/thoughts/big-tech-vs-openclaw/
1•headalgorithm•17m ago•0 comments

Anofox Forecast

https://anofox.com/docs/forecast/
1•marklit•17m ago•0 comments

Ask HN: How do you figure out where data lives across 100 microservices?

1•doodledood•17m ago•0 comments

Motus: A Unified Latent Action World Model

https://arxiv.org/abs/2512.13030
1•mnming•17m ago•0 comments

Rotten Tomatoes Desperately Claims 'Impossible' Rating for 'Melania' Is Real

https://www.thedailybeast.com/obsessed/rotten-tomatoes-desperately-claims-impossible-rating-for-m...
3•juujian•19m ago•2 comments

The protein denitrosylase SCoR2 regulates lipogenesis and fat storage [pdf]

https://www.science.org/doi/10.1126/scisignal.adv0660
1•thunderbong•21m ago•0 comments

Los Alamos Primer

https://blog.szczepan.org/blog/los-alamos-primer/
1•alkyon•23m ago•0 comments

NewASM Virtual Machine

https://github.com/bracesoftware/newasm
2•DEntisT_•25m ago•0 comments

Terminal-Bench 2.0 Leaderboard

https://www.tbench.ai/leaderboard/terminal-bench/2.0
2•tosh•26m ago•0 comments

I vibe coded a BBS bank with a real working ledger

https://mini-ledger.exe.xyz/
1•simonvc•26m ago•1 comments

The Path to Mojo 1.0

https://www.modular.com/blog/the-path-to-mojo-1-0
1•tosh•29m ago•0 comments

Show HN: I'm 75, building an OSS Virtual Protest Protocol for digital activism

https://github.com/voice-of-japan/Virtual-Protest-Protocol/blob/main/README.md
5•sakanakana00•32m ago•1 comments

Show HN: I built Divvy to split restaurant bills from a photo

https://divvyai.app/
3•pieterdy•34m ago•0 comments

Hot Reloading in Rust? Subsecond and Dioxus to the Rescue

https://codethoughts.io/posts/2026-02-07-rust-hot-reloading/
3•Tehnix•35m ago•1 comments

Skim – vibe review your PRs

https://github.com/Haizzz/skim
2•haizzz•36m ago•1 comments

Show HN: Open-source AI assistant for interview reasoning

https://github.com/evinjohnn/natively-cluely-ai-assistant
4•Nive11•37m ago•6 comments

Tech Edge: A Living Playbook for America's Technology Long Game

https://csis-website-prod.s3.amazonaws.com/s3fs-public/2026-01/260120_EST_Tech_Edge_0.pdf?Version...
2•hunglee2•40m ago•0 comments

Golden Cross vs. Death Cross: Crypto Trading Guide

https://chartscout.io/golden-cross-vs-death-cross-crypto-trading-guide
3•chartscout•43m ago•1 comments

Hoot: Scheme on WebAssembly

https://www.spritely.institute/hoot/
3•AlexeyBrin•46m ago•0 comments

What the longevity experts don't tell you

https://machielreyneke.com/blog/longevity-lessons/
2•machielrey•47m ago•1 comments

Monzo wrongly denied refunds to fraud and scam victims

https://www.theguardian.com/money/2026/feb/07/monzo-natwest-hsbc-refunds-fraud-scam-fos-ombudsman
3•tablets•52m ago•1 comments

They were drawn to Korea with dreams of K-pop stardom – but then let down

https://www.bbc.com/news/articles/cvgnq9rwyqno
2•breve•54m ago•0 comments

Show HN: AI-Powered Merchant Intelligence

https://nodee.co
1•jjkirsch•56m ago•0 comments
Open in hackernews

Tell HN: Perplexity Has Unspecified Character Limits for Session Export

3•eth0up•1mo ago
Hello all,

I discovered, the hard way, that exporting Perplexity sessions to PDF results in substantial content loss when the page is ~90 pages.

After opening a ticket on the matter, a brief dialogue with a rep proved unhelpful and confusing. It was stated that the Export as PDF feature only exports individual "threads", and that to export an entire session, each so-called thread must be individually selected and exported. This is simply wrong.

In practice, there is no method to select threads through the Top-Right/ 3-dot menu/Export as PDF option. Testing this with various sessions from 1 to 170 page exports showed no indication that threads were relevant.

Exports under 90 pages tend* to retain all content, while a 93-page didn't, but a 95 and 170 page export did. This indicates that the character limit (if that's the cause) is variable, as 170 pages is almost guaranteed to contain more characters than 90.

The fundamental point here, whatever the cause, is that data loss is inevitable under the present UI with its absence of documentation, notices etc.

*I observed changes after submitting the ticket and modifications have already been made. The situation was worse before, and now less worse, but still applies.

Comments

eth0up•1mo ago
It should also be noted that no simple option exists within the Perplexity UI to export an entire session. Select-All/Copy/Paste captures only single or small groups. To test this: generate a session with a dozen or so pages (or open an extant session) and try to copy the content and paste to txt file. You'll see it doesn't work.

This does not apply to Claude, Gemini, Grok or GPT, all of which presently permit Select All

Edit: I also think it's worth discussing the actual advantages of using Perplexity over a dedicated model, eg Claude (my personal choice). Obviously, to offer so many models as Perplexity does, and for the same dollar amount, quality must be pruned somewhere. I have yet to purchase a subscription for a single model provider, but I am going to confidently guess that the session degradation, context preservation, memory and general coherence would be superior in a dedicated model. I'm not terribly informed on LLM details, but surmise some token austerity is involved in Perplexity's perplexing generosity. But I think it's safe enough to say, they must cut costs where they can, and the session export bug may rather be a feature of austerity, which I'm pretty sure is among more than a few more features disguised as bugs.

And yes, I'm mildly bitter. I've lost many hours of work due to being foolish enough to believe Export as PDF means export as pdf. A smart person doesn't believe what they read