Tootfinder

Oh whew, finally a new #HTTP verb!
https://www.theregister.com/devops/2026/07/13/http-gets-a-query-method-so-comple…

HTTP gets a QUERY method so complex searches can stop pretending to be POST
New verb carries request content while remaining safe, idempotent, and cacheable

from my link log —
Designing a query system for a Rust compiler.
https://bal-e.org/speed/krabby/2026/query-system-design/
saved 2026-08-04 ht…

Meine Kinder fragen mich, warum ich noch linear fernsehe. Hier sind zwei gute Gründe:

1. Allein wäre ich nie auf die Idee gekommen, die siebenteilige Verfilmung von Jahrmarkt der Eitelkeiten in der #Mediathek abzurufen. Ich bin zufällig in die Ausstrahlung bei #One reingeraten und habe dann a…

from my link log —
Query-based compiler architectures.
https://ollef.github.io/blog/posts/query-based-compilers.html
saved 2020-06-25 ht…

⚡ The core is a single native tool, rlm_query, that spawns a child Pi process with the same tools. The child can call rlm_query too — so the agent recursively decomposes a problem, delegates chunks to copies of itself, and combines the results.
🎯 One-shot or interactive: run ypi "Refactor the error handling in this repo", or bunx ypi without installing. Available via bun, npm, curl or as the minimal pi-recursive extension for plain Pi.

Funny to see this article from 2024 getting some traffic today. There's supposed to be a solar eclipse, and as I wrote, you can expect a particular search query to trend because us humans are predictable in our stupidity.
https://coywolf.com/news/seo/people-se

People searching ‘eyes hurt after eclipse’ spike after solar eclipse
Google searches for "eyes hurt after eclipse" spiked after the April 8, 2024 solar eclipse, driven by fake glasses, eye strain, and unprotected viewing.

Shopware 6: "Log the old and new value when this field changes" sounds simple until you write it — a second query, race conditions, deleted rows with no old value left to fetch. Shopware computes the diff inside the write transaction itself. One footnote in the docs, and that's it.

Shopware 6 Hidden Gems #15: ChangeSet — before/after values for every write, no second query - Winkelwagen GmbH - an ecommerce agency
The requirement sounds harmless every time: "Log when a product price changes, with old and new value." Or: "When an order's email changes, notify the old

https://www.youtube.com/results?search_query=やよぴったん
このへんから好きな奴を

Er det en ny, ondsindet form for overvågning eller hvad??
https://ordnet.dk/ddo/ordbog?query=-skopi

from my link log —
The curious case of the table-locking UPDATE query on PostgreSQL.
https://blog.heroku.com/curious-case-table-locking-update-query
saved 2019-12-19

The Curious Case of the Table-Locking UPDATE Query
Check out how I debugged a long-running UPDATE database query modifying a single row, and why such a seemingly simple query caused so much harm.

This Tool Unmasks the Shadowy World of Ads that Track Your Location https://www.404media.co/this-tool-unmasks-the-shadowy-world-of-ads-that-track-your-location/

Bunny just made their DNS service free. Crazy good.
https://bunny.net/blog/were-making-bunny-dns-free/

Yay, I added my #yannix flake to search.nixos.org. You can now find my #nix / :nixos: #NixOS stuff on the official search page:
📦 my packages:

Long before anyone typed a query into Google,
Roger Summit built a system that let researchers search vast databases from a remote terminal in minutes.
His creation, Dialog, was the first commercial online search service,
and it laid groundwork the modern web would later inherit.
Summit has died at 95
…

Roger Summit, who built online search decades before Google, dies at 95
Long before anyone typed a query into Google, Roger Summit built a system that let researchers search vast databases from a remote terminal in minutes. His creation, Dialog, was the first commercial online search service, and it laid groundwork the modern web would later inherit. Summit has died at 95.

Gerade einmal geschaut, was sich in Sachen Digitalisierung der Werke Thomas Manns (seit diesem Jahr ja #gemeinfrei) getan hat. Die #ulb #düsseldorf hat sich um diese Sache nach Ausweis der …

from my link log —
Query cycles: A Rust compiler murder mystery.
https://ferrous-systems.com/blog/query-cycles-a-compiler-murder-mystery/
saved 2026-07-28

Query cycles: A compiler murder mystery
I HAVE NO TOOLS BECAUSE I HAVE DESTROYED MY TOOLS WITH MY TOOLS

Do you have a long running calculation freezing up your shiny app? {callr} or {crew} might help: https://discindo.org/post/asynchronous-execution-in-shiny/

🎯 Core systems already work: query planner, buffer cache, storage engine, B-tree indexes, wire protocol (psql connects), JSON/JSONB, window functions, foreign keys, EXPLAIN / EXPLAIN ANALYZE, regex, PL/pgSQL pieces, and standard SQL with transactions, aggregates and joins.

Shopwares Criteria::setTitle() is an awesome way to tag your queries throughout the infrastructure!
http://winkelwagen.de/2026/07/10/shopware-6-hidden-gems-1-criteriasettitle-name-tags-for-your-sql-queries/

Which Model Is Actually Serving You? IRIS: Budgeted Black-Box Auditing of Model Substitution and Routing Dilution in LLM Gateways
Yuewei Zhang, Zhi-Hai Zhang, Hanzhang Qin
https://arxiv.org/abs/2607.20860 https://arxiv.org/pdf/2607.20860 https://arxiv.org/html/2607.20860
arXiv:2607.20860v1 Announce Type: new
Abstract: Commercial LLM gateways mediate access to hosted models, but the served backend may not match the advertised one: it may substitute a cheaper model on every request or route only a fraction $\epsilon$ of requests to it. Prior black-box auditors often need a privileged signal (log-probabilities, token ranks, or reference samples) or a target-specific probe, fix the query budget in advance, and return a yes/no verdict. We present $\mathrm{IRIS}$, an audit that needs only the returned text: it asks endpoints to generate random numbers or strings, fingerprints the backend, and is the first to combine, in one text-only audit, detection of whole-stream substitution and fractional dilution, attribution of the served backend, routing-fraction ($\epsilon$) estimation, and a query budget it sizes itself. A cheap pilot fits the exponential query-error decay and freezes that budget before any suspect query is issued. On an intra-family Qwen3 ladder $\mathrm{IRIS}$ verifies the backend at $0.99$ AUROC and sharpens attribution as queries accumulate; across a commercial OpenRouter library it catches $\epsilon{=}0.3$ dilution on margin-qualified pairs at $0.85$ mean power ($0.017$ false-positive rate) and recovers $\epsilon$ to within $0.04$ for enrolled diluents; and a live cross-provider audit flags $14$ of $15$ same-model provider pairs by genuine quantization and kernel deviations, corroborated on third-party MET traces. Against comparable black-box auditors, $\mathrm{IRIS}$ matches or beats detection on shared tasks, and adaptive allocation lifts the matched-budget target-hit rate from $73$% to $87$%. Further experiments cover adversarial gateways, knob identifiability, unseen diluents, and false-positive control.
toXiv_bot_toot

#PSA : The bots here are adopting the same patterns that they use elsewhere.
Be very dubious of “firstlast##@mastodon.social“ accounts with tone discrepancies. Like responding to a snarky gripe by asking a verbosely faux-deep query about a tangent.
Because they are LLM training bots.

🔧 Three pieces mirror the Python RLM design: a system prompt, a $CONTEXT file plus bash as the REPL, and the native rlm_query sub-call. Depth 0 and 1 get full tools; the leaf at max depth runs bash only. jj workspaces isolate child edits when available.
🛡️ Five guardrails guarantee termination: dollar budget, wall-clock timeout, call limit, depth limit and PATH scrubbing. Deeper agents are told to be more conservative, preferring direct action over spawning more children.

RE: #Berlin|er

Generative AI’s per-query footprint can be small compared to flying, eating meat, or even Zoom. Until agents show up and compute goes through the roof.
How I updated my AI footprint calculator accordingly https://jonippolito.substack.com/p/agents-are-where-ais-energy-bil…

if only #mastodon would support *any* kind of shitty search for posts scrolling by... but even when i know to query "from:@… <term>" i get 0 posts, but 3 completely irrelevant profiles, for any term, even terms that i see in my feed which are in the toots by this …

Did I just hack up an abomination using xmllint (from libxml2) and jq to query the latest #PyPI version of a package? Perhaps.
Do you no longer have to specify the target version when using pkgbump on PyPI packages? Absolutely.
#Gentoo

Tomorrow! Join Grist Labs co-founder Stan for a demo of Grist's new MCP server.
We'll start small and then show off a complex workflow that leverages multiple integrations, so there's something for everyone.
Connect Claude, ChatGPT, Gemini, or your own local models to your docs over the Model Context Protocol, then list and query tables, read and write rows, and build schema, all via OAuth with scoped access.
Register:

from my link log —
Querying PostgreSQL tables from DuckDB.
https://duckdb.org/2022/09/30/postgres-scanner.html
saved 2022-09-30 https://

Querying Postgres Tables Directly from DuckDB
DuckDB can now directly query tables stored in PostgreSQL and speed up complex analytical queries without duplicating data.

Polars is a lightning fast DataFrame library/in-memory query engine with parallel execution and cache efficiency. And now you can use is with the tidyverse syntax: #rstats

More Efficient Tidyverse Code, Using Polars in the Background
Polars is a cross-language tool for manipulating very large data. However, one drawback is that the R implementation has a syntax that will look odd to many R users who are not used to Python syntax. The objective of tidypolars is to improve the ease-of-use of Polars in R by providing tidyverse syntax to polars.

⚡ Instead of a flat text file, every command lands in a queryable database.
atuin search --exit 0 --after "yesterday 3pm" make
Filter by success, time, directory or session — the kind of query a plain .bash_history can never answer.

⚡ 50x more token-efficient than raw transcript search: ~917 tokens per query vs ~45,734 for raw search. It structures history into sessions, events & indexed fields, returning ranked, cited matches instead of dumping raw text

Tomorrow! Join Grist Labs co-founder Stan for a demo of Grist's new MCP server.
We'll start small and then show off a complex workflow that leverages multiple integrations, so there's something for everyone.
Connect Claude, ChatGPT, Gemini, or your own local models to your docs over the Model Context Protocol, then list and query tables, read and write rows, and build schema, all via OAuth with scoped access.
Register:

💨 Response time of 134ms versus 1,440ms for Mem0, with retrieval running inside the voice turn instead of after it
💰 430 memory tokens per query versus 6,956 for Mem0 and 1,899 for EverMemOS: queries route first, rank second, then inject only Top-K into the context

📋 #queuesql is an #opensource #Laravel package that turns any write query — delete, update, insert, upsert — into parallel batched queue jobs

Uh, I do NOT want this. I do not want this AT ALL. No option to opt out on Tahoe 26.6.2.
"Local files support for Excel Agent.
Added support for Local files for Excel Agent in modern Excel workbooks stored on your device, giving you the flexibility to query Copilot Chat no matter where your files are located."

🧩 Service coverage spans application services, data services, eventing, identity, infrastructure, billing and container-backed workloads — including #DynamoDB with GSI, LSI, Query, Scan, TTL and transactions, EventBridge, Step Functions, Athena, Glue, EMR, Firehose, #Bedrock Runtime and S3 Ve…

🔍 Tracing — per-request & per-job waterfalls of every query, cache op & HTTP call, off by default & persisted only when slow, errored or sampled to stay cheap at scale
🐞 Unified error tracking — web, queue, command & browser exceptions fingerprinted into a grouped inbox with stacktraces, sparklines & assign/ack/mute/resolve

RE: #science #smithsonian #censorship #openScience #sciOp #p2p #torrent #museum

📈 Full observability suite — Core Web Vitals (RUM), SLOs with burn-rate alerts, incident tracking with MTTR, release health with regression guard & custom business metrics
🤖 First-class #MCP server — let Claude Code, #Cursor or Copilot query live errors, APM, traces, logs, SLOs & rele…