Tootfinder

AI for writing documentation | The FreeBSD Forums
<https://forums.freebsd.org/threads/ai-for-writing-documentation.101700/>
@…

I started using SearXNG recently, which is a metasearch engine you can install locally to keep all your search data away from search engine companies.
https://docs.searxng.org/
There are also a bunch of public instances you can try out.

Mintlify, which uses AI to help companies generate software documentation, raised a $45M Series B led by a16z and Salesforce Ventures at a $500M valuation (Rashi Shrivastava/Forbes)
https://www.forbes.com/sites/the-prompt/20

Anthropic Uses This Startup’s AI To Answer People’s Questions About Claude Code
It’s just one of 20,000 companies that use Mintlify’s AI system to generate and maintain FAQs and user guides about their software products.

@… If you are ever get bored and want to understand why the lack of dependency management is no biggie, I wrote a post some 15 years ago about this on my then blog, hosted at Opera (back when I worked there).
When it was removed (because Opera shutdown personal blogs) the Slackware documentation site rehosted it. I see it is actually still here and while not fully…

"Good documentation is a tell-tale sign of a great product and a company that puts users first. There exist good products with bad or no documentation, but there are very few poor products with great documentation."
https://daringfireball.net/linked/2026/05/12/kagi-snaps

Kagi Snaps
Link to: https://help.kagi.com/kagi/features/snaps.html

So, spent most of the day getting Handbrake compiled with the `svt-av1-psyex` and `svt-av1-essential` forks of the SvtAv1 encoder. However, I did win that battle and both are now available in the `sisyphus-client` containers. Definitely a pain in the ass, but it's nice to have options depending on what you want to use. Also updated the documentation to make sure I don't forget which is which.
I still need to implement some pipelines to build these, but holy crap the Handbrak…

Just went through a code file and replaced "target" with "recipient" in a bunch of docs and variable names, because I had had two senses of "target" and because I want to encourage less-combat-focused thinking among potential devs (which at the moment and for the foreseeable future is just me). Zero changes to code functionality (unless I messed up) and took maybe 30 minutes. Got me thinking about the goal of software development and who would view that time as "wasted".
From a short-sighted "measure user-facing functionality" standpoint there was basically zero progress made. From a "cultivating-code-understanding-entities" perspective, there are both immediate and long-term gains. I got a chance to review the entire file at a high level which improves my understanding of things even if no changes had been made. The changes increase clarity of a lot of documentation, as well as variable naming clarity in code which deals with both senses of "target" which now has clear separation between variables that refer to targets and recipients. If there are any other devs in the future, this will help them a lot. Even if not, by increasing my own development efficacy, the expected values for project-gets-shipped and overall-project-impact just went up. I've also helped myself not introduce the same ambiguity in all my future projects, because I'll remember this time investment, and I've bent the ideological flavor of the protect towards neutral/beneficial interactions between "initiators and recipients" instead of hostile interactions between "initiators and targets".
Overall, well worth the time investment if we consider all the subtler benefits, and not just a very crude measure of user-facing behavior. (Secretly, this is also a post about LLMs.)
#programming

Last month, I became aware of #GitHub #agentic #workflows. I succeeded in creating one, to extract structured data out of release notes, something that was done before manually.
Here are my

The Synthetic Exercise World Format provides fictional countries, companies, sectors, and threat actors with structured metadata for neutral CTI examples, exercises, interoperability tests, and standards documentation without referencing real-world sensitive entities.
I just released version 1.0.
#cti #opensource

My changes to pkgbase documentation in the FreeBSD wiki prior to 19th August 2025:
<https://wiki.freebsd.org/action/info/PkgBase?action=info>
The more recent page (fourth edition, outdated):
<

Does VSCode (eg in pylance) have a feature to display a clickable link to the language standard library documentation (eg #python #vscode #askfedi

Seen Multivariate Adaptive Regression Splines (MARS) in some of the tidymodels documentation and in Kuhn's books, but never really looked into them. As a model, I don't find them particularly interesting, but I'll say this, the {earth} vignettes are some of the best I've read. (not linked on CRAN page, but included w/pkg and at Milborrow's site)

A survey of 1,692 US physicians: over 80% use AI professionally and the most common use cases are medical research summarization and clinical care documentation (American Medical Association)
https://www.ama-assn.org/press-center/ama-

AMA: AI usage among doctors doubles as confidence in technology grows
Cautious optimism remains about the promise of AI.

That T2 should spend some time on documentation; the first dialog box asks for a superuser pwd and it isn't supplied anywhere. Quick install documentation is from 2017. And, btw, over the last couple of years I've regularly had issues with wifi connections at install in KDE (hello Kwallet). Am I the only one? Obviously!

from my link log —
ACME CA comparison.
https://poshac.me/docs/v4/Guides/ACME-CA-Comparison/
saved 2026-05-09 https://dotat.…

ACME CA Comparison - Posh-ACME
Documentation for the Posh-ACME PowerShell module

{dtrack} makes documentation of data wrangling part of the analysis and creates pretty flow charts: #rstats

Documentation continues. Done with the `queue` and `worker` endpoints, probably about 50% of the way done with this. I still need a good way to package all of this up with the binary which means putting the `openapi.yaml` file somewhere on the net so it can be pulled if need be. I'll deal with that later...lol
#openapi

The Mastodon documentation... which I'm sure most people trying to fix their instance have read

A few years ago, I gave a talk at RISC-V on #documentation, with some hints on automatically generating it. They are doing this now in practice for #Baochip:

#PostgreSQL running in the browser with an S3 storage backend - that's pretty impressive
https://www.zerofs.net/postgresql-in-the-browser

@… Es gibt eine Schriftart, die Barcodes erzeugt. https://graphicore.github.io/librebarcode/documentation/code128.html
Man könnte den Barc…

This has got to be one of Peter Girnus’ best pieces:
https://xcancel.com/gothburz/status/2052011313249890375?s=46

boss: we want to train a ServiceBot on our #documentation
me: here you go <gives #PDFs>
boss: ServiceBot says these PDFs are too big
me: isn't parsing vast amounts of #data to answer questions and dist…

Thanks to a work-in-progress Homebrew cask by Andi Péter (https://codeberg.org/GramEditor/gram/issues/15#issuecomment-11159294), I was just able to quickly install and play with the new Gram code editor (

📖 Automatic API documentation:
PostgREST generates #OpenAPI specs automatically from your schema. Hook up Swagger-UI − interactive, always up-to-date docs for your API. No more manual doc maintenance.
🏗️ Versioning through DB schemas:
Instead of managing API versions in code, PostgREST uses #PostgreSQL

Have You Ever Thought About Drones in MISP?
To better support the documentation and analysis of drone-related incidents, several new resources have been integrated into MISP.
#drone #drones #intelligence

The experimental HTML export of @… still lacks a lot of features, but already has enormous potential IMO.
I've been able to work around most of the missing features by manually adding the HTML elements using the typed HTML module (

sometimes it really makes sense to read the documentation before starting to debug ...🙈

@… A fair question and I wish it were possible! Unfortunately our fiscal sponsor on Open Collective would not allow Kickstarter contributions as expenses (as Kickstarter by definition is prepayments), per the documentation:

Encouraging signs: you're reading documentation for a tool you're considering using, which points you to a GitHub repo for the source.
The repo contains a single file from a commit a decade ago saying they've moved the code elsewhere.
Definitely raises more than a few questions about how up to date the doc I was just reading is...

Something I learned late in my academic life and since trying to implement in most of my teaching: When you do science, you actively have to revisit all your documentation regularly and think of ways to paint a bigger picture of your findings.
When you just document, your viewpoint is sort of chronological or even biographical.
Explaining a new finding in the order you found the puzzle pieces leading to that discovery, usually is not the best way to explain or write about it.
…

It strikes me that one of the things LLMs rob us of is the exact kind of serendipity that much of OSS relies on.
I was just looking something up on https://docs.python.org, which lead me to stumble on a line of documentation entirely irrelevant to what I was trying to do - But it caught my eye because…

Experts are calling it “the worst voter suppression bill ever seriously considered by Congress.”
As the U.S. Senate prepares to vote on a Trump-backed voter ID bill known as the SAVE Act,
millions of citizens who lack easy access to its required forms of documentation are now at risk of disenfranchisement.
“Republicans are singularly focused on making it harder to vote and pursuing this MAGA fever dream,”
explains Ari Berman, national voting rights correspondent for …

Disenfranchise Tens of Millions? Trump’s SAVE Act Targets Women, Poor, Rural & Trans Voters
Experts are calling it “the worst voter suppression bill ever seriously considered by Congress.” As the U.S. Senate prepares to vote on a Trump-backed voter ID bill known as the SAVE Act, millions of citizens who lack easy access to its required forms of documentation are now at risk of disenfranchisement. “Republicans are singularly focused on making it harder to vote and pursuing this MAGA fever dream,” explains Ari Berman, national voting rights correspondent for Mother Jones. “It …

Almost 600 lines of OpenAPI documentation later and I'm done. I think I'll package the `/doc` endpoint with the Docker images since I can just copy over the static directory and have it live in the container. I'll add directions on how to add the docs to the standalone binaries as well which should cover that use case. Overall, it turned out pretty well.
#sisyphus

Ten years contributing to an open source project teaches you things no documentation ever could. Jason Gerlowski, PMC Chair of Apache Solr, is at #bbuzz26 to share what a decade of OSS security experience actually looks like in practice.
Learn more: https://2026.berlinbuzzwords.de/session/oss-security-lessons-from-10-years-at-apache-solr/
Get your ticket: https://2026.berlinbuzzwords.de/tickets/

Estoy probando libretranslate en mi local. https://docs.libretranslate.com/
Un traductor auto hospedado para no depender del traductor de google.

#ReleaseSunday 🎉 Quite a few https://thi.ng/column-store updates over the past month, including further performance optimizations, more tests and documentation updates...
Just also added a small section an…

quick and easy way to build a website for your r package: #rstats #package

Build websites for R packages
Generate an attractive and useful website from a source package. pkgdown converts your documentation, vignettes, README, and more to HTML making it easy to share information about your package online.

AI for writing documentation – The FreeBSD Forums
<https://forums.freebsd.org/threads/ai-for-writing-documentation.101700/> | <https://…

I did a short talk at work a few weeks ago that was basically "if AI is going to be used by users, any commitment to energy sustainability implies reducing the resource cost of AI - and that means good public documentation and comprehensive APIs."
Wish I'd have included some of the figures involved:
https://

Vision Agents vs API Calls
We benchmarked browser-use vision agents against auto-generated API endpoints on the same Reflex app. 47 steps and 495k tokens vs 8 calls and 12k tokens.

Woohoo! I just typed
`(mapcar (lambda (n) (:documentation (meta n))) (keys (oblist)))`
into the REPL of my frankenlisp, and printed out the documentation of every function it knows.
I am getting close to the point that other people can play with this...
#Lisp

"It places page cache pages in a writable scatterlist, separated from the legitimate write region by nothing more than an offset boundary. The design assumes every AEAD algorithm will confine its writes to the intended destination, but nothing in the API enforces this, and nothing documents it as a requirement.
Unfortunately, one AEAD algorithm breaks this silent invariant."
"No other standard AEAD algorithm in the kernel [uses memory that doesn't belong to it as a scratch pad]. GCM, CCM, and regular authenc all confine their writes to the legitimate output area. authencesn alone writes past the boundary."
I'm actually amazed that there's only one bug here. Somehow almost everyone just managed to do the right thing, despite no mechanism enforcing it and no documentation describing it. That's just amazing. It's a testament to the skill of those developers, despite an incredibly bad design.
#copyfail

@… In principle I agree, but one problem seems to be precisely that some things aren’t questioned. There’s certainly a need for documentation and calculation, but this isn’t necessarily the same as Word and Excel files—not just wrt. file formats, but in a more existential sense.

@… In principle I agree, but one problem seems to be precisely that some things aren’t questioned. There’s certainly a need for documentation and calculation, but this isn’t necessarily the same as Word and Excel files—not just wrt. file formats, but in a more existential sense.

@… video group membership is certainly not required to start X.
#FreeBSD #documentation

from my link log —
Nornir: an automation framework in Python.
https://nornir.readthedocs.io/en/latest/
saved 2020-03-23 https://dotat.a…

Following an official github documentation about integrating Claude with Github MCP ends up with github token in a plain text in ~/.claude.json. Other guides I found did the same, except revealing a different github secrets there.
That means I am really really really surprised that there are so few security incidents like axios. I'd guess the situation is much worse.
#claude

I don't know what has changed - although I have a suspicion that it is a reaction by other countries against el cheato's tariff nonsense - but I am observing a significant uptick in the amount and detail of documentation needed in order to export products from the US to foreign countries.

AWS launches Amazon Connect Health, an AI-driven suite to automate clinical documentation, billing and procedure coding, and patient identity verification (Matt Day/Bloomberg)
https://www.bloomberg.com/news/articles/2026-03-05/…

~The Claude Desktop App installs its browser extension into 7 different browsers without consent automatically. This is very much not cool.~
Edit: I misread the article, and this is not what is happening. Instead, these files that are installed allow for easier installing of the actual browser extensions, which is much less evil. No browser extensions are automatically installed. Thanks to @…

Anthropic secretly installs spyware when you install Claude Desktop — That Privacy Guy!
Anthropic's Claude Desktop silently installs a Native Messaging bridge into seven Chromium browsers, including browsers Anthropic's own documentation says it does not support, and browsers the user has not even installed.

The Microsoft SQL Server Docker image names all start with mcr.microsoft.com/mssql/server
Yet their documentation is found on https://mcr.microsoft.com/artifact/mar/mssql/server/about
Trying to browse to https://mcr.microsoft.com/mssql/server results in a 404. That's quite a missed opportunity. 🤷🏻‍♂️

Can confirm this for Arc, Brave, Edge, Chromium, and Vivaldi on my machine:
#Anthropic secretly installs spyware when you install Claude Desktop
https://www.thatprivacyguy.com/blog/anthro

Anthropic secretly installs spyware when you install Claude Desktop — That Privacy Guy!
Anthropic's Claude Desktop silently installs a Native Messaging bridge into seven Chromium browsers, including browsers Anthropic's own documentation says it does not support, and browsers the user has not even installed.

Now the pain in the ass part: documenting all of the routes. I think I'm gonna use Scalar because it's easy and I'm pretty lazy. Carved off the `/doc` directory to host the `openapi.yaml` and `index.html` files. So far, got one single endpoint documented...lol
#openapi #sisyphus

Oh. my. god. For the first time in my life I now have access to a *wide angle camera* on my phone! 😱 🥳
Today I was brave enough to finally update my phone to the latest #SailfishOS 5 (I could absolutely not afford a bricked phone the last ~2 years), and it was the smoothest update ever, so my fears were unfounded. 🙃
However - in typical

I’m planning to teach folks about making their own website later this year, but I’m conflicted about which static site generator to use. Jekyll is what I’m most experienced with, but it seems to be more complex to install and stuff. Jekyll also doesn’t have stuff like multi language websites and alike.
Tho Hugo seems something I’d like to avoid, since apparently themes end up being completely different frameworks from each other so the documentation doesn’t always apply and switching themes on a ready website seems harder?
Are there any other options out there? Ideally easier to install than jekyll, easy to set up on codeberg/github pages, extensible and more complete but without Hugos issues.
I’ll probably migrate my website to whatever I choose as a way to learn the new generator.
Edit: Zola seems interesting, anyone has opinions about that one?
Do note that the target audience includes people who never touched a terminal or any programming language before, so not needing much of those paradigms is preferred (I could, and probably will, give them a basic template to begin with tho)

Anyone have thoughts on dropping Ubuntu 22.04 support in the ngscopeclient CI build moving forward since 26.04 is coming out very soon?
We won't be actively going out of our way to break 22.04 (i.e. adding a dependency not available there) but it will no longer be routinely tested.
The main reason is some Vulkan packaging issues that require special handling for 22.10 and older, the CI scripts will be a lot simpler if I can just forget that (the documentation mentions the nec…

The font and colors are subject to change, (though I'm kind of digging the dark blue), but the next portion of the Warlock project is coming together! This will serve as the public documentation and information site for the project, for folks not wanting to solely use Github. Still a lot of work to do, but making good progress on it.
OH, and the screenshot is being generated by the following "code"

::: section .showcase
::: .blocks-2
::: .block

# Warlock Ne…

Man kann mit dem Flatfile #CMS @… so viel mehr machen, als eine einfache Website wie meine mit kleinem Blog, aber auch dafür ist der Adminaufwand im Vergleich zu WordPress minimal. Habe mir aber nun vorgenommen, mich in den nächsten 2 Monaten wieder etwas intensiver mit …

An underwhelming experience, my first setup of T2SDE. Technically absolutely something very special and blazingly fast and snappy. But at the moment too many usability gaps (installer, upgrading, documentation) to use it (even in beta) as a daily driver. At least for me. A pity.

I don’t want to set up a #smarthome, but the current timer switch just doesn’t cut it. Unfortunately, the documentation for smarthome stuff seems to assume that you already know everything…
I’d like to use an existing headless #RaspberryPi with a RaspBee II to control a single Ikea Tre…

Finally happening
https://www.theguardian.com/world/2026/apr/22/spain-champions-migrations-benefits-with-regularisation-scheme-but-queues-are-long

Yesterday I learned that if you want to edit a custom format e.g. cinematic/spatial video in final cut, you have to be very specific about how you import it. I typically AirDrop them from the phone, but I couldn't figure out how to pull up the cinematic editor.
Apple's documentation is shit, and the only reason I figured it out is because ChatGPT sent me to a Reddit thread where someone linked to a Ripple Training video from 3-4yrs ago that had a small section about this. (1/2…

Structured theme presets with typography controls
Directional controls, progress bar & fullscreen support built in
📖 Full documentation: https://slidewire.dev
🔗 Source code:

My wife and I have passports, as do our kids (from an early age), so we aren't dependent on smudged, barely readable copies of our now-ancient birth certificates.
But it's not about me. It's about every American citizen who has a right to vote without being subjected to expenses, red tape, and potentially arbitrary decisions by political stooges regarding supporting documentation.
SAVE is a dangerous, un-American fraud.

Enzo Health, whose AI tools help home health and hospice agencies automate tasks like patient intake and documentation review, raised a $20M Series A led by N47 (Brock E.W. Turner/Axios)
https://www.axios.com/pro/health-tech-deals/2026/05/04/e…

Exclusive: Enzo Health raises $20M for post-acute software
The company will begin prepping for a Series B raise next year.

And it is a bloated, hard to use product. We moved to Adobe PDF from Foxit. Every boot it launches its crap to ensure you have a license. Multiple popups. Scanning the indexes stalls for 30 to sixty seconds just about every click. When I rebuild the documentation indexes after a release now takes several days where Foxit did it in an hour. And that is on a gigabit network. When on VPN it takes a full five days.

@… it was not a waste of time :-) I enjoy IT support and untangling things such as documentation.

As is somewhat customary, spring time is maintenance time and so this past week was again "infrastructure week" in #ThingUmbrella land, updating various tool scripts, switching to Rclone for uploading/syncing API docs, updating/pinning Prettier version & config, reformatting thousands of source files, re-uploading hundreds of MB of documentation for all current 214 projects, …

In-memory column store database with indexing (WIP)

I don’t want to set up a #smarthome, but the current timer switch just doesn’t cut it. Unfortunately, the documentation for smarthome stuff seems to assume that you already know everything…
I’d like to use an existing headless #RaspberryPi with a RaspBee II to control a single Ikea Tre…

I don’t want to set up a #smarthome, but the current timer switch just doesn’t cut it. Unfortunately, the documentation for smarthome stuff seems to assume that you already know everything…
I’d like to use an existing headless #RaspberryPi with a RaspBee II to control a single Ikea Tre…

@… I’ll document the answer to that question later :)
I was also surprised to find that the RSS documentation itself points to Atom for a couple of things.

MCP stands for Model Context Protocol and is a api for agents. Storyblok has made their api into an MCP and Bernhard is here to tell us about that from the context of a psychologist.
Storyblok is a Content management system which can now have agents mess with it for you.
MCP standard connects agents to external systems. An mcp server runs to enable that.
Isn't the agent smart enough to use the api?
Well, they need documentation to figure out how anyway, so MCP rolls that in I guess? Avoid filling the context window with api docs.
Overlay specification from openapi let's you describe the api with a json doc.
The agent can query the server to search for commands, get the spec for the one they want, then call it.
"Skills" are troublesome because they gry loaded to dev machine and never updated. MCP stuff downloaded live on demand.
There exist tools to convert a rest apt to MCP but they apparently fail often because the machine ideally wants something different to a rest api.
#devWorld

@… for ZFS (the default),
zfs mount -a
@…
#FreeBSD

There are ONE HUNDRED AND TEN distinct colour specifications in the standard Doxygen CSS stylesheet.
Holy shit.
Who needs 110 distinct colours to present technical documentation?

The long test is looking good for the `av1an`, `mkvmerge`, and `cleanup` modules, the Golang client is doing exactly what its supposed to which is a welcome change from yesterday. I need to do long tests on the `ffmpeg` and `handbrake` modules next, but I'll handle those tomorrow.
After that, I'll update documentation, release the `2.0.0` client, and deprecate the old client.
#sisyphus

OpenAI releases ChatGPT for Clinicians, a tool for medical tasks like documentation and research, free for verified physicians, pharmacists, and more in the US (OpenAI)
https://openai.com/index/making-chatgpt-better-for-clinicians

Oh $DEITY I am *so* fed up with the primitiveness and uselessness of modern software IDEs. I so wish I could use a powerful modern graphical IDE, like we had in the 1980s.

@…
Continuing the "gymnastics" theme:
― what sporty word can we use to describe the twenty-four month wait for FreeBSD documentation of a feature that was added in April 2024?
Cc @…

So, my initial review of the Yaelink SIP-T42S before I have the in hand:
- The documentation exists, but can be a pain in the ass to track down exactly what you need.
- Looks like I've got everything already setup to serve the boot configurations and should work with my FreePBX installation. I'll probably get these setup as backups in case I lose a couple of Digium D40 phones during the competition.
- Firmware is still available for the phone, but it is EoL.
- En…

I sometimes believe that Phabricator is AI
― active impediment.
Like, where documentation goes to die.
Process without progress.
<#FreeBSD

@… there's already a separate licence for FreeBSD documentation:
<https://www.freebsd.org/copyright/freebsd-doc-license/>