sueden.social ist einer von vielen unabhängigen Mastodon-Servern, mit dem du dich im Fediverse beteiligen kannst.
Eine Community für alle, die sich dem Süden hingezogen fühlen. Wir können alles außer Hochdeutsch.

Serverstatistik:

2 Tsd.
aktive Profile

#orgmode

16 Beiträge16 Beteiligte2 Beiträge heute

Should I dare to do a "why #Markdown is one of the worst lightweight markup syntax languages there is"-session at a UX/UI-dominant #barcamp? 😜

C'mon, do push me over that cliff! 😆

Background: karl-voit.at/2017/09/23/orgmod 👉 it's related but would be a different focus since I won't push #orgdown that much - people can be happier with other LMLs as well as long as it's not the #MD hell. (Sneak preview: I'm writing a long article on all the MD issues in order to explain it once and for all since the Mastodon discussions are really annoying to me.)

public voit - Web-page of Karl Voit · Org Mode Syntax Is One of the Most Reasonable Markup Languages to Use for TextOrg Mode Syntax Is One of the Most Reasonable Markup Languages to Use for Text
#LML#AsciiDoc#rSt

#Emacs #OrgMode is looking for someone to maintain Worg, the community-driven documentation.

See the call for volunteer on the mailing list and on Reddit:

👉 list.orgmode.org/87o6wirw8t.fs

👉 reddit.com/r/orgmode/comments/

You can read Worg on orgmode.org/worg/ and access to its sources on git.sr.ht/~bzg/worg

Thanks for boosting this 🙏

list.orgmode.orgJoin the Org Mode project as the Worg maintainer

gptel-org-tools update.

1. Cloned to
https://codeberg.org/bajsicki/gptel-org-tools, and all future work will be happening on Codeberg.
2. Added
gptel-org-tools-result-limit and a helper function for it. This sets a hard limit on the number of characters a tool can return. If it's over that, the LLM is prompted to be more specific in its query. Not applied to all tools, just the ones that are likely to blow up the context window.
3. Added docstrings for the functions called by the tools, so LLMs can look up their definitions.
4. Improved the precision of some tool descriptions so instructions are easier to follow.
5. Some minor improvements w/r/t function names and calls, logic, etc. Basic QA.

Now, as a user:
1. I'm finding it increasingly frustrating that Gemma 3 refuses to follow instructions. So here's a PSA: Gemma 3 doesn't respect the system prompt. It treats it just the same as any other user input.
2. Mistral 24B is a mixed bag. I'm not sure if it's my settings or something else, but it fairly consistently ends up looping; it'll call the same tool over and over again with the exact same arguments. This happens with other models as well, but not nearly as frequently.
3. Qwen 2.5 14B: pretty dang good, I'd say. The Cogito fine-tune is also surprisingly usable.
4. Prompting: I have found that a good, detailed system prompt tends to /somewhat/ improve results, especially if it contains clear directions on where to look for things related to specific topics. I'm still in the middle of writing one that's accurate to my Emacs set-up, but when I do finish it, it'll be in the repository as an example.
5. One issue that I still struggle with is that the LLMs don't take any time to process the user request. Often they'll find some relevant information in one file, and then decide that's enough and just refuse to look any further. Often devolving into traversing directories /as if/ they're looking for something... and they get stuck doing that without end.

It all boils down to the fact that LLMs aren't intelligent, so while I have a reasonable foundation for the data collection, the major focus is on creating guardrails, processes and inescapable sequences. These will (ideally) railroad LLMs into doing actual research and processing before they deliver a summary/ report based on the org-mode notes I have.

Tags:
#Emacs #gptel #codeberg #forgejo #orgmode #orgql #llm #informationmanagement #gptelorgtools

PS. Links should work now, apparently profile visibility affects repo visibility on Codeberg. I would not have expected that.

PPS. Deleted and reposted because of strong anti-bridge sentiment on my part. Screw Bluesky and bots that repost to it. Defederated: newsmast.*

Zusammenfassungskarte des Repositorys bajsicki/gptel-org-tools
Codeberg.orggptel-org-toolsTooling for LLM interactions with org-mode. Requires gptel and org-ql.

Das #Barcamp #Graz 2025 wird in den #lightningtalks der Linuxtage angekündigt.

Save-the-date : 17.05.2025

Lightning talks sind #Kurzvorträge bei denen man sich kurzfristig zu einem (max. 5 Minuten langen) #Vortrag meldet.

Außerdem diesmal noch dabei:
#pijul, #msys2, #openwrt und #orgmode, Methoden zu #spam-Verhinderung, #softskills,
#private.coffee ,

pretalx.linuxtage.at/glt25/tal

pretalx.linuxtage.atLightning Talks Grazer Linuxtage 2025Kurzvorträge (max. 5 min) zu verschiedenen Themen. Quick talks (max. 5 min) about different topics.

gptel-org-tools update.

Edit: there's some kind of issue with
@Codeberg@social.anoxinon.de which prevents the link from working (returns 404). The old (but up to date) repo is here: https://git.bajsicki.com/phil/gptel-org-tools

1. Cloned to
https://codeberg.org/bajsicki/gptel-org-tools, and all future work will be happening on Codeberg.
2. Added
gptel-org-tools-result-limit and a helper function for it. This sets a hard limit on the number of characters a tool can return. If it's over that, the LLM is prompted to be more specific in its query. Not applied to all tools, just the ones that are likely to blow up the context window.
3. Added docstrings for the functions called by the tools, so LLMs can look up their definitions.
4. Improved the precision of some tool descriptions so instructions are easier to follow.
5. Some minor improvements w/r/t function names and calls, logic, etc. Basic QA.

Now, as a user:
1. I'm finding it increasingly frustrating that Gemma 3 refuses to follow instructions. So here's a PSA: Gemma 3 doesn't respect the system prompt. It treats it just the same as any other user input.
2. Mistral 24B is a mixed bag. I'm not sure if it's my settings or something else, but it fairly consistently ends up looping; it'll call the same tool over and over again with the exact same arguments. This happens with other models as well, but not nearly as frequently.
3. Qwen 2.5 14B: pretty dang good, I'd say. The Cogito fine-tune is also surprisingly usable.
4. Prompting: I have found that a good, detailed system prompt tends to /somewhat/ improve results, especially if it contains clear directions on where to look for things related to specific topics. I'm still in the middle of writing one that's accurate to my Emacs set-up, but when I do finish it, it'll be in the repository as an example.
5. One issue that I still struggle with is that the LLMs don't take any time to process the user request. Often they'll find some relevant information in one file, and then decide that's enough and just refuse to look any further. Often devolving into traversing directories /as if/ they're looking for something... and they get stuck doing that without end.

It all boils down to the fact that LLMs aren't intelligent, so while I have a reasonable foundation for the data collection, the major focus is on creating guardrails, processes and inescapable sequences. These will (ideally) railroad LLMs into doing actual research and processing before they deliver a summary/ report based on the org-mode notes I have.

Tags:
#Emacs #gptel #codeberg #forgejo #orgmode #orgql #llm #ai #informationmanagement #gptelorgtools

Zusammenfassungskarte des Repositorys phil/gptel-org-tools
Phil's Gitgptel-org-toolsTooling for LLM interactions with org-mode. Requires gptel and org-ql.
Antwortete ZeStig :emacs: :nix: :rust: :gnu: :archlinux:

@zstg I've never used #orgmode before, but I may take a look at it. I'm certainly all for my notes being in plain-text files rather than in a proprietary format.
The one drawback, which I'm not sure I can overcome, is that I often need to share and concurrently edit notes with coworkers, and it is probably neither practical nor reasonable to ask them to use org mode.

Did some minor styling today of my agenda today.
The clocking table looks now better and the agenda looks a bit cleaner in general.

Clocking was a bit smoother with Toggl as I had it automated quite nicely with the iOS focus modes.

But in general it feels a bit like coming home.
It's such a nice thing to record your thoughts along side your tasks and in such frictionless environment.

one of my 'hello world' experiments with #Emacs #orgmode is using Tangle to manage the homepage of radio.teledyn.com — the /challenge/ is having the master on my laptop but the /published/ version is across the LAN. I haven't tried using tramp to publish directly, only because my default reflex is to open a termnal and scp.

but on today's edit I noticed: how is it even /possible/ there is still no Font-Awesome code for #Pixelfed — hey, it's 2025! Let's get with it!

As part of my journey remove dependencies on US companies I now fully switched back to #orgmode from #Todoist
In addition I switched back from #Toggl (which is actually an Estonian company) to org mode clocking as well.

So far it has been pretty painless as I completely separated my personal agenda from my work agenda.
They live in the same directories but on different computers.
This way I don't have to make any big adjustments to the configuration.

I lose a bit of mobility but I already noticed the advantage of taking notes alongside your tasks again.
It's just such a great feature to have.

I had a #Shortcuts workflow on my iPhone that triggered Toggl.
This way I knew when I started and stopped working for the day.
I replaced this with a dedicated calendar and then Shortcuts creates an entry when I arrive at work and one when I leave work.
This is just for reference, however.

Fortgeführter Thread

Meeting notes can be found at list.orgmode.org/87ecxkb1pv.fs

TL;DR: timezone support; search in exported HTML; lualatex as default pdf export engine; exporting
non-latin documents; nested special blocks;
parser-based fontification; elisp inspector; inline special blocks;
org-special-block-extras; noweb references; native macOS support for org-capture;
breadcrumb-mode

Earlier notes on the meetup page:
orgmode.org/worg/orgmeetup.htm

#OrgMeetup #orgmode #meetup #emacs

CC: @sacha

list.orgmode.org#17 [[bbb:OrgMeetup]] on Wed, Apr 9, 19:00 UTC+3

Are you using Bluesky primarily as a source for links but are overwhelmed by the massive inflow of posts and doom? I wrote a silly #Emacs package to address this personal need of mine.

hotsky.el fetches links from your #Bluesky feed, sorts them by frequency, and presents them in an #orgmode buffer.

github.com/rtrppl/hotsky

Behind the scenes: the package uses #cURL. It was put together in the last 48 hours, so there still might be some hidden issues. You need a Bluesky app-password for this.