max's Avatar

max

@notplants

experiments in online gardening permacomputing, mycology, queer poetry they/them https://mfowler.info https://sunbeam.city/@notplants

509
Followers
426
Following
474
Posts
05.05.2023
Joined
Posts Following

Latest posts by max @notplants

im curious how you define "handcraft RL environment"?

does handcraft mean not using the model to design the environment the model runs in? or is this a specific term?

i'm out of the loop, if you feel like explaining

02.03.2026 20:02 πŸ‘ 2 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

no seed round, but here is an on-pds budget alternative bsky.app/profile/notp...

02.03.2026 19:02 πŸ‘ 2 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

yup I am using ghostty without tmux (somehow I like the simplicity of it) -- would definitely be appreciative if a fix for copying softwraps without hard newlines lands πŸ‘Ύ

01.03.2026 00:21 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

I currently have a weird workaround workflow of copying claude code output into a text file so I can remove the newlines before pasting it into another terminal. with opencode i can just copy and paste directly, but not sure what tui magic they use

28.02.2026 18:00 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

hi thanks for your work!

not sure if intended or user error, but I am stumped figuring out how to copy out claude code output without softwrap newlines appearing as real newlines in claude code in ghostty

in opencode, I don't have this issue so i wonder what the difference is between the tuis

28.02.2026 17:58 πŸ‘ 0 πŸ” 0 πŸ’¬ 2 πŸ“Œ 0

people often object to this because testimony of direct perception can be falsely given, but thats a legal point, and doesnt refute to the basic fact that this is just the ground of knowing anything whether you like it or not

26.02.2026 21:48 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

ty for intel :) I haven't tried adding a devShells.default to a flake yet

but afaict, getting the built binary on the path for the non-dev case is basically all i need to do

26.02.2026 15:22 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

@fierybrain.bsky.social could you say more what you mean?

I had also been thinking about this as a way of backing up static web content fwiw

26.02.2026 15:02 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Preview
Fixing the 'dumb opinions' problem: Addison, Johnson, Walpole What do 18th century writers say about posting online?

Have you noticed how some people have bad takes? Horrid!

Couldn't we empower them to engage with unfamiliar topics in an informed way? How do we make context travel?

In our latest dripline post, @literaryarsonist.press uses the 1700s as a lens to explore the structure of online discourse.

25.02.2026 21:40 πŸ‘ 12 πŸ” 2 πŸ’¬ 0 πŸ“Œ 2

like if there are any conventions of what form is most expected ...

so many projects I look at have such a complicated flake.nix but I'm not entirely sure why

25.02.2026 22:09 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

I actually had been hoping to find someone to look at this flake.nix -- I feel like I don't have a sense of what is actually mostly useful/sane to include to make something useful to another person

25.02.2026 22:07 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

code and tests aren't the moat. they never were. no need for agents to copy a project. forking was always an option. the moat of an oss project is its community, its governance, and its ecosystem connections. agents will never be able to replicate that. don't hide. let's build in the open, together

25.02.2026 19:30 πŸ‘ 169 πŸ” 21 πŸ’¬ 4 πŸ“Œ 5

added a flake.nix to make this easier to install and run for nix users

bsky.app/profile/notp...

25.02.2026 17:46 πŸ‘ 1 πŸ” 0 πŸ’¬ 2 πŸ“Œ 0

or once you have pds-git-remote installed, you can simply run:

git clone pds://notplants.bsky.social/pds-git-remote

(minus the chicken and egg problem)

25.02.2026 17:44 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

:-) yes it is here

curl "https://bsky.social/xrpc/com.atproto.repo.listRecords?repo=notplants.bsky.social&collection=net.commoninternet.pdsgit.state" | jq

25.02.2026 17:43 πŸ‘ 1 πŸ” 1 πŸ’¬ 1 πŸ“Œ 0

the details of oauth are currently a bit murky to me (maybe always will be lol) but I would like to add this in one form or another

25.02.2026 16:04 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

and ty for the link to git-credential-oauth

I like the idea, and know that I do want to integrate this with oauth in some way. the larger project this was actually a subtask for definitely involves app authenticating w atproto via oauth, then somehow passing a token to pds-git-remote seamlessly

25.02.2026 16:03 πŸ‘ 1 πŸ” 1 πŸ’¬ 1 πŸ“Œ 0

and even a pds being called a "Personal Data Server" i feel like the name sort of implies one should be able store whatever kind of data they want in there

25.02.2026 16:00 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

@filippo.abyssdomain.expert this is also how I thought tangled worked until I looked into it more deeply

after talking with the tangled devs, I understand why they made the choice they did, but I still think its interesting to also make an atproto-git thing that doesn't require more than a pds

25.02.2026 15:59 πŸ‘ 2 πŸ” 1 πŸ’¬ 1 πŸ“Œ 0

this project was made via opus primarily via this plan (codeberg.org/notplants/pd...) and has been minimally but not thoroughly tested. currently an exploration-of-concept

24.02.2026 04:45 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

```
# log in to your pds
git-remote-pds auth login --pds-url your-pds.example.com --handle alice.example.com

# push an existing repo
cd my-project
git remote add pds pds://alice.example.com/my-project
git push pds main

# clone it elsewhere
git clone pds://alice.example.com/my-project
```

24.02.2026 04:42 πŸ‘ 1 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0

"this is not the most efficient way to do git, and is not meant as a replacement for tangled, or for anything really

however it seems to work, and can be used without installing anything additionally on your server

pds-git-remote is a git remote helper that stores repositories on a pds directly"

24.02.2026 04:41 πŸ‘ 2 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0
Preview
pds-git-remote ⚘ use your pds as a git remote if you want to ⚘

use your pds as a git remote, if you want to

codeberg.org/notplants/pd...

24.02.2026 04:39 πŸ‘ 33 πŸ” 7 πŸ’¬ 4 πŸ“Œ 2
lexa.fyi's teal-coloured garden with a row of seven floral outlines of different colours.

lexa.fyi's teal-coloured garden with a row of seven floral outlines of different colours.

When spores.garden generates your theme, it also determines if you get a special spore✨. It will be displayed in your garden to start, but other gardeners can steal it, enabling it to travel throughout the network of gardens! Can you find any spores in the wild?

23.02.2026 19:25 πŸ‘ 21 πŸ” 6 πŸ’¬ 2 πŸ“Œ 2

@jackvalinsky.com ty for the pointers to the other pds implementations. I added them to the readme.

and sounds like you found the co-op cloud source. in short: a network of tech co-ops + devops framework for open source applications oriented around "the configuration commons"

23.02.2026 13:46 πŸ‘ 1 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0
Preview
Matrix - Decentralised and secure communication You're invited to talk on Matrix. If you don't already have a client this link will help you pick one, and join the conversation. If you already have one, this link will help you join the conversation

@nathan.pds.vpsm.sh the matrix room for coop cloud is quite active these days if you want to swing by matrix.to#/#coopcloud-...

23.02.2026 13:32 πŸ‘ 0 πŸ” 0 πŸ’¬ 0 πŸ“Œ 0

curious to hear if it works for you if you try it out

23.02.2026 03:55 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0
Preview
bluesky-pds bluesky-pds

made a co-op cloud recipe for a bluesky-pds

git.coopcloud.tech/coop-cloud/b...

23.02.2026 03:52 πŸ‘ 18 πŸ” 2 πŸ’¬ 2 πŸ“Œ 0

but I'm also hearing in this exchange, there is a whole other category of unparseable, where its not tiredness, and could potentially be weird cases where the grammar is undecided even according to an idealized construction, which is new to me

19.02.2026 21:02 πŸ‘ 1 πŸ” 0 πŸ’¬ 2 πŸ“Œ 0

even if its just being tired, that to me implies that in the moment of tiredness, there was the experience/existence of a system (even internally) which could produce the tired output and was somehow at odds with a more pure representation

19.02.2026 21:01 πŸ‘ 0 πŸ” 0 πŸ’¬ 1 πŸ“Œ 0