Den 2026-09-23 skrev Stefan Ram <ram@zedat.fu-berlin.de>:

Johann 'Myrkraverk' Oskarsson <johann@myrkraverk.invalid> wrote or quoted:

If you're wondering how to format text for Usenet, I like Emacs for this >>task. I then paste into Thunderbird for posting, but that's because I'm >>not inclined yet to figure out how to configure GNUS.

To wrap paragraphs, I wrote a script in Python that uses dynamic
programming to find the globally best wrap and made it available
to vim via a key mapping. It was used to wrap this paragraph here.

<snip>

My problem is the unique 'mark up' I need to weed out...

Also, some questionable tips are hiding among the good ones. I will keep
them but add comments. I think it illustrates nicely how careful you
need be when reading LLM produce.

/Martin

--- PyGate Linux v1.5.19
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)

Den 2026-09-24 skrev Martin Sch””n <martin.schoon@gmail.com>:

<snip>

My problem is the unique 'mark up' I need to weed out...

Also, some questionable tips are hiding among the good ones. I will keep
them but add comments. I think it illustrates nicely how careful you
need be when reading LLM produce.

OK, let's see how I managed. Org-mode kind of markup used:

* Heading 1
** Heading 2
~code~
+ list bullet
*bold*
/emphasis/

# Python style inline comment by me.

###################################################################

Mistral Small 4 on bets practice for managing virtual environments
using conda.

Managing virtual environments with Conda (Anaconda/Miniconda) is
essential for isolating project dependencies and avoiding
conflicts. Here are the best practices. By following these
practices, you ensure reproducible, conflict-free environments for
your projects.

* Create isolated environments
Always create a dedicated environment for each project to
avoid dependency conflicts.

Create a new environment with a specific Python version:

~conda create --name my_project python=3.9~

Activate the environment:

~conda activate my_project~

*Best Practice:*
+ Use descriptive names (e.g., `my_project_dev', `data_science_env').
+ Specify Python versions explicitly (`python=3.9' instead of just
`python').

* Export and recreate environments
To share or replicate an environment, export its dependencies to
a YAML file.

Export the environment to a YAML file:

~conda env export --name my_project > environment.yml~

Recreate the environment from the YAML file:

~conda env create --file environment.yml~

*Best Practice:*
+ Commit environment.yml to version control (e.g., Git) for
reproducibility.
+ Use `conda env export --from-history' to export only explicitly
installed packages (avoids unnecessary dependencies).

* Update environments safely
Avoid breaking dependencies by updating packages carefully.

Update all packages in the environment:

~conda update --all~ # This is /not/ recommended by NSC!

Update a specific package:

~conda update numpy~

*Best Practice:*
+ Test updates in a staging environment before applying them to
production.
+ Use `conda list' to check installed versions before updating.

* Remove unused environments
Clean up old environments to save disk space.

List all environments:

~conda env list~

Remove an environment:

~conda env remove --name old_env~

*Best Practice:*
+ Regularly prune unused environments (`conda clean --all' removes
cached packages).

* Use conda-forge for additional packages
Prefer conda-forge for community-maintained packages.

Add conda-forge as a channel:

~conda config --add channels conda-forge~

Install a package from conda-forge:

~conda install some_package~

*Best Practice:*
+ Prioritize conda-forge over the default Anaconda channel for
broader package availability.

* Avoid mixing conda and pip
If you must use pip, install it first in the Conda environment.

~conda install pip~

~pip install some_pip_package~

*Best Practice:*
+ Prefer Conda packages when possible to avoid conflicts.
+ If using pip, document it in environment.yml under `pip:'.

# According to NSC you should /never/ install packages using conda
# /after/ you have used pip. Do /all/ conda (mamba) installs before
# using pip!

* Use `conda clean' regularly
Remove unused packages and caches to free up space.

~conda clean --all~

*Best Practice:*
+ Run this periodically to keep your Conda installation lean.

* Use environment variables for paths
Set environment variables to avoid hardcoding paths.

Set a variable in the current shell:

~export MY_DATA_DIR=/path/to/data~

Or set it in the Conda environment:

~conda env config vars set MY_DATA_DIR=/path/to/data~

~conda activate my_project~

**Best Practice:**
+ Use `.env` files for project-specific variables (load them with
`python-dotenv`).

* Use mamba for faster dependency resolution
Mamba is a drop-in replacement for Conda that resolves dependencies
much faster.

Install mamba:

~conda install mamba -n base -c conda-forge~

Use mamba instead of conda:

~mamba install numpy pandas~

*Best Practice:*
+ Use mamba for large environments to speed up installation.

* Document dependencies clearly
Include a README.md in your project with:
- Environment setup instructions.
- Key dependencies.
- How to activate the environment.

* Example Workflow
** Create environment
~conda create --name my_project python=3.9~

~conda activate my_project~

** Install packages
~conda install numpy pandas scikit-learn~

~pip install some_pip_package~

** Export environment
~conda env export --from-history > environment.yml~

** Share environment
Share environment.yml and README.md with collaborators.

####################################################################

I hope this is readable.

Note that some advice above is /not/ good practice according to the instructions supplied by NSC and NCS provides plausible motivation
for their advice.

https://nsc.gitlab-pages.liu.se/berzelius-training/managing-conda-environments-best-practices/2.%20best-practices/

I have tested none of this so far. The weather has been too nice :-)

/Martin

--- PyGate Linux v1.5.19
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)