search

exwm / yt_clipper

Mark up YouTube videos and quickly generate clipped webms.
https://openuserjs.org/scripts/elwm/yt_clipper
MIT License
50 stars 12 forks source link

Add documentation for linux installation #44

Open rehandaphedar opened 2 days ago

rehandaphedar commented 2 days ago

What does this PR contribute?: Adds documentation for installing yt_clipper on Linux.

Which issues does this PR address?: Currently, there are neither prebuilt binaries nor instructions for installing from source on Linux,

Notes: This is mainly intended as a temporary fix until Linux binaries can officially be supported.

exwm commented 12 hours ago

Hi, thanks for your contribution.

I think it would be better if the installation directory was added as a configurable variable at the top of the script. I think it is unusual to want to clutter the home directory with a package.

I'm using poetry during development and it may be worthwhile to add alternative installation instructions using poetry. See https://python-poetry.org/. Using poetry will ensure a compatible python version and will use locked dependency versions, but users may understandably not want to install poetry. I don't mind if the instructions are not detailed, that can be improved later.

exwm commented 11 hours ago

It may be useful to mention git pull for updating yt_clipper and for checking out specific tagged versions. This is another area where poetry or another tool like pipx may work better than plain venv and pip, as they can sync the dependencies with the lock file.

rehandaphedar commented 10 hours ago

I think it would be better if the installation directory was added as a configurable variable at the top of the script. I think it is unusual to want to clutter the home directory with a package.

Oh yeah, I was also installing mine in ~/.local/share/yt_clipper.

users may understandably not want to install poetry

This is what I was thinking of. I think while initially figuring out how to install this package, I didn't know that pipx can have separate python versions per package. After reading more about it, pipx seems to be the best fit to me.