The current repository is quite cluttered with many versions of the same images.
There is no need to keep all these different versions around if we have a pipeline that:
Resizes the original to the desired resolutions.
Converts the resulting images into the (compressed) formats we need (now: avif and jpg)
For local development you would run the pipeline once too and ensure the files are in the .gitignore. It is an open question of how to determine the level of compression that satisfies a trade-off between visual quality and file size in an automated fashion.
The current repository is quite cluttered with many versions of the same images. There is no need to keep all these different versions around if we have a pipeline that:
For local development you would run the pipeline once too and ensure the files are in the
.gitignore. It is an open question of how to determine the level of compression that satisfies a trade-off between visual quality and file size in an automated fashion.