Open lanerussell opened 9 months ago
I'm experiencing this as well, even with a persistent copy of the entire /photo-stream folder.
Same here. Interested in following this issue.
I've tried changing a few lines in https://github.com/benubois/jekyll_image_processing without much effort and rebuilding from there, but without success. I'm not sure whether the issue lies within this resource or photo-stream.
Experiencing this as well. Takes well over an hour to generate, and photos/large, photos/thumbnails etc. are not even created.
Even if you manually create the folders and set permissions to 777, nothing is put into the folders and with each restart of the containers all the photos are regenerated for a stupidly long time.
Hey guys, may I ask you go give this experimental feature a go? https://jekyllrb.com/docs/configuration/incremental-regeneration/
Basically just add incremental: true
in your _config.yml
file.
My results so far have been reducing generation time from 20 seconds to 6 seconds on subsequent starts with no modifications on the image folder.
I am still going to debug if some of my changes are causing some extra iterations over images or something but I am not very experienced with Ruby and Jekyll to be honest.
One more thing to add to the initial question: yes, I think the behavior is intended. Generating the photos is part of the Jekyll build process which is always being triggered via the serve
command. There is a --skip-initial-build
option though which will "skip the initial site build which occurs before the server is started".
This actually looks like the behavior you would expect, but what about the first time starting photo-stream? You would have to manually build the page because the _site
folder should be empty anyway (also as per default this is not persisted).
Is this behaviour intended? I can understand the site being regenerated, but it seems that all photos are resized into thumbnail, tint, and large every time the container restarts. With my small collection of ~160 photos, this takes over 20 minutes, during which time the site is totally down.
I'm running the site in docker-compose, and am using volumes to keep a persistent copy of the site: