I take a ton of photos but I don’t like Instagram so I added a mosaic feed on my site that I dump photos onto every 3-6 months.
The gallery is a static Jekyll page hosted on GitHub Pages. The images live separately in Amazon S3, so the site’s Git repository doesn’t have to carry the photo collection. A YAML file lists the images, and a small JavaScript gallery loads them in batches as you scroll. Thumbnails keep the initial download smaller; clicking a photo opens the larger version.
There’s also a less visible part of publishing photos: deciding what information travels with them. Image metadata can include GPS coordinates, timestamps, and camera details. The processing pipeline strips metadata before uploading, applies the image’s orientation, and creates the thumbnail and larger copy. The collection also has a cleanup script that replaces original filenames with simple numbered names. I published most of the anonymization code in tidyData, my collection of small tools.
You can see the collection at /photos.