The Peanut Gallery
May 14th, 2006
7:17 am
Peanut Gallery is an auto-web-gallery that does not require any database
back-end. It uses information from the file itself (filename, creation
date, and path) to populate a high-speed, indexed, and robust cache so
users can find pictures quickly and easily.
So why did I design just another web gallery? Well, I've used Gallery, Coppermine, and a variety of other gallery programs out there. The one thing that I don't like about them is that their designs keep the pictures in disarray. Therefore, it's a task to keep my albums current, without duplication, with the web gallery. I am anal, and keep my photo's in appropriately named directories, with descriptive filename's. I don't want to duplicate that process.
I set out to find an auto-gallery that used relied and update itself based on the directory structure. If you changed a filename, for example, the cache would update as well. I found picKLE, which "generates thumbnails and resampled images on the fly and caches them." This project was the closest match to what I was looking for. However, like many of the related projects, picKLE is not maintained, and is fraught with performance issues.
I use the Peanut Gallery on this site. It's a good demonstration of the power and speed of it's initial stages. It is still in it's infancy and I have not registered it yet with SourceForge. I do plan on releasing the source soon.
Note: I'm looking for an algorithm (preferably that uses ImageMagick) that will recognize faces. I want it to be able to identify the height/width of an image so that when it crops the image for a thumbnail, it will not crop out people's faces. Let me know if something like that exists. Contact me via the links on the top of this page. Thanks.
