Quantam wrote:
vampirefrog wrote:
https://vgmrips.net/packs/releases
I was really hoping for something that didn't require parsing HTML pages

So, after playing around with it some (and appreciating just how slow the VGMRips server is), there are a couple ways we could go to make this efficient (new pages that could be added to the site):
1. A list of all files currently in the archive containing at least: the file path, file size, and file last modified date/time, either as CSV or XML
2. An XML-based representation of the latest releases page
https://vgmrips.net/packs/latest that provides either all of them on a single page or paged, with some large number of items on each page (e.g. 500 packs). To be efficient this version would need to contain the info from the release's page (e.g.
https://vgmrips.net/packs/pack/earthwor ... -boy-color, which would preferably include the ZIP file's date/time as well), so that only a single XML feed needs to be pulled to get all the information necessary to determine new/updated files and download them.
Optionally, either of these could be combined with an HTTP parameter that lets the user specify the oldest date to return - the latest date they have packs for. This would cut down on disk/database accesses and amount of data the server needs to transmit for packs the user already has, but makes the CSV/XML data uncacheable.