I loved working with physical paper and collage, but I don't really have the space/time these days. Instead I work with digital art in many different ways, and I've found that scans of old book pages can lead in some similar directions. There are a lot of different archive collections with free scans, and there's a great centralized collection of old, public domain works at the Internet Archive. You can download PDFs or zips of jpegs, but they aren't always of the maximum resolution. For a while I was using cemerson's bookmarklet: a JS bookmarklet that walks an archive.org reader's image URLs and prompts you through downloading them one at a time.

I also had to maximize the on-screen page-size, inspect the image to grab its actual URL, open that in its own tab, and click the bookmarklet. But it worked! I downloaded a lot of weird books. And often it was the blank, water-stained pages that were best!

And then I accidentally nuked all of my bookmarks (yes, they were backed up, but I didn't check for a month and so the backup got overwritten, long story).

So, since I had to get the bookmarklet back anyway, I thought I could take the opportunity to make it easier.

I had Claude Code turn it into a real Chrome extension — same URL logic under the hood, but now it's a toolbar icon. Click it, it auto-detects the current book's page count and image URL, you confirm a range, it downloads the lot at full resolution while you go to find more weird books to mine.

It's for personal use — unpacked, not on the Web Store — and it only touches freely-viewable scans. It doesn't touch material that's borrowed rather than public domain. I'm not trawling the whole site and downloading everything it finds, so as far as I know I'm not violating any TOS.

The code is on GitHub: ia-book-page-downloader. If you also trawl archive.org for source material and don't want to load-unpacked an extension, the original bookmarklet still works fine.