A collection of notes and bookmarks.

Twenty Five Years of The New York Times Crossword Data

I first started reverse-engineering the .puz file format about 6 years ago. Its a very efficient crossword format that goes to great lengths to compact the data into a small payload. With that done, the The New York Times crossword archive had interesting information. This is a CSV of collected data ... read more

Feature Request #282

If I ever open a new document or email, then change my mind and close it without adding any sort of text - please, for the love of god, don't prompt me to Save it before closing. You know who you are. read more

Setting up Dreamhost email to work in Gmail

You've registered a domain with Dreamhost and you might even be hosting a site there. In an ideal world you'd use Gmail with the domain for your email but you don't really want to shell out the money and have yet another new login. Having your email forward to your ... read more

Instagram media by account in JSON

Scraping inline JSON for fun and frustration This will work until Instagram go ahead and make changes to their JS objects. Until then, go crazy. So... within the Instagram page source is a SCRIPT tag that contains pretty much everything (almost) that you'd need to make a feed of your ... read more

Mining The New Yorker for Haikus

To be fair, these are not properly formatted Haikus but they are fun to extract and search for. This demo focuses on a current article from The New Yorker by pulling one from a JSON feed. However you can easily adapt this to any piece of text. The main PHP ... read more