NYC got a proper snowstorm back in January. A few photos from that day, for posterity.






About a year ago I posted a photo of my old iPod and added “someday I’ll share all the surgery I’ve done to keep it working”. I think that day has arrived!

We visited The Herb Lubalin Study Center of Design and Typography aka Lubalin Center a few weeks ago. We loved it. Make an appointment via email, and don’t miss it if you’re into design and typography and have a couple of free hours in NYC.









I live relatively close to the Queensboro Bridge, but I recently crossed it by foot for the first time. I was surprised and kind of disappointed to discover that there’s no way to get to the Roosevelt Island from it, given that it passes right above it.
Upon some light research I learned there was a way to move between the bridge and the island —the Elevator Storehouse, a building that connected to the bridge’s platform and had elevators to move cars and people up and down— but it was closed in 1955 and knocked down in 1970.

It’s the building attached to the left of the bridge in this 1955 postcard I got from eBay. There are not a lot of other pictures, and none are in high resolution, but I’ve found some interesting stuff in these links:
It was also featured in this 1963 episode of Naked City:
Cars drive so fast through the bridge nowadays that if feels wild this was a thing back then. Fascinating!
I just found out about ShadeMap, a “global simulation of mountain, building, and tree shadows for any date and time.” I’m not sure how accurate it is (and I’m not willing to check myself), but it’s a really cool tool to play with.
Saw it via this Howtown short. Howtown is cool too!
Something I didn’t know until fairly recently is that with every new US administration, the White House website gets relaunched and the old one is archived. Makes sense — starting the site from scratch is probably the cleanest method to keep the historical records clear.
The beautiful consequence of this is that the old WH sites, which can be consulted, are frozen in time. I’m particularly in love with the second version of the site during Clinton’s administration, active between around 1996 and 1999, if I’m not mistaken.
Everything feels so innocent and honest that I spent a long time digging through the archive. You could send an email directly to Vice President Al Gore, but they would only reply via postal mail. You could also check a “photo scrapbook” of the First Lady’s latest travels. Or you could consult the “Inside the White House” newsletter for kids, whose Spring 1999 edition focused solely on the presidential pets.

I really miss this era of the web. One of the last bastions of this philosophy (simple and wholesome content with fun, creative design) is Neocities, which I find surprisingly refreshing and inspiring. I’ve been spending more and more time there as a break from the constant negativity and ongoing enshittification everything else has to offer.
I started using Swarm nearly 15 years ago to keep track of the places I visit, and I’ve stuck with it because I love seeing my journey through life mapped out. It’s been handy for finding specific places I’ve been to too, but the app’s search tools suck, and they’ve even removed some features recently.
It’s a bit sad to see how few people use it these days. From what I know, Foursquare’s main business now revolves around selling location data and services, so keeping a consumer app that only a small group of people uses must be hard to justify. This has made me worry about the future of the activity data I’ve collected over the years.
Swarm also doesn’t have a usable web interface, so almost everything relies entirely on the mobile apps. Sharing or looking up anything in a browser is extremely hard —a classic walled garden.
Over the holidays, I asked ChatGPT to write a simple PHP script to download all my check-in data via the Foursquare API, and, to my surprise, it worked perfectly on the first try. All I needed to do was create an app in their developer console, link it to my account, and run the script. In the end, I got a 14 MB JSON file with all my check-ins, including coordinates, venue categories, and addresses for every one of them.
The next step was clear: I built a map to display all those check-ins, along with some search functionality, so my data wouldn’t only be backed up safely, but also accessible from anywhere.
The result is Paseo, a little tool that I’m quite proud of.

This is how it looks in motion:
This is a personal project, so I haven’t set it up for others to use. However, if anyone’s interested, I could easily create a folder for them. I’m using Leaflet for handling the map, a customized Mapbox tileset for the visuals, and the messiest JS code you can imagine. ChatGPT helped me with some of it, so who knows where it borrowed the code from—but hey, it works.
Here are a few things I think turned out pretty well:
F to search and R to recenter the map. It’s also mobile-friendly, and I’ve been using it a lot lately to revisit old spots.One thing I’m particularly proud of is avoiding a pet peeve of mine: the “and 1 other” paradox. This happens when list truncation isn’t smart enough and displays something like “A, B, C, and one more” instead of simply showing “A, B, C, D.”


Anyway, feel free to check out the map here. If you’re still using Swarm (or used it a lot in the past), let me know! I might be able to set it up for you too. If you don’t know how to contact me, you can just ping me on Bluesky.