jmkb

I do not enjoy looking at a golf course. (A minigolf course is a different kettle of fish.. of course.)

Thanks, this is a really fun way to view the data. I don't think the no-google/no-yelp filters will really accomplish this goal though. Most of the ones I examined were either variations in the restaurant name or not really restaurants (gas station convenience store, hotel with breakfast, catering LLC with no storefront, etc.) The google and yelp datasets in NYC are really quite good.

Thanks for building and maintaining this lovely project.

I'm an OSM mapper & enjoy linking mapped benches to their profiles on OpenBenches. It gets hairy, though, when a single bench features multiple memorial plaques. Wonder if you'd consider revising your data model to permit multiple inscriptions per bench?

Decades ago, my partner used Google Voice for texting -- really handy, texts just showed up in the gmail inbox, and could be replied to from there. She didn't like cellphones, but usually carried one of the old "Kindle Keyboard" models with unlimited 3G data. The Kindle had simple web browser that could load the low-spec gmail interface, so in essence she had a fully functional SMS device, with no monthly charges.

Notification of incoming texts was the only problem. I jailbroke the thing and started trying to schedule network requests, thinking I'd add some kind of new message counter on the home screen. This proved hard. But it occurred to me that the best place for the counter would be right next to the Kindle's device name, at the top of the screen. And the device name could be updated from her Amazon account.

So I automated a web browser on the home server to log into Amazon and update the device name to "My Kindle (x)" where x was the number of unread Google Voice texts. The Kindle would update the name on the home screen in less than a minute. This worked for years!

(Eventually that Kindle was stolen. I wanted to update its name to something foul but the device disappeared from her account too quickly.)

I grew up with rotary phones, and dialing (on a physical dial) really was fun.

Disassembling my first junk phone and playing with the rotary relay was even more fun.

But the best was learning to dial numbers by manually pulsing the switch on the handset cradle. Timing was tricky but luckily there was some leeway. (And it helped that phone numbers were shorter then.)

I guess it's time to stop searching from the url bar (sorry, "awesomebar") and create a local html file as the startup and new tab page, with search boxes for all your fav engines. Just like the good old days.

I found that upgrading mine with a full keypad -- including the ABCD column on the right -- did wonders for my rep. And of course a switch to toggle oscillators, so I could send red box tones.

Just to be clear, I find conversion of viable multifamily buildings into mansions generally repugnant.

The opposite process also happens, though... not just a 3- and 4-story residential buildings demolished for large residential towers (plenty of that in NYC) but single-family brownstones carved up into smaller dwelling units.

The map caption "Conversions Have Erased Some Gains in NYC's Housing Stock" would make a less deceptive headline. The map shows data from a defined timespan ("between 2010 and 2021") and shows that there's not a single community district that hasn't added housing units. The headline, meanwhile, refers to a vague "70-plus years" timeframe. That's long enough that you could probably find hundreds of buildings that have bounced between single- and multi-family several times.

And in fact the underlying research characterizes any building that was ever recorded as multifamily and currently isn't as a "loss." I fear it risks romanticizing the 20th century's cramped tenements.

(And again, just to be clear, I find conversion of viable multifamily buildings into mansions generally repugnant.)

Brian Eno developed tape-delay audio tech, released a few mobile apps, is a founder of the Long Now Foundation... not a scientist but definitely a bit science-y.

Eventually, I imagine a new licensing concept will emerge, similar to the idea of music synchronization rights -- maybe call it "training rights." It won't matter whether the text was purchased or pirated -- just like it doesn't matter now if an audio track was purchased or pirated, when it's mixed into in a movie soundtrack.

Talent agencies will negotiate training rights fees in bulk for popular content creators, who will get a small trickle of income from LLM providers, paid by a fee line-itemed into the API cost. Indie creators' training rights will be violated willy-nilly, as they are now. Large for-profit LLMs suspected or proven as training rights violators will be shamed and/or sued. Indie LLMs will go under the radar.

Picture an analog clock and it can be 0.008 hours/hour... about one percent of an hourhand's movement.

Just to be clear: this is an individual user's "diary" (blog) hosted at openstreetmap.org, not in any way an official statement from the OpenStreetMap Foundation.

And I agree with the author that the design of Bing Map Builder looks problematic. The biggest problem being that it's segregating its contributors from the general OSM population in the name of "privacy."