Debugging help is indispensable!

Unsurprisingly it was a CORS error... everything is _always_ a CORS error. It works now!

(account switching) hm, good findings, I was just getting this up and running in R. I'll compare the details of these responses and should be able to alter the OIM response to work better. Some of those response headers don't need to be included in tilejson so I should clean that up anyway.

My eyes just glaze over and I hit enter until all of those purple screens are gone

We were lucky to get you at the beginning of this bowling Renaissance!

Perfect! Yes the Springfield georeferencing was done by a user who got in touch out of the blue and got to work on those maps, still not sure who that was!

Great! Please don't hesitate to reach out if you need something or find issues with the site. Would love to learn more about this book sometime!

Glad to hear it and thanks for trying it out! "therapeutic" needs to go on the testimonials page if I ever make one

In the meantime checkout this walkthrough that gives an overview of the process: about.oldinsurancemaps.net/walkthroughs..., and you can find other docs/info from there as well. Some of the docs are a bit out of date, but still close enough to get started. Don't hesitate to DM me with questions!

You bet. I've added the 1887 vol. 1 map here: oldinsurancemaps.net/map/sanborn0.... You can make an account then go to that page and click "Load documents". Just let it run for a bit (I'll check on it too later)

We wrote a bit about it here: forum.openhistoricalmap.org/t/street-wid...

@ktflghm.bsky.social I'd be happy to add any of these volumes to @oldinsurancemaps.net if you would like to georeference them there! The workflow is straight-forward and you can download any geotiffs or use XYZ/WMS tile layers in whatever software you want. www.loc.gov/collections/...

On @oldinsurancemaps.net oldinsurancemaps.net/san-francisc... you can find hundreds of georeferenced sheets from 1950, done by the team at the University of Richmond's Digital Scholarship lab.

Yeah it's a strange thing about the SF Sanborns, all the streets are drawn very wide. For the team working on these, I added a parcel boundary layer (used #pmtiles @protomaps.com !) so they could split each street and georeference each block individually, snapping to parcel lines.

Seems to me that our current dependence on GPS is fed by an expectation that we always go the quickest route and know exactly where we are at any moment. If that expectation were muted we would be happier to miss a turn sometimes and ask for directions when needed.