Progress notes from the team, roughly in order. Not proofread, not curated — just whoever had five minutes to post something.
We're in. Team is Theo (me), Priya, Jordan, Sam, and Elena — five people, 48 hours, one enormous abandoned building. This year's HeritageHack theme is historical preservation, and Elena convinced us Ashworth Hollow Sanatorium was the right pick: closed since 1974, never properly documented, and apparently there are still boxes of paper blueprints sitting in the county archive that nobody's scanned. Plan is: Elena and Priya handle blueprint sourcing and drone scanning this week, Sam builds the backend for storing and processing the scan data, Jordan does the site (sorry in advance, Jordan), and I keep everyone on schedule and make sure we actually have something to demo. Kickoff meeting went long because we spent twenty minutes arguing about whether it's "sanatorium" or "sanitarium." For the record: sanatorium. Moving on.
Got out to the site this weekend with the drone and a LIDAR rig borrowed from the engineering department (thanks to whoever left that unlocked, we owe you one). Exterior scans for Pavilions 1 through 3 came out clean — a little noisy around the collapsed sections of roof, but nothing the cleanup pass can't handle. Cross-referencing against the 1908 and 1952 blueprint sets now to line up the mesh with the original floor plans. Numbers so far: roughly 180,000 to 220,000 vertices per pavilion depending on complexity, which is more than I expected for a building this size. Next up: interior scans, assuming the county lets us back in before the weather turns.
Hi, I'm the person who built this website, and I would like to formally apologize for the About page. I don't know what's wrong with the CSS grid on that page and at this point I'm scared to find out. Every time I fix one thing, two more things break, like some kind of layout hydra. Anyway! Site's up, nav works, forms don't actually do anything yet (sorry), and I've been told the gallery will eventually have real images instead of gray boxes, which, fingers crossed. If anyone wants to help me fight the CSS grid, my door (inbox) is open. Bring snacks.
Small hiccup with the Pavilion 4 scan. The mesh processor keeps throwing out-of-memory errors partway through, which is strange because Pavilion 4 isn't actually bigger than the others — if anything it should be the simplest one, it was the last wing added and it's mostly just corridors according to the blueprints. Might be a corrupted scan file, might just be a bug in the batch script. Going to re-run it from scratch this week. Not worried yet, but if anyone's dealt with weird memory spikes in a photogrammetry pipeline before, I'm all ears.
set up a cron job last week to clear temp render files every night so we don't run out of disk space before the demo. simple script. shouldn't be interesting enough to write a blog post about. and yet. the job keeps running outside its schedule, sometimes multiple times an hour, and every time it does it seems to regenerate a version of the pavilion 4 mesh instead of deleting anything. checked the crontab three times. it says once a day. it is not running once a day. going to rewrite it from scratch instead of debugging priya's science project further. will report back.
Okay, this might be my favorite day of this entire project. Spent the afternoon at the county historical society archive and they finally let me see the full 1931 expansion records — the ones that were supposedly "in processing" for like six months. Turns out there's a complete survey from the original 1931 expansion, done by a surveyor named E. Whitcombe, with hand-drawn floor plans for the whole new wing. It's beautiful, honestly, and in way better condition than I expected for something 93 years old. Going to spend the weekend cross-referencing it against Priya's scans. This is exactly the kind of primary source we needed to make the model actually accurate instead of just a nice-looking guess. More soon.
Taking the site down from active updates for a bit while we sort some things out. Nothing wrong with the project, we're still on track for the demo. Just need some time to go through the new material properly before we—