Michael Sumner

Does multi-page mean (multi)subdataset analogously to multi-variable in NetCDF? I'm not familiar with the tiff spec, mostly only with GeoTIFF usage details and the abstractions "Pages" could be like bands,...

I'm getting the same error still as the OP ?? ```python import virtualizarr virtualizarr.__version__ '1.1.1.dev12+g9e7d430' from pathlib import Path import numpy as np from PIL import Image from virtualizarr import...

there's a multi tiff here: https://github.com/OSGeo/gdal/raw/refs/heads/master/autotest/gcore/data/twoimages.tif ```bash gdalinfo /vsicurl/https://github.com/OSGeo/gdal/raw/refs/heads/master/autotest/gcore/data/twoimages.tif Driver: GTiff/GeoTIFF Files: /vsicurl/https://github.com/OSGeo/gdal/raw/refs/heads/master/autotest/gcore/data/twoimages.tif Size is 20, 20 Image Structure Metadata: INTERLEAVE=BAND Subdatasets: SUBDATASET_1_NAME=GTIFF_DIR:1:/vsicurl/https://github.com/OSGeo/gdal/raw/refs/heads/master/autotest/gcore/data/twoimages.tif SUBDATASET_1_DESC=Page 1 (20P x 20L x 1B)...

I think that's fine, multiple chunks, I think that's how they're intended to be interpreted, not as different arrays.(It's very old style)

oh niice!! I wanted this kind of separation of the manifest, awesome

I have *never* seen a netcdf that uses bounds exclusively to specify a regular grid. that's cool, does it occur in the wild? can you point to a real example?...

also GDAL doesn't recognize this way of georeferencing, is it mentioned in CF anywhere? I've only ever seen bounds used to disambiguate the centre vs corner thing for (usually degenerate)...

my opinion, create new functions '_sf()', '_Spat()', '_wk()', maintain the old behaviour for as long as sf and sp work. New formats, new functions for when change inevitably comes again

I do think it's worth having some logic to pick a crs family for a given set of longlat data. Essentially anything is fine at Tasmania size (laea, aeqd, lcc,...

and because, I suppose, the first level (local) is to have approximately equal area and equal distance and conformality across the entire map, which is trivial as long as you...