more details on several raster formats. More todos

Author: Luke Zeitlin

src/gis/roadmap.clj

@@ -1,7 +1,7 @@
 ^{:kindly/hide-code true
   :clay             {:title  "GIS roadmap"
                      :quarto {:author   :luke-zeitlin
-                              :type     :collaboration
+                              :type     :post
                               :date     "2025-10-23"
                               :draft    true
                               :category :gis
@@ -15,23 +15,63 @@
 ;; At present this document can serve as a location to add notes, todos and thoughts.
 ;; Some future version of it may serve as a landing page.
 ;; # Why Clojure?
-;; A few words about why clojure is good for GIS...
+;; ## TODO:
+;; A few words about why Clojure is good for GIS...
 ;; -  Clojure's ability to interact with multiple host environments (Java, JS, Python, etc).
 ;; - Same language on the front and backend
 ;; - Ergonomic data workflow
 ;; - Interactive development
 ;; # GIS Task Categories 
 ;; ## Image formats
 ;; ### Raster
-;; #### GeoTIFF
-;; A guide for interacting with this format. Read, write, display.
+;; #### Cloud Optimized GeoTIFF (COG)
+;; ##### What is a COG?
+;; **TIFF** is an extensible image file format. **GeoTIFFs** are TIFFs with geospatial metadata
+;; such as coordinate reference / projection information, Geographic extent,
+;; pixel resolution and so on.
+;; **Cloud Optimized GeoTIFFs (COGs)** are geoTIFFs arranged for efficient access over HTTP. This
+;; is achieved with tiled layouts allowing for partial access (using HTTP range requests),
+;; and resolution pyramids. This allows for fetching only the portion of the COG required.
+;;
+;; More details on the format at: [cogeo.org](https://www.cogeo.org)
+;;
+;; #### How to interact with COGs with Clojure
+;; Many libraries for creation and manipulation of GeoTIFF data in other languages rely on
+;; the [GDAL binary](https://gdal.org). 
+;; This is performant and battle-tested but has a large footprint and is vulnerable to
+;; version mismatches between the client language and GDAL.
+;;
+;; Some pure language implementations (that do not depend on GDAL) exist, such as
+;; [geotiff.js](https://geotiffjs.github.io/) for JavaScript and [GeoTools](https://geotools.org/) in Java.
+;; For the sake of
+;; simplicity we recommend using interop with pure language GeoTIFF library where possible.
+;;
+;; ##### TODOs:
+;; - GeoTools / Java interop example
+;; - geotiff.js / JavaScript interop example 
+;; - Example of a containerized Python bindings with GDAL setup.
+;;
 ;; 
-;; - Interop Java ([GeoTools example](https://gist.github.com/the80srobot/3042990))
-;; - Python bindings (rasterio)
-;; - Thought: pathway to a pure clojure library for geotiffs (whithout GDAL dep)?
-;; - display / frontend (openlayers, leaflet, other)
 ;; #### NetCDF
+;; NetCDF is a format for metadata enhanced, compressed, multidimensional arrays.
+;; Often used for weather, climate, remote sensing and
+;; oceanographic data.
+;; Both [GeoTools](https://geotools.org/) and [NetCDF-Java](https://docs.unidata.ucar.edu/netcdf-java) are acceptable ways to to handle and create
+;; netCDF files in Java-hosted Clojure. In JS we can read NetCDF files with
+;; [netcdfjs](https://github.com/cheminfo/netcdfjs)
+;; ##### TODOs:
+;; - GeoTools interop example
+;; - NetCDF-Java example
+;; - Which one is best?
+;; - netcdfjs example (maybe interactive with scittle/react/clay)
+;;
 ;; #### HDF
+;; Another widely used multidimensional array format in the scientific GIS world.
+;;
+;; [HDF in clojure](https://scicloj.github.io/clojure-data-scrapbook/projects/data-formats/hdf/index.html)
+;; #### Zarr
+;; ##### TODO:
+;; - how to read in Clojure
 ;; 
 ;; ### Vector
 ;; #### GeoJSON
@@ -43,15 +83,21 @@
 ;; #### Notes:
 ;; Maybe [Factual/geo](https://github.com/Factual/geo) is good for some of this?
 ;;
-;; ## Serverless
+;; ## Cloud Optimized
 ;; ### COG
 ;; See GeoTIFF
+;; ### ZARR
 ;; ### PMTiles
 ;; - Reading: Timeverse PMtiles (Java interop)
 ;; - Creation: CLI / Babashka?
 ;; ### Notes:
 ;; [The Cloud Native Geo guide](https://guide.cloudnativegeo.org/) may be a useful starting
-;; pount for some of this
+;; pount for some of this.
+;;
+^:kindly/hide-code
+(kind/image {:src "cogeo-formats-table.png"
+             :style {:width "30em"}})
+;; *Image from https:/guide.cloudnativegeo.org* 
 ;;
 ;; ## Image processing
 ;; Routing / Network analysis
@@ -65,7 +111,7 @@
 ;;
 ;; ## Spatial reference systems
 ;; ### Coordinate system conversions
-;; - [Coordinate Systems article](https://mgimond.github.io/Spatial/chp09_0.html) maybe a good starting point for a clojure oriented article on the same.
+;; - [Coordinate Systems article](https://mgimond.github.io/Spatial/chp09_0.html) maybe a good starting point for a Clojure oriented article on the same.
 
 ;; ### H3
 ;; - again, [Factual/geo](https://github.com/factual/geo) may be a good starting point.