Closed StefanKarpinski closed 9 years ago
I was thinking about this, too. What do you think would be appropriate scope for that? Just global coordinate transforms (i.e., Geodesy.jl), or map projections as well (Mapping.jl)? Matlab has an all-encompassing Mapping Toolbox, but most of what's in there is outside my (borderline, at best) expertise. Is the Julia package philosophy more towards fewer, bigger packages, or lots of small packages with dependencies? Either way, I agree it would be good to put the coordinate transforms into a more fundamental package.
I would say that if two things mutually depend on each other then they should be one package. Slightly less clear-cut, but if there's no real reason to use one but not the other, then one package. In general, I would say that Julia tends to be less granular than Python and more granular than Matlab.
I'm about done splitting out existing coordinate functionality into Geodesy.jl
and tightening up the code. Is JuliaGeospatial
a good organization name? Should we wait to create an organization until the ecosystem is fleshed out a little more?
I'm all for more/better geospatial tools in Julia. We do have a mailing list for this kind of thing (julia-geo@googlegroups.com), but it hasn't seen much activity lately. It's not particularly well-advertised for one.
I haven't been working too much with geospatial stuff lately (my GDAL.jl has languished over the last few months), but if there is interest in beefing up the Julia geospatial infrastructure, I'll contribute where I can.
@garborg when you're ready, maybe post an announcement about Geodesy.jl
to the julia-geo mailing list?
@wkearn, I have been working with shapely
/fiona
/geopandas
in python, and am interested in the GDAL
package too, but have my hands busy with other packages. What's your take on first priorities for JuliaGeo[spatial], or your package on JuliaGIS?
@yeesian Good idea -- posted.
closed by #61. The discussion has moved over to the julia-geo mailing list.
Since physical global coordinates are such a fundamental thing and not really specific to OpenStreetMap, it may make sense to factor the coordinates code out into its own package. A good example of this sort of thing is the
Color
package, which implements a huge amount of colorimetry functionality: https://github.com/JuliaLang/Color.jl.