Richard Beare

It turns out that the greater melbourne map is a surprisingly good example of the odd formats! The western edge of the bbox cuts through the Bellarine Peninsula, and that...

I've done some thinking about the unrolling functions - probably best to go back and do them with loops. I have the prototypes ready when you have integrated the other...

There is some very ancient work in these branches https://github.com/richardbeare/osmplotr/tree/SFExperiments https://github.com/richardbeare/osmplotr/tree/SFSeaLand There's a lot of work involved in reliably doing water/land shading, and the quickest way to get something reliable...

Would suggest keeping the small maps somewhere as test cases. I wasn't sure where to do that. They definitely don't look good in the vignette.

Good idea - I've done that.

I can't see a nice way of doing it in the current structure, although I'm not sure if I've found all the relevant parts of the code. I guess an...

I'll try to make time - I obviously forgot about it after it first came up.

I have tracked down what I believe to be the problem, and will attempt to generate tests to show it. The root of the issue is [here](https://github.com/InsightSoftwareConsortium/ITKMinimalPathExtraction/blob/2ff1ab58f30842ac57d76c8bc9462fcbc7523e67/include/itkSpeedFunctionToPathFilter.hxx#L94) The fast marching...

I have been trying to reproduce the behaviour in the existing tests and have lots of other problems. Going to take some work to sort it out. I now think...

Thinking further - it seems to me that there is no approach using the TargetOffset that will guarantee that all pixels in the immediate neighbourhood have the arrival function evaluated....