Dane Springmeyer

My guess is this will go away if I stop building against ancient mapnik 2.0. So, todo is to only build against more recent packages.

http://dbsgeo.com/latlon/

okay, removed those files, can you try again? You may need to clone a refresh repo to avoid conflicts now in history.

very cool, thanks, seems worthwhile to merge into a separate branch, perhaps `python`, to live as a working example? I'm confused however - how you are triggering a segfault exactly?

Ah, maybe it needs to be `[FASTBUILD=0]` if deps changed? Will try that now...

@BergWerkGIS thank you. For future reference can you add docs for how the binary deps package works? I'd like to ensure we have docs in the readme to answer questions...

@BergWerkGIS Thank you. Going to re-open with one more question. I noticed mapnik master appveyor was also failing with this error `fatal error LNK1104: cannot open file 'libboost_filesystem-vc140-mt-1_61.lib'`. I think...

This looks to me that you are pulling Mapnik master, which requires C++14 features which likely don't work with visual studio 2015. I would recommend sticking with Mapnik 3.0.x rather...

> There must be exactly one source file present on the command line. I think that is fine. The `mapnik-gyp` build only tries to build one .cpp per compile line.

This is the configure output on unix for harfbuzz: ``` Build configuration: Unicode callbacks (you want at least one): Glib: false ICU: false UCDN: true Font callbacks (the more the...