I rote a script to automatically convert a bunch of equirectangular projections to dymaxion projections. I had to refactor the other main apps slitely, because they were structured very poorly.
Why did I ever think this was something people wanted? By default, the raster maps should look the same as the vector maps and their analysis. Also pseudocylindrical pole-point projections look really terrible by this default (though that's kind of their fault for being so bad).
That's read /oʊ ˈɡʊdi/.
I added the Goode Homolosine projection. I was reluctant; its complexity to quality ratio is just a little too high. But I need it for the paper I'm writing. So I may as well add it to the program.
I've realized that my edge breaking system is way overzealous, and I curbed it somewhat. I think it's probably easier for users to break excess edges than to have to replace incorrectly broken ones.
I added a distortion comparison for a map projection I invented that I actually like, played around with the political map, renamed Hammer to not mention Aitoff since that can be confusing to people who can't tell the difference, removed Natural Earth because why does Tom Patterson keep publishing these they don't bring anything new to the table, and I removed some minor artifacts from Sillouette.png.
All I wanted was to hide the category labels for every other bar in my bar chart. It does that automatically when it gets resized, and it had been annoyingly irregularly hiding some of my labels already. Is there an option for the threshold at which that happens, or to change the manner in which it chooses which ones to hide? No. Is there a list of Labels for the ticks that I can access and manually edit? No. I'll just set every other x value to the empy string, then. JavaFX takes it upon itself to sum the y values of all bars with matching x values and combine them into one bar.
I finally had to just make the chart smaller to trigger the automatic reformatting. Lord Ruler, I hate this.
I just can't justify using this much computing power on an initial guess
for a bisection scheme. For Newton-Raphon it might have made sense. I'm
really not sure why that didn't converge.
I fixed up American polyconic, and I'm kind of glad I have it, even though it's a terrible map projection whose only application is making gores and Cassini is better for that, anyway.
Now that I'm out of my parameter optimization phase (heh), I've changed the Tobler hyperelliptical projection to be more like he actually published them. Also, I guess the graticule is broken again.
I actually fixed the graticule. Mostly. It still has a few issues, but overall, it's working great. There are a few projections with which it's not working. And also, apparently Albers is just broken.
I got a copy of Kartographische Netzenwuerfe and was able to look up the original equations for Wagner's projections (since I couldn't find a reliable source for them anywhere else). So now they are correct. And a bit simpler, thanks to Wagner's tables.
I redid everything, realizing that I could just encode _everything_ in the octant table. Why didn't I just do that before? Why did I keep trying to write code?
