#408 Fill pattern generation in Magic has multiple issues. Issue created by RTimothyEdwards I am entering this issue so that it is recorded in the issue tracker and not just a pull request. The pattern generation parameters in the magic tech file has multiple problems: 1. The density for all metal layers and diffusion is too high, approximately 70% for most layers, while a more reasonable target is 50% or slightly below. 2. The resizing used to eliminate fill shapes that are clipped by keep-out areas is incorrect in a few cases, resulting in some fill shapes either non-rectangular or smaller than allowed. 3. The poly fill is not aligned with the diffusion fill, and in fact is being suppressed by the diffusion fill, resulting in too-low poly density. 4. The density fill script needs rewriting, to handle GDS files as well as .mag files, to automatically find the .magicrc startup script, and to make use of compressed GDS for the pattern fill files, among other things. 5. The density check script has incorrect calculations for the area as counted by whole tiles, having the original 100 count for a 10 x 10 stepped area, which was carried over from the original script used as the source for this one, whereas the steps were changed in this file to step by half the width of a tile, so the total tile area is 4 for a 2 x 2 stepped area. 6. The density check script has incorrect calculations for pro-rating the density of partial tiles on the right and top sides of a layout. These items are being addressed and will result in a pull request. IHP-GmbH/IHP-Open-PDK