I used KiCad-Diff to quickly identify changes in the PCB, as well as the normal Gerber viewer to diff between the changes and the unmodified project.
I could have squashed the three commits down to one, but my version of kiCad (5.1.5) on Linux seemed to shift the entire project (in .kicad_pcb form) by a small offset, making it difficult to analyze any changes (intentional and otherwise) made to the circuit (this could have been an artifact of the KiCad-Diff). The Gerber viewer showed no translation.
When plotting the Gerber and Drill files, I included ALL layers (even the empty ones) and used the following settings for Gerber:
Plot footprint values
Plot footprint references
Exclude PCB edge layer from other layers
Exclude pads from silkscreen
Use Protel filename extensions
And for the Drill files:
Excellion: minimal header
Excellion: Oval Holes: Use route command
Map File Format: PostScript
Drill Origin: Absolute
Drill Units: Millimeters
Zeroes Format: Suppress leading zeroes
I diffed in the Gerber viewer against commit 15dacd6; and verified that there were no unexpected changes. The KiCad-Diff did show up a small translation offset for the entire board, but only on the first change (whatever the change) to the board. No matter how many times I tried, and down to literally one operation and then saving; it always translated by the same small amount on the first save. I expect that there is some difference in the way 5.1.5 generates the .kicad_pcb file, OR KiCad-Diff (based on python pcbnew) itself had some issue. However, the offset did NOT affect the Gerber files, as no translation shows up there.
I used KiCad-Diff to quickly identify changes in the PCB, as well as the normal Gerber viewer to diff between the changes and the unmodified project.
I could have squashed the three commits down to one, but my version of kiCad (5.1.5) on Linux seemed to shift the entire project (in .kicad_pcb form) by a small offset, making it difficult to analyze any changes (intentional and otherwise) made to the circuit (this could have been an artifact of the KiCad-Diff). The Gerber viewer showed no translation.
When plotting the Gerber and Drill files, I included ALL layers (even the empty ones) and used the following settings for Gerber:
And for the Drill files:
I diffed in the Gerber viewer against commit 15dacd6; and verified that there were no unexpected changes. The KiCad-Diff did show up a small translation offset for the entire board, but only on the first change (whatever the change) to the board. No matter how many times I tried, and down to literally one operation and then saving; it always translated by the same small amount on the first save. I expect that there is some difference in the way 5.1.5 generates the
.kicad_pcbfile, OR KiCad-Diff (based on python pcbnew) itself had some issue. However, the offset did NOT affect the Gerber files, as no translation shows up there.