I have 10 mil plated holes defined as vias, and some defined as pad in a footprint. In the drill file they appear in different rows:
I guess it can be sorted out easily with the fabricator, but is it a bug?
I see the same thing. I don’t see a reason to separate out a TH via drill vs a TH pad drill unless the via is a blind or buried one.
While something unconsolidated triggers my OCD too, I wouldn’t call this a bug but more of a nit. I doubt most fabricators even look at it. They just load it into the machine. The machine doesn’t have pedantic OCD.
If it’s classified as a bug, I’d vote it near the bottom of any bug list unless a dev is really bored.
In the Gerber specification, the .AperFunction attribute for drills takes different values, depending on whether they are ViaDrill or ComponentDrill. It looks like KiCad follows the Gerber recommendation of unique apertures for different functions.
Their narrative is that different compensation is needed at fabrication time for the different drill types. But, as original poster Alexy_Faynburd said, the fabricator will easily merge these drills if it can be done for their processes. Not so easy to find the 9 out 1193 holes which need handling differently.
My nature would be not to remove the information contained in these differences, and continue to have a conversation with the fabricators if there was a chance of ambiguity.
I wouldn’t spend much time on this. The fabricator usually recompiles the .drl files and rearrange the drill sequences for optimal use of their tool magazine.