mousebites like OshPark, with holes touching sides.

[bootchk]

Prerequisites

• I have read the documentation and the proposed feature is not implemented.

Description

The request is to generate a number of drills in a mousebite that will fit in a tab.

This might put a drill very near the side of a tab.
But when you consider the round corners for milling on an inside corner,
this still produces a good tab:
with a high proportion of material drilled out of the mousebite.

Here is the Oshpark drawing of a mousebite.

Note the tab is 2.5mm wide and holes are 0.5mm diameter, with 1mm spacing.
So three holes fit in 2.5mm, if the holes touch the edge.
(From their website, width actually given as 0.1 inch, 0.02 inch drill, 0.04 inch spacing.)

Here is one fascimile I could generate with KiKit.
Here is KiKit tab width: 2.5mm, mousebite hole spacing: 1mm, mousebite prolong:0.5 .
I assume the KiKit algorithm generates two drills because that is all that would fit

Here is KiKit tab width: 2.5mm, mousebite hole spacing: 1mm, mousebite prolong:0.

It seems like the hole spacing is actually 1.25.

Note my boards are very small, say 5 mm wide.

This issue is low priority.
Note that OshPark will generate mousebites for a panel made by KiKit.
So one actually doesn't need to generate mousebites in KiKit to order from OshPark.
But you optionally can, to ensure that the mousebites are where you want them.

I could be wrong.
I tried other combinations, such as hole spacing 0.99 and 0.8, with no luck.
I haven't looked at the KiKit Python code.
Possibly its a floating point comparison issue.

Thanks, KiKit is great. I am a FOSS contributor to GIMP, especially plugin infrastructure.
The KiKit and KiCad plugin architecture is very nice.

[yaqwsx]

The code for generating mouse bites is relatively short and, I hope, self-explanatory:

KiKit/kikit/panelize.py

Lines 1531 to 1557 in a8d1c05

	def makeMouseBites(self, cuts, diameter, spacing, offset=fromMm(0.25),
	prolongation=fromMm(0.5)):
	"""
	Take a list of cuts and perform mouse bites. The cuts can be prolonged
	to
	"""
	bloatedSubstrate = prep(self.boardSubstrate.substrates.buffer(SHP_EPSILON))
	offsetCuts = []
	for cut in cuts:
	cut = cut.simplify(SHP_EPSILON) # Remove self-intersecting geometry
	cut = prolongCut(cut, prolongation)
	offsetCut = cut.parallel_offset(offset, "left")
	offsetCuts.append(offsetCut)
	for cut in listGeometries(shapely.ops.unary_union(offsetCuts).simplify(SHP_EPSILON)):
	self.renderedMousebiteCounter += 1
	length = cut.length
	count = int(length / spacing) + 1
	for i in range(count):
	if count == 1:
	hole = cut.interpolate(0.5, normalized=True)
	else:
	hole = cut.interpolate( i * length / (count - 1) )
	if bloatedSubstrate.intersects(hole):
	self.addNPTHole(toKiCADPoint((hole.x, hole.y)), diameter,
	ref=f"KiKit_MB_{self.renderedMousebiteCounter}_{i+1}",
	excludedFromPos=True)

.

The idea is:

• the cut line formed by the tab is simplified and prolonged by the specified amount,
• we place the holes evenly along the cut line (hence, the actual pitch might slightly deviate),
• a hole is created only when its center is placed on the board substrate. ←This is where I think the behavior your report comes from