Conversation
|
@liam-o-marsh does it work for you right now? |
|
@briling inkscape 1.3.2 and it works. why? |
|
Weird... 2W ago I tried with Inkscape 1.3 (0e150ed, 2023-07-21) and got weird errors, now they're gone. |
| ''' | ||
|
|
||
| l = abs(y1-y0) | ||
| assert 0.5*h <= r0 , f"bad h parameter, should be less or equal to {2*r0}" |
There was a problem hiding this comment.
"bad" parameter actually works -- gives a weird looking bond but I thought it might be not useless
|
Finally found that for me even the old version breaks when the bubbles are of the same size, even though it's supposed to treat this case separately |
|
that's weird, it works on my side... both with the exact same size and with an epsilon difference. |
|
Haven't saved it then because I was in a hurry and now can't reproduce the bug! It happened with the second circle being a copy of the first one (probably in some specific circumstances)... |
2 participants
also, a weird edge-case was patched for borderless circles
I might need to run this version on a couple of inkscape versions
the second commit contains a few refactors, but you can also choose not to include it if you want