I believe some type 1 conflicts are currently missed, even if #26 is remedied, because of this line which is currently while l < l_1, when it should, as far as I know, actually be while l <= l_1, since otherwise the final node in the next layer will never be processed and thus any conflicts arising from its incoming edges will remain unmarked.
This would also agree with the pseudocode from this thesis on page 33, which uses zero-based indexing, unlike the original Brandes and Köpf paper, but still uses while l <= l_1.
I believe some type 1 conflicts are currently missed, even if #26 is remedied, because of this line which is currently
while l < l_1, when it should, as far as I know, actually bewhile l <= l_1, since otherwise the final node in the next layer will never be processed and thus any conflicts arising from its incoming edges will remain unmarked.This would also agree with the pseudocode from this thesis on page 33, which uses zero-based indexing, unlike the original Brandes and Köpf paper, but still uses
while l <= l_1.