3 ms·
I'm definitely nowhere near 25 years. My perspective of "hard" LC problems is that they are typically distributed over a system anyways, where the optimal part
by scollet 5y ago
I'm definitely nowhere near 25 years.
My perspective of "hard" LC problems is that they are typically distributed over a system anyways, where the optimal parts make the whole. That, or they are domain-specific analogs, like complex solvers, or involve graduate graph theory, etc.
What is your perception? I haven't done LC in about a year and a half.