Mathematicians say they have finally built a solution to the graph sandwich problem, a stubborn puzzle that has resisted answers for decades.
The graph sandwich problem asks whether a graph that fits between two given graphs, one setting the required edges and one setting the allowed edges, can satisfy some target property. Researchers have used this framework since the 1990s to sort graph-recognition problems into complexity classes, and several specific versions have stayed open despite repeated attempts. The new report says one of those holdout cases has now been resolved, but it offers little on who did the work, what property was involved, or how the proof holds up. That is a thin foundation for a headline this confident.
If the result is real, it matters because graph sandwich problems are not just an academic curiosity. They show up in scheduling, computational biology, and network design whenever you need to know if a structure exists between two boundaries. Closing a known-hard instance would hand computer scientists a cleaner map of what is actually solvable in that space.
Until more details surface, "finally" is carrying more weight than the evidence can support.