A ko is semi-stable if a long cycle occurs only with one or several single passes in it. An example is molasses ko.
RobertJasiek: Temporary discussion: Molasses ko is a symmetric example; there are also asymmetric semi-stable shapes. The "superclass" is on the same level as the superclasses "fighting ko" and "disturbing ko"; i.e., "semi-stable ko" is a necessary and essential, although rare superclass. Overlooking that class let Ing create his wrong classification of kos into only fighting kos and disturbing kos, so that molasses ko could not be assigned to a superclass clearly. When players were exposed to Ing rules in the 1994 European Go Congress, the cutest players immediately noticed this application problem. In the meantime, a full theory for the ko types of basic kos is available.
Dieter: Okay then but I repeat what uberdude/hyperpapeterie has said before, namely that pages which do not really make an attempt to clarify something, at best referring to external pages, are at least doubtful parts of the library. They serve completeness but do not serve understanding. I'll leave this page as another ko-rules expert page. I hope that people who care about its existence will give it more reason for existence.
RobertJasiek: Suppose the page did not exist, but only the Ko Types page would exist. On that page, not every term can be defined, or that page would become unreadable. So this page here has justification already because it enables description of an otherwise undescribed (and essential) term!