wenzelm [Sat, 08 Nov 2014 22:10:16 +0100] rev 58951
removed obsolete global-only options, which did not work out anyway (due to complexity of local_theory sandwich);
wenzelm [Sat, 08 Nov 2014 21:31:51 +0100] rev 58950
optional proof context for unify operations, for the sake of proper local options;
wenzelm [Sat, 08 Nov 2014 17:39:01 +0100] rev 58949
clarified name of Type.unified, to emphasize its connection to the "unify" family;
tuned low-level operation;
wenzelm [Sat, 08 Nov 2014 16:55:41 +0100] rev 58948
proper Envir.norm_type for result of Type.raw_unifys;
wenzelm [Sat, 08 Nov 2014 16:42:04 +0100] rev 58947
avoid slow metis proof;
wenzelm [Sat, 08 Nov 2014 16:35:24 +0100] rev 58946
proper Envir.norm_type for result of Unify.unifiers (amending 479832ff2d29 from 20 years ago);
wenzelm [Sat, 08 Nov 2014 15:45:00 +0100] rev 58945
tuned;
wenzelm [Sat, 08 Nov 2014 15:44:41 +0100] rev 58944
updated some sledgehammer proofs -- much faster;
wenzelm [Sat, 08 Nov 2014 15:40:29 +0100] rev 58943
updated sledgehammer proof after breakdown of metis (exception Type.TUNIFY);
wenzelm [Sat, 08 Nov 2014 15:01:05 +0100] rev 58942
recovered type matching, which was broken in 8a765db7e0f8 (see also 8a765db7e0f8, 2db1d3d2ed54);
NB: "match" operates on direct substitution without variable chasing, in contrast to "unify" (and Unify.matches!) which work on cascaded env;