eliminated dead code;

parallel preparation of document variants, within separate directories;

Present.init/finish/no_document are not thread-safe -- eliminated futile CRITICAL sections; tuned;

eliminated redundant doc_prefix1; tuned;

renamed doc_prefix2 to dump_prefix;

tuned;

tuned terminology for document variants;

replaced File.check by specific File.check_file, File.check_dir; clarified File.full_path -- include parts of former Thy_Load.get_file; simplified Thy_Load.check_file -- do not read/digest yet; merged Thy_Load.check_thy/deps_thy -- always read text and parse header; clarified Thy_Header.read -- NB: partial Path.explode outside of args parser combinator; Thy_Info.check_deps etc.: File.read exactly once;

tuned;

preencode value of "need" selectors in Kodkod bounds as an optimization

ignore "need" axioms for "nat"-like types

added "simp:", "intro:", and "elim:" to "try" command

optimize Kodkod axioms further w.r.t. "need" option

optimize Kodkod bounds of nat-like datatypes

more optimizations of bounds for "need"

optimize Kodkod bounds when "need" is specified

always destroy constructor patterns, since this seems to be always useful

reintroduced "show_skolems" option -- useful when too many Skolems are displayed

reword Nitpick's wording concerning potential counterexamples

prevent an exception if "card" is empty (e.g., "nitpick [card]")

add option to function to keep trivial ATP formulas, needed for some experiments

add option to relevance filter's "all_facts" function to really get all facts (needed for some experiments)

tuned lemma

added lemmas

added lemma

support non-ground "need" values

recover Isabelle symlink for public distribution, notably website;

standardized headers;