2016-10-14  Hendrik Tews  <hendrik@askra.de>

	* add test for shared block with different types

2016-10-14  Hendrik Tews  <hendrik@askra.de>

	* added old safeUnmarshal comparison plots (can't find
	  the data any more)

2016-10-14  Hendrik Tews  <hendrik@askra.de>

	* copy web page into repo

2016-10-14  Hendrik Tews  <hendrik@askra.de>

	* update to camlp5 and OCaml 4.03
	* new test for empty array
	* several new tests for a non-regular recursive type

2007-04-19  Hendrik Tews  <tews@tandem.cs.ru.nl>

        * simplified internal flag list parsing
        * added new Verbose_type_ids flag to output just the type ids
        * generate and remember unique block id's and print them for with
          Verbose_blocks

2006-12-19  Hendrik Tews  <tews@tandem.cs.ru.nl>

        * finish docs, readme
        * add license
        * release polishing

        * clean for release
        * rename into memcheck

2006-12-18  Hendrik Tews  <tews@tandem.cs.ru.nl>

        * trace feature added

2006-12-17  Hendrik Tews  <tews@tandem.cs.ru.nl>

        * modification flags for check
        * flag parsing
        * diversified messages

2006-12-12  Hendrik Tews  <tews@localhost.localdomain>

        * finished walk through messages
        * seems to work

2006-12-12  Hendrik Tews  <tews@debian>

        * added physical equality hash
        * introduced type-check state
        * unified array block and variants/tuples
        * added some verbosity

2006-12-12  hendrik Tews  <tews@tandem.cs.ru.nl>

        * lots of changes

2006-12-06  Hendrik Tews  <tews@tandem.cs.ru.nl>

        * substitution, head normal form

        * ocaml_type type and fixing in heapcheck.ml
        * got elsa type descriptions generated/compiled/fixed

2006-12-06  Hendrik Tews  <tews@debian>

        * added records, variants, type constructors to generate

2006-12-05  Hendrik Tews  <tews@tandem.cs.ru.nl>

        * start project with third approach: Generate literate type
        descriptions with camlp4 and use relatively heavy checking
        functions. Types and type constructors must have unique
        identifiers. For type constructors I use the filename together
        with the location (don't know how that will work with functors).
        Ids for types will be generated when checking the type.

