home *** CD-ROM | disk | FTP | other *** search
- Submitted-by: bagwill@swe.ncsl.nist.gov (Bob Bagwill)
-
- Jeffrey Kegler (jeffrey@netcom.com) wrote:
- >Submitted-by: jeffrey@netcom.com (Jeffrey Kegler)
- >Here's the problem. In the absence of a test method standard, the
- >implementor of the base standard must come as close as possible to the
- >original standard.
-
- In the absence of a test method (and, presumably test assertions)
- what does it mean to "come as close as possible to the original
- standard"? How does an independent implementor measure their
- closeness? IMHO, there would be three ways to ensure "closeness":
-
- 1) Don't implement. Buy the product from the original inventor.
-
- Of course, that implies:
-
- a) no competition, because there's only one "real" implementation
- b) no way of requiring any particular behavior,
- because the current version of the product *is* the standard.
- c) making sure the inventor stays in business, lives forever,
- and is perfectly fair and honest in all their licensing
- practices
-
- Of course, if the standard doesn't match an existing implementation,
- there *is* no product to buy that's close to the standard.
-
- 2) Implement, but use the inventor's test suite.
-
- That implies:
-
- a) the inventor has a test suite
- b) the inventor will test your implementation, or
- c) the inventor will sell or license the test suite fairly, honestly, etc.
- d) The inventor, a potential competitor, determines whether
- your implementation passes, directly or indirectly.
-
- In the past, inventors have reserved the right to decide:
-
- a) who can have their product tested
- b) who can buy or license the test suite
- c) how much testees pay
- d) if the inventor runs the tests
- e) if the implementor must provide source
- and last but not least,
- f) if a given implementation passes or fails
-
- Of course, an implementor always has recourse to our swift,
- inexpensive, and just tort system if they disagree
- with the inventor.
-
- 3) Implement using the paper standard,
- and use lawyers to "prove" that your implementation meets the standard.
-
- :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-)
-
- Obviously, these strategies are far superior to having openly specified
- test methods, test assertions, and test suites which may not
- accurately interpret the divinely-inspired prose of the paper
- standard.
-
- --
- Caveat lector, these represent my humble opinions, and none other.
-
- Bob Bagwill
- bagwill@swe.ncsl.nist.gov
-
-
- Volume-Number: Volume 30, Number 99
-
-