> JavaMail is a poorly defined specification, which makes the role of the RI > more critical not less. It has also been around forever achieving de-facto > status and changing things runs the risk of making this implementation > unusable.
In fact much of the actual code makes up the API, and not the RI. It is very poorly designed, and part of it's poor design (too much code, too few interfaces) is also what makes it so hard for 3rd parties like us to make any alternative implementations. > That you have not compared behaviour to the RI for a spec that is known to > be ambiguous illustrates my concerns about your approach. All the unit > testing in the world won't identify a problem in how you expect > it to work. Unfortunately Sun haven't factored all the concrete classes in JavaMail into Interface and RI, the comparison will often not be with the behaviour of the RI, but with the behaviour of Sun's own API. d.
