+1 While I don't have a strong opinion on whether full constructors are implemented (we had them, now we don't, whatever :)), I definitely don't think they should be mandated.
--G ----- Original Message ----- From: "Sam Ruby" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Friday, June 14, 2002 11:50 AM Subject: Re: TCK issue: beans with full constructors? > Greg Truty wrote : > > > > *the Bean class must provide zero-argument constructors so it can be created ** > > using Beans.instantiate(),* > > Russell Butek wrote: > > > > The TCK requires generated beans to have full constructors. > > > Note: these two statements are not necessarily in conflict. Witness the > following code generated by the RI from > http://www.xmethods.net/idemo/wsdl/ISupplier.wsdl: > > public PO() { > } > > public PO(supplier.POHeader header, supplier.POLine[] lines) { > this.header = header; > this.lines = lines; > } > > I believe that the Apache, IBM, (and possibly Macromedia?) positions > should be simply that it is inappropriate for the RI and TCK to > implement a such a feature as it is not in the spec. While I agree with > Russell's original note, this isn't the forum to discuss the merits of > the feature - that's the domain of the expert group. Meanwhile, it is > not appropriate for the RI or TCK to be used as "back doors" to get > standardization of features upon which the expert group did not agree to. > > - Sam Ruby >
