Sascha Brawer <[EMAIL PROTECTED]> writes: > However, we'd need to do a lot more for this to become useful. Maybe this > is obvious to everyone on the list, but in case it isn't, let me outline > what would be needed for CORBA support: > > 1. Use an IDL-to-Java compiler tool (like Sun's idlj) to generate > Java code for standard IDL files provided by the OMG. For example, > package org.omg.IOP is derived from IDL, and not directly available > at ftp.omg.org/pub/ javartf/. > > Having done this, we'd have roughly the same as is documented on > the Javadoc pages for Sun's J2SE 1.4. AFAIK, this should be enough > to run an ORB on a Classpath-based JRE.
That would be good. > 2. Provide an actual implementation for all those interfaces and > abstract classes. I didn't think the JDK actually provided an ORB. To my knowledge everyone goes to a vendor for an ORB implementation. My lack of actually using CORBA from Java is showing here. > 3. Write a GNU version of an IDL-to-Java compiler (i.e., a replacement > for idlj). More tools always welcome in the classpath-tools subproject. > (1) without (2) is somewhat useful: users could run CORBA applications if > they license an ORB from someone. Also, (1) is a prerequisite for > supporting CORBA in Classpath. But the big amount of work is (2). > > By the way, could we legally use Sun's idlj for (1), or do we have to do > (3) first? Yes, you can use idlj or javah as needed. -- Brian Jones <[EMAIL PROTECTED]> _______________________________________________ Classpath mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/classpath
