> [Original Message] > From: Peter Edworthy <[EMAIL PROTECTED]> > To: <[email protected]> > Date: 6/10/05 8:23:48 AM > Subject: Re: [arch] VM/Classlibrary interface > ...snip... >
> I don't like changes in Java.Lang, although I know they should be > invisible it just doesn't feel like a very clean interface design, > overlaying extra functionality without extending the class or interface. > This could, of cause, all be prejudice and hence having to now go off and > judge what benefits it gives over a more clear separation. > No, your observations are not AT ALL prejudical. This lack of a "clean interface design " is part of what I have been wrestling with in my mind about how to build a JVM below the foundational class of java.lang.Object and friends (namely, Class, System, etc.) Some various other comments on the list have also hinted at this issue, as well as some even suggesting that various class libraries could be used on top of the JVM. Ultimtely, the discussion will devolve back down to GNU Classpath, as Geir has kept prodding us all in that direction (starting with the May announcement and concommitant FAQ), and it correctly avoids licensing problems with other vendors' class libraries. Therefore, concerning java.lang.* corporately, one of the JVM issues is going to be a proper relationship between the JVM and java.lang, partially due to its inherent functionality, partly due to what parts of it are implemented as native. A closely related issue will therefore also be how the JVM and java.lang work with JNI. Just for the to-do list. Dan Lydick
