RE: [ve-dev] Re: [Design] Init string parsing

["Ed Burnette" <Ed.Burnette@xxxxxxx>]

Title: Message

Rich, this seems well thought out but that first part concerned me when I read it. You say:

"The parse tree will be modeled after the AST node tree. It will be in EMF. This way it can be modeled and serialized out. It won't be an exact duplicate of the AST structure because there are far too many nodes in AST for what we need. We will take the necessary handful, plus some extensions, and what we use in the VE. "

If you look at JDT discussions in the past several months, you'll find a lot of people asking about DOM vs. AST and fretting about the overlap. And yet now you are adding another representation of source code. Is there no other way?