[jsr294-modularity-eg] Simple Module System Proposal

Stephen J. McConnell mcconnell at dpml.net
Wed Aug 19 13:20:44 EDT 2009


On Wed, 2009-08-19 at 12:02 -0400, BJ Hargrave wrote:

> So here the black holes are: syntax in the module-info source files is
> not specified in the JLS and is thus module system specific and, at
> compiler-time, a module system must be plugged into the compiler to
> compute the module membership of a type. 

BJ:

Thanks for the clarification. 

So, in effect - we are talking about one 'black-hole'. Could we could
actually refer to simply as an issue (after all over the top rhetoric
can sometimes derail substantive technical discussion)?

So - let's rephrase the issue:

  * in the current spec you are saying that there exists syntax 
    that is not specified in the JLS that is specific to a 
    module system implementation

and as such, 

  * these undefined information structures can only be computed 
    through the usage of a specific module system implementation 
    (noting that I have dropped the reference to 'a module system 
    plugged into the compiler' as that may be a separate and 
    distinct issue).

Is that a correct summation?

If it is, could you please provide references to the specific
specification documents that form you idea of the baseline. I just want
to confirm that you and I are actually working off the same script and
that as soon as we engage into specifics we are basically on the same
page.

Cheers, Steve.

-- 
Stephen J. McConnell
mailto:mcconnell at dpml.net
http://www.dpml.net
mobile: +61 4 5800 3980



More information about the jsr294-modularity-observer mailing list