Stefan Bodewig wrote:
On Wed, 27 Nov 2002, Conor MacNeill <[EMAIL PROTECTED]>
wrote:


What comes next is the notation to express the use of derived type,
and the underlying implementation in the core. In Mutant I advocated
an explicit statement of the substitution


This here?

<http://cvs.apache.org/viewcvs.cgi/*checkout*/jakarta-ant/proposal/mutant/docs/Attic/features.html?rev=1.1#Extensibility>

Yes.

Let me say at this juncture that Mutant is and will remain dead. I'm on an evolutionary tack now. Nevertheless there are some ideas I was able to explore and some refactorings which can be applied to Ant1 without breaking things.


So the example from the link above would become

    <copy todir="dest">
      <classfileset dir="../../bin/ant1compat/">
        <root classname="org.apache.tools.ant.Project"/>
      </classfileset>
    </copy>

yes? Sorry, haven't had the time to read the patch to know.

Yes, this is correct.


This one looks easier to write and read from a build file writer's perspective. What has been the reason to use a different notation in Mutant?

As you state below it is when the same type is used in multiple set/add methods that the above breaks down. Also the XSI based file would be able to be validated by a schema aware parser, I think (hope), whereas the above cannot be validated (which may not be a big deal, although some seem to like having the ability to do so.)




Using only one piece of information (the type name) has limits.


I think I remember it.  A problem arises when you have two different
nested elements (element names) that both accept the same class.  Say
you have <mypath> that is derived from Path, what would it be used for
in

<javac ...>
  <mypath .../>
</javac>

src, classpath, sourcepath, bootclasspath and extdirs would be
possible. Has this been the reason for your choice Conor?

Yes

Conor



--
To unsubscribe, e-mail:   <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>



Reply via email to