With the switch in Ant 1.6 to UnknownElements for everything in the build,
dynamically resolved at runtime, it's now much more difficult to know if
a build is correct, i.e. unless one executes all the targets, one doesn't
know for sure that all tasks used can be correctly resolved.

It happened to me a few time already to break a build after updating it to
use XML namespaces and forgetting to put some tasks in the appropriate
namespace. Almost always, all my tasks/types can be statically resolved
given the special antlib: XML namespace declarations.

I would thus request a new mode/switch in Ant that would try to resolve
all tasks/elements in the build file, and report the ones which cannot
be resolved, in order to ascertain whether a build is 'Ant' correct.
I don't care whether the build is actually run or not, all I want is
a mode of execution similar to the old Ant ways where unresolved tasks,
even in targets not executed, can be successfully resolved.

Using that switch would allow me to stop breaking builds... Thanks, --DD

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

Reply via email to