Because the typical use case spawn is to have the spawned process outlive the Ant process. Pumping the out/err streams of the process would prevent the Ant VM from exiting, no? I'm no specialist of System.exec.
If you can provide a patch that allows Ant to pump the streams while still alive, and doesn't prevent it from exiting normally as usual, we'll sure consider it. This stuff is usually platform-dependent and difficult to get right though. --DD On 6/23/06, Aaron Davies <[EMAIL PROTECTED]> wrote:
Well why not? I'm looking through the Execute code, and it looks like ant deliberately throws away the output of a spawned process. What's the design philosophy behind this? On 6/23/06, Dominique Devienne <[EMAIL PROTECTED]> wrote: > I don't think it is. Unless you spawn a shell that executes your > command, doing the redirection. --DD > > On 6/23/06, Aaron Davies <[EMAIL PROTECTED]> wrote: > > Is it possible to capture the stdout and stderr of a spawned Java > > process (i.e. a <java spawn="true" ... />)? It's rather hard to debug > > without that. > > -- > > Aaron Davies > > [EMAIL PROTECTED] > > > > --------------------------------------------------------------------- > > To unsubscribe, e-mail: [EMAIL PROTECTED] > > For additional commands, e-mail: [EMAIL PROTECTED] > > > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [EMAIL PROTECTED] > For additional commands, e-mail: [EMAIL PROTECTED] > > -- Aaron Davies [EMAIL PROTECTED] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
