Return-Path: Delivered-To: apmail-jakarta-ant-user-archive@apache.org Received: (qmail 14885 invoked from network); 8 May 2002 20:50:54 -0000 Received: from unknown (HELO nagoya.betaversion.org) (192.18.49.131) by daedalus.apache.org with SMTP; 8 May 2002 20:50:54 -0000 Received: (qmail 7780 invoked by uid 97); 8 May 2002 20:50:11 -0000 Delivered-To: qmlist-jakarta-archive-ant-user@jakarta.apache.org Received: (qmail 7691 invoked by uid 97); 8 May 2002 20:50:11 -0000 Mailing-List: contact ant-user-help@jakarta.apache.org; run by ezmlm Precedence: bulk List-Unsubscribe: List-Subscribe: List-Help: List-Post: List-Id: "Ant Users List" Reply-To: "Ant Users List" Delivered-To: mailing list ant-user@jakarta.apache.org Received: (qmail 7567 invoked by uid 98); 8 May 2002 20:50:10 -0000 X-Antivirus: nagoya (v4198 created Apr 24 2002) Message-ID: <6E78A44566D3FF429E7B08BC305A8B321DE9B0@DYN-EXCH-001.dynamicsoft.com> From: Kevin Grey To: "'Ant Users List'" Subject: RE: setting environment variables using ant Date: Wed, 8 May 2002 16:49:50 -0400 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2653.19) Content-Type: text/plain X-Spam-Rating: daedalus.apache.org 1.6.2 0/1000/N X-Spam-Rating: daedalus.apache.org 1.6.2 0/1000/N The same occurs for JDK 1.3.1. I implemented a pretty trivial mechanism to "start"/"kill" processes. It would execute the process with an associated key name, and hold the Process object in a Static Map. The script continues as normal until you tell it to die using the associated key. It doesn't always work that way though, so I added "waitfor" functionality which would wait for the process to die. Using RMI I'd tell the process to exit, then do a "waitfor" to wait for it to actually exit. In this case the process is our AppServer which we wrote, so I can implement/utilize RMI to make the process exit. It'd be really nice if it were possible to forcefully kill a process reliably... ~~K -----Original Message----- From: Hal Hildebrand (web) [mailto:Hal.Hildebrand@hellblazer.com] Sent: Wednesday, May 08, 2002 4:31 PM To: Ant Users List Subject: Re: setting environment variables using ant The start and forget aspect is a problem with the Java runtime exec, not Ant. It's a child process issue in that the parent process death kills all the children Strangely, under JDK 1.4 - at least under WinXP/2000 - I find that runtime exec *does* spawn new processes that don't die when the parent Java process dies. This is extremely annoying in the opposite way. I launch a bunch of parallel tasks for testing, and if I ^C the parent Ant process, they stick around. I'm not sure of the "fire and forget" feature of JDK 1.4 - it may only be a Windows aberration. But it's a welcome aberration. So if this *is* consistent and supported behavior across all platforms, then it's trivial to make a "fire and forget" execute task. ----- Original Message ----- From: "Scott Ganyo" > > classpath it should use. But then again, ANT is not an > > application launcher > > either, although it can do it (hangs around, and filters the > > output though, > > since the 'start and forget' capability of doesn't > > seem to make it > > into ANT)... > > This is a real shame, too, since I do use Ant to launch applications. (The > reason I need to do that is that Ant is the thing that knows how to create > the correct classpath for the applications that it builds for me.) > > Scott > -- To unsubscribe, e-mail: For additional commands, e-mail: -- To unsubscribe, e-mail: For additional commands, e-mail: