Garbage collection is non-deterministic - the JVM gets to decide when to
run it. My guess would be that ANT has a reference to ALL the files and
doesn't let go of any of them until all the tasks have run. You can
probably get around this with a little re-coding; I could be wrong, but
I think Ant-Call (and its variants) might cause the task objects to be
derefernced for every loop, meaning a little more CPU time, but
hopefully no oyt-of-memory errors. You might also look at either the
ANT scriping task, or the Ant-contrib (from sourceforge) for-each tag.
Dave Bartmess wrote:
Does the garbage collection have a "time-out" it's based on, or is it
strictly the lack of references to the classes?
I'm seeing problems with memory errors when doing a series of tasks on a
set of files, and running out of memory. But if I run the tasks on the
files one at a time manually (through ant still, but "written out"
instead of a single call to do all), it seems to work...
Thanks
--
Robert r. Sanders
Chief Technologist
iPOV
(334) 821-5412
www.ipov.net
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
