<?xml version="1.0" encoding="utf-8"?>
<feed xml:lang="en" xmlns="http://www.w3.org/2005/Atom"><title>Recent changes to 4: JIComTransport fix for file descriptor leak</title><link href="https://sourceforge.net/p/j-interop/patches/4/" rel="alternate"/><link href="https://sourceforge.net/p/j-interop/patches/4/feed.atom" rel="self"/><id>https://sourceforge.net/p/j-interop/patches/4/</id><updated>2010-08-06T19:03:34Z</updated><subtitle>Recent changes to 4: JIComTransport fix for file descriptor leak</subtitle><entry><title>JIComTransport fix for file descriptor leak</title><link href="https://sourceforge.net/p/j-interop/patches/4/" rel="alternate"/><published>2010-08-06T19:03:34Z</published><updated>2010-08-06T19:03:34Z</updated><author><name>Robert Clark</name><uri>https://sourceforge.net/u/finiteautomaton/</uri></author><id>https://sourceforge.net42f54c3336717416991f8211bf987c74c81eb8f2</id><summary type="html">I'm attaching a modified version of the JIComTransport class \(as well as a patch of the same\) that fixes the file descriptor leak we're seeing when blocking reads are down without specifically creating a Selector. As far as I can determine, the external
behavior of the class is unchanged.

I've testing this by running a new WMI request every 5 seconds in a new thread for 100 iterations. The file descriptor count has been constant, aside from a temporary increase when the read is actually occurring.</summary></entry></feed>