<?xml version="1.0" encoding="windows-1252"?>
<node id="896493" title="Re: Proposal how to make modules using fork more portable" created="2011-03-30 14:14:14" updated="2011-03-30 14:14:14">
<type id="11">
note</type>
<author id="647953">
sundialsvc4</author>
<data>
<field name="doctext">
&lt;p&gt;
I am always of the opinion that threads and processes should terminate &lt;em&gt;themselves,&lt;/em&gt; and that they should, where possible, &amp;ldquo;stick around, waiting for more work to do,&amp;rdquo; instead of coming into existence to service one request and then flaming-out.
&lt;/p&gt;&lt;p&gt;
The &lt;tt&gt;kill(9)&lt;/tt&gt; protocol is &lt;em&gt;definitely&lt;/em&gt; one that was &amp;ldquo;taken from Unix,&amp;rdquo; and it really does not have a good parallel in the Windows &lt;i&gt;(i.e.&lt;/i&gt; DEC VAX 11/780...) world-view. &amp;nbsp; And yet, I have never recommended &amp;ldquo;management by the use of overwhelming deadly-force&amp;rdquo; &amp;nbsp; &lt;tt&gt;;-)&lt;/tt&gt; &amp;nbsp; anywhere.
&lt;/p&gt;&lt;p&gt;
Building process contexts, and tearing them down, are best thought of as expensive operations. &amp;nbsp; And, the maximum number of processes that may exist at one time is usually something that you want to govern.
&lt;/p&gt;</field>
<field name="root_node">
896360</field>
<field name="parent_node">
896360</field>
</data>
</node>
