<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
		>
<channel>
	<title>Comments on: Followup to &#8220;A Very Subtle Bug&#8221;</title>
	<atom:link href="http://blog.nelhage.com/2010/03/followup-to-a-very-subtle-bug/feed/" rel="self" type="application/rss+xml" />
	<link>http://blog.nelhage.com/2010/03/followup-to-a-very-subtle-bug/</link>
	<description>It's software. It's made of bugs.</description>
	<lastBuildDate>Fri, 03 Feb 2012 17:48:35 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=3.2.1</generator>
	<item>
		<title>By: lacos</title>
		<link>http://blog.nelhage.com/2010/03/followup-to-a-very-subtle-bug/comment-page-1/#comment-575</link>
		<dc:creator>lacos</dc:creator>
		<pubDate>Thu, 04 Mar 2010 15:18:25 +0000</pubDate>
		<guid isPermaLink="false">http://blog.nelhage.com/?p=160#comment-575</guid>
		<description>&lt;p&gt;Hi Nelson,&lt;/p&gt;

&lt;p&gt;thanks for the followup post. We continued the discussion with Paolo Bonzini in private, and in the end, I released a &lt;a href=&quot;http://freshmeat.net/projects/lbzip2/releases/313240&quot; rel=&quot;nofollow&quot;&gt;new version (0.23)&lt;/a&gt; of lbzip2.&lt;/p&gt;

&lt;p&gt;Before this release, lbzip2 ignored SIGPIPE, no matter what action it inherited, and an EPIPE write() condition was honored as any other fatal error, ie. with exit status 1. I still consider this way of handling EPIPE completely OK, but since I could implement process termination with SIGPIPE fairly cleanly for this case, I thought I&#039;d follow the &lt;strong&gt;default&lt;/strong&gt; behavior of traditional UNIX(R) utilities more closely, and also improve lbzip2&#039;s compatibility with GNU tar.&lt;/p&gt;

&lt;p&gt;Any user who wants the previous behavior can bequeath a SIG_IGN action to lbzip2 for SIGPIPE, as in&lt;/p&gt;

&lt;p&gt;&lt;code&gt;$ (trap &#039;&#039; SIGPIPE &amp;&amp; lbzip2 ...)&lt;/code&gt;&lt;/p&gt;
</description>
		<content:encoded><![CDATA[<p>Hi Nelson,</p>

<p>thanks for the followup post. We continued the discussion with Paolo Bonzini in private, and in the end, I released a <a href="http://freshmeat.net/projects/lbzip2/releases/313240" rel="nofollow">new version (0.23)</a> of lbzip2.</p>

<p>Before this release, lbzip2 ignored SIGPIPE, no matter what action it inherited, and an EPIPE write() condition was honored as any other fatal error, ie. with exit status 1. I still consider this way of handling EPIPE completely OK, but since I could implement process termination with SIGPIPE fairly cleanly for this case, I thought I&#8217;d follow the <strong>default</strong> behavior of traditional UNIX(R) utilities more closely, and also improve lbzip2&#8242;s compatibility with GNU tar.</p>

<p>Any user who wants the previous behavior can bequeath a SIG_IGN action to lbzip2 for SIGPIPE, as in</p>

<p><code>$ (trap '' SIGPIPE &amp;&amp; lbzip2 ...)</code></p>]]></content:encoded>
	</item>
</channel>
</rss>

