<page xmlns="http://projectmallard.org/1.0/"
      xmlns:its="http://www.w3.org/2005/11/its"
      type="topic"
      id="introduction">

  <info>
    <link type="guide" xref="index"/>
<!--
    <link type="guide" xref="index#features"/>
    <link type="seealso" xref="anotherpageid"/>
-->
    <revision pkgversion="3.20" date="2016-04-04" status="stub" />

    <credit type="author">
      <name>Christian Hergert</name>
      <email its:translate="no">[email protected]</email>
      <years>2016</years>
    </credit>
 
    <include href="legal.xml" xmlns="http://www.w3.org/2001/XInclude"/>

    <desc>Welcome to <app>Sysprof</app>!</desc>
  </info>

  <title>Introduction</title>

  <p><app>Sysprof</app> is a system profiler for Linux that targets
  the GNOME desktop.</p>

  <section id="what-is-a-profiler">
    <info>
      <link type="guide" xref="index"/>
      <desc>Differences between tracing and sampling</desc>
    </info>
    <title>What is a Profiler?</title>

    <p>A profiler is an application that records information about an
application or system while it runs. That information can be explored to
gain insight into how the application could be changed to perform
better.</p>

    <p>Two common categories of software profilers exist, commonly referred
to as either tracing or sampling profilers. What is meant by tracing
profiler is that every function call executed by the program is known to the
profiler. A sampling profiler works by inspecting the state of the
program on a regular frequency and therefore does not see every function
call executed by the program.</p>

    <p>Both tracing and sampling profilers have their advantages. A
notable advantage of a sampling profiler is that the overhead is much
less than that of a tracing profiler, making it easier to use for
software that requires interactivity.</p>

    <p><app>Sysprof</app> is a sampling profiler.</p>
  </section>

</page>