From: Lennart Poettering <lennart@poettering.net>
Date: Mon, 9 Jul 2012 14:04:21 +0000 (+0200)
Subject: man: explain why people should/shouldn't use sd_journal_print()
X-Git-Tag: v187~156
X-Git-Url: http://www.chiark.greenend.org.uk/ucgi/~ianmdlvl/git?p=elogind.git;a=commitdiff_plain;h=6aae0ed2d5915c28b27857fe0cf16e01b21a1882;hp=40b90434832c373c1394bc502523bee39c279c24;ds=sidebyside

man: explain why people should/shouldn't use sd_journal_print()
---

diff --git a/man/sd_journal_print.xml b/man/sd_journal_print.xml
index c40498b38..2d5d3639c 100644
--- a/man/sd_journal_print.xml
+++ b/man/sd_journal_print.xml
@@ -168,6 +168,23 @@ sd_journal_send("MESSAGE=Hello World, this is PID %lu!", (unsigned long) getpid(
                 not desired it can be turned off by defining
                 SD_JOURNAL_SUPPRESS_LOCATION before including
                 <filename>sd-journal.h</filename>.</para>
+
+                <para><citerefentry><refentrytitle>syslog</refentrytitle><manvolnum>3</manvolnum></citerefentry>
+                and and <function>sd_journal_print()</function> may
+                mostly be used interchangably
+                functionality-wise. However, note that log messages
+                logged via the former take a different path to the
+                journal server than the later, and hence global
+                chronological ordering between the two streams cannot
+                be guaranteed. Using
+                <function>sd_journal_print()</function> has the
+                benefit of logging source code line, file names, and
+                functions as meta data along all entries, and
+                guaranteeing chronological ordering with structured
+                log entries that are generated via
+                <function>sd_journal_send()</function>. Using
+                <function>syslog()</function> has the benefit of being
+                more portable.</para>
         </refsect1>
 
         <refsect1>