1
<?xml version="1.0" encoding="utf-8"?>
3
<!ENTITY mdash "—">
5
<!-- $Id: changes-2.0.6-2.1.0.xml 4672 2010-04-19 10:26:23Z tfheen $ -->
6
<group from="2.0.6" to="2.1.0">
11
<para>The changelog in 2.1.0 included syntax errors, causing
12
the generated changelog to be empty.</para>
16
<para>The help text for <code>default_grace</code> was wrongly
17
formatted and included a syntax error. This has now been fixed.</para>
21
<para><code>varnishd</code> now closes the file descriptor used
22
to read the management secret file (from the <code>-S</code>
26
<change type="bug" ref="669">
27
<para>The child would previously try to close every valid file
28
descriptor, something which could cause problems if the file
29
descriptor ulimit was set too high. We now keep track of all
30
the file descriptors we open and only close up to that number.
35
<para>ESI was partially broken in 2.1.0 due to a bug in the
36
rollback of session workspace. This has been fixed.</para>
40
<para>Reject the <code>auth</code>command rather than crash if
41
there is no <code>-S</code> parameter given.</para>
45
<para>Align pointers in allocated objects. This will in theory
46
make Varnish a tiny bit faster at the expense of slightly more
51
<para>Ensure the master process process id is updated in the
52
shared memory log file after we go into the background.</para>
55
<change type="bug" id="679">
56
<para><code>HEAD</code> requests would be converted to GET
57
requests too early, which affected <code>pass</code>
58
and <code>pipe</code>. This has been fixed.</para>
62
<para>Update the documentation to point out that the TTL is no
63
longer taken into account to decide whether an object is
64
cacheable or not.</para>
68
<para>Add support for completely obliterating an object and all
69
variants of it. Currently, this has to be done using inline C.</para>
73
<para>Add experimental support for the <code>Range</code>
74
header. This has to be enabled using the parameter
75
http_range_support.</para>
78
<change type="bug" id="684">
79
<para>The <code>critbit</code> hasher could get into a deadlock
80
and had a race condition. Both those have now been fixed.</para>
85
<name>varnishsizes</name>
88
<para><code>varnishsizes</code>, which is
89
like <code>varnishhost</code>, but for the length of objects,
90
has been added..</para>