1
<?xml version="1.0" encoding="utf-8"?>
2
<html xmlns="http://www.w3.org/1999/xhtml">
4
<title>Change log for Varnish 2.0.4</title>
5
<link rel="stylesheet" type="text/css" href="changes.css"/>
8
<h1>Change log for Varnish 2.0.4</h1>
9
<h2>Changes between 2.0.3 and 2.0.4</h2>
13
<p>Make Varnish more portable by pulling in fixes for
14
Solaris and NetBSD.</p>
17
<p>Correct description of -a in the manual page.</p>
20
<p>Ensure we are compiling in C99 mode.</p>
23
<p>If error was called with a null reason, we would crash on
24
Solaris. Make sure this no longer happens.</p>
27
<p>Varnish used to crash if you asked it to use a
28
non-existent waiter. This has now been fixed.</p>
31
<p>Add documentation to the default VCL explaining that
32
using <span class="code">Connection: close</span> in <span class="code">vcl_close</span>
33
is generally a good idea.</p>
36
<p>Add minimal facility for dealing with TELNET option
37
negotiation by returning WONT to DO and DONT requests.</p>
40
<p>If the backend is unhealthy, use a graced object if one is
44
<p>Make <span class="code">server.hostname</span>
45
and <span class="code">server.identity</span> available to VCL. The latter
46
can be set with the <span class="code">-i</span> parameter
47
to <span class="code">varnishd</span>.</p>
50
<p>Make <span class="code">restart</span> available
51
from <span class="code">vcl_error</span>.</p>
54
<p>Previously, only the TTL of an object was considered in
55
whether it would be marked as cacheable. This has been changed
56
to take the grace into consideration as well.</p>
59
<p>Previously, if an included ESI fragment had a zero size,
60
we would send out a zero-sized chunk which signifies
61
end-of-transmission. We now ignore zero-sized chunks.</p>
64
<p>We accidentially slept for far too long when we reached
65
the maximum number of open file descriptors. This has been
66
corrected and <span class="code">accept_fd_holdoff</span> now works
70
<p>Previously, when ESI processing, we did not look at the
71
full length, but stopped at the first NULL byte. We no longer
72
do that, enabling ESI processing of binary data.</p>
78
<p>Make sure system "..." returns successfully to ensure
79
test failures do not go unnoticed.</p>
82
<p>Make it possible to send NULL bytes through the testing