summaryrefslogtreecommitdiff
path: root/doc/upgrade.html
blob: cf46aee3717c8bda82e3a5accad71fb349d79339 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
    <meta http-equiv="Content-Language" content="en" />
    <title>skalibs: how to upgrade</title>
    <meta name="Description" content="skalibs: How to upgrade" />
    <meta name="Keywords" content="skalibs installation upgrade" />
    <!-- <link rel="stylesheet" type="text/css" href="http://skarnet.org/default.css" /> -->
  </head>
<body>

<p>
<a href="index.html">skalibs</a><br />
<a href="http://skarnet.org/software/">Software</a><br />
<a href="http://skarnet.org/">skarnet.org</a>
</p>

<h1> What has changed in skalibs </h1>

<h2> in 2.3.1.3 </h2>

<ul>
 <li> Bugfix release, no functional changes. </li>
</ul>

<h2> in 2.3.1.2 </h2>

<ul>
 <li> The child_spawn* family of functions now resets all signals
to their default values when spawning a program on systems with
<a href="http://pubs.opengroup.org/onlinepubs/9699919799/functions/posix_spawn.html">posix_spawn()</a>.
This should not change anything - it's just paranoia. </li>
 <li> Minor bugfixes. </li>
</ul>

<h2> in 2.3.1.1 </h2>

<ul>
 <li> Bugfix release, no functional changes. </li>
</ul>

<h2> in 2.3.1.0 </h2>

<ul>
 <li> wait_status() now returns 256 + signal number when WIFSIGNALED(). </li>
 <li> new macro wait_estatus() to get WIFSIGNALED information in an exit code. </li>
 <li> new functions in djbunix.h: hiercopy(), hiercopy_tmp() </li>
</ul>

<h2> in 2.3.0.1 </h2>

<ul>
 <li> wait_status() now returns 256 (instead of 126) when WIFSIGNALED(). </li>
</ul>

<h2> in 2.3.0.0 </h2>

<ul>
 <li> buffer_getvall and buffer_putvall have a different interface. (And now work.) </li>
 <li> buffer_read and buffer_write are now deprecated. </li>
 <li> various siovec additions, such as siovec_trunc </li>
</ul>

<h2> in 2.2.1.0 </h2>

<ul>
 <li> unixconnection added. </li>
 <li> sig_name and sig_number added. </li>
</ul>

<h2> in 2.2.0.0 </h2>

<ul>
 <li> TAI64 and TAI64N conversion functions revisited.
API slightly changed, made more consistent. </li>
</ul>

<h2> in 2.1.0.0 </h2>

<ul>
 <li> unixmessage_drop added, part of unixmessage/skaclient hardening against fd leakage </li>
 <li> unixmessage_sender_flush return code aligned with the convention in the rest of skalibs </li>
 <li> unixmessage_sender_flush now handles short writes </li>
 <li> skaclient_start(_async) prototype changed: now takes an option argument </li>
</ul>

<h2> in 2.0.0.0 </h2>

<ul>
 <li> The packaging system has completely changed. </li>
 <li> Programs should now link with <tt>-lskarnet</tt> followed by the appropriate
sysdeps links. </li>
 <li> No compatibility is ensured. Most APIs haven't changed, but no guarantee
is offered. </li>
 <li> The most important change is probably the disparition of <tt>struct tai</tt>
and <tt>struct taia</tt>, replaced with the <tt>tai_t</tt> and <tt>tain_t</tt>
types. Attosecond precision has been removed - processor speed is almost capped now,
and it looks like nanosecond precision will be enough for the foreseeable future. </li>
 <li> Buffer implementation has changed, and some interfaces too, mainly the
buffer_flush and bufalloc_flush return code. </li>
 <li> skaclient, skalibs' client-server communication suite, has been
entirely reimplemented: it's now using the new "unixmessage" set of primitives
for easier fd-passing. </li>
</ul>

</body>
</html>