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
108
109
110
111
112
113
114
115
116
117
|
Snappy v1.1.1, October 15th 2013:
* Add support for uncompressing to iovecs (scatter I/O).
The bulk of this patch was contributed by Mohit Aron.
* Speed up decompression by ~2%; much more so (~13-20%) on
a few benchmarks on given compilers and CPUs.
* Fix a few issues with MSVC compilation.
* Support truncated test data in the benchmark.
Snappy v1.1.0, January 18th 2013:
* Snappy now uses 64 kB block size instead of 32 kB. On average,
this means it compresses about 3% denser (more so for some
inputs), at the same or better speeds.
* libsnappy no longer depends on iostream.
* Some small performance improvements in compression on x86
(0.5–1%).
* Various portability fixes for ARM-based platforms, for MSVC,
and for GNU/Hurd.
Snappy v1.0.5, February 24th 2012:
* More speed improvements. Exactly how big will depend on
the architecture:
- 3–10% faster decompression for the base case (x86-64).
- ARMv7 and higher can now use unaligned accesses,
and will see about 30% faster decompression and
20–40% faster compression.
- 32-bit platforms (ARM and 32-bit x86) will see 2–5%
faster compression.
These are all cumulative (e.g., ARM gets all three speedups).
* Fixed an issue where the unit test would crash on system
with less than 256 MB address space available,
e.g. some embedded platforms.
* Added a framing format description, for use over e.g. HTTP,
or for a command-line compressor. We do not have any
implementations of this at the current point, but there seems
to be enough of a general interest in the topic.
Also make the format description slightly clearer.
* Remove some compile-time warnings in -Wall
(mostly signed/unsigned comparisons), for easier embedding
into projects that use -Wall -Werror.
Snappy v1.0.4, September 15th 2011:
* Speeded up the decompressor somewhat; typically about 2–8%
for Core i7, in 64-bit mode (comparable for Opteron).
Somewhat more for some tests, almost no gain for others.
* Make Snappy compile on certain platforms it didn't before
(Solaris with SunPro C++, HP-UX, AIX).
* Correct some minor errors in the format description.
Snappy v1.0.3, June 2nd 2011:
* Speeded up the decompressor somewhat; about 3-6% for Core 2,
6-13% for Core i7, and 5-12% for Opteron (all in 64-bit mode).
* Added compressed format documentation. This text is new,
but an earlier version from Zeev Tarantov was used as reference.
* Only link snappy_unittest against -lz and other autodetected
libraries, not libsnappy.so (which doesn't need any such dependency).
* Fixed some display issues in the microbenchmarks, one of which would
frequently make the test crash on GNU/Hurd.
Snappy v1.0.2, April 29th 2011:
* Relicense to a BSD-type license.
* Added C bindings, contributed by Martin Gieseking.
* More Win32 fixes, in particular for MSVC.
* Replace geo.protodata with a newer version.
* Fix timing inaccuracies in the unit test when comparing Snappy
to other algorithms.
Snappy v1.0.1, March 25th 2011:
This is a maintenance release, mostly containing minor fixes.
There is no new functionality. The most important fixes include:
* The COPYING file and all licensing headers now correctly state that
Snappy is licensed under the Apache 2.0 license.
* snappy_unittest should now compile natively under Windows,
as well as on embedded systems with no mmap().
* Various autotools nits have been fixed.
Snappy v1.0, March 17th 2011:
* Initial version.
|
