summaryrefslogtreecommitdiffstats
path: root/CHANGELOG.md
blob: 495d821e4875ee4844f1737784f94dacea349f55 (plain) (blame)
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
118
119
120
121
122
123
# Changelog
All notable changes to GNU Radio will be documented in this file.

The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/)
and this project adheres to [Semantic
Versioning](http://semver.org/spec/v2.0.0.html), starting with version 3.7.12.0.

Older Logs can be found in `docs/RELEASE-NOTES-*`.

## [3.8.0.0] - 2019-08-09

Witness me!

Tonight, we release GNU Radio 3.8.0.0.

It's the first minor release version since more than six years, not without
pride this community stands to face the brightest future SDR on general purpose
hardware ever had.

Since we've not been documenting changes in the shape of a Changelog for the
whole of the development that happened since GNU Radio 3.7.0, I'm afraid that
these release notes will be more of a GLTL;DR (git log too long; didn't read)
than a detailed account of what has changed.

What has _not** changed is the fact that GNU Radio is centered around a very
simple truth:

> Let the developers hack on DSP. Software interfaces are for humans, not the
> other way around.

And so, compared to the later 3.7 releases, nothing has fundamentally modified
the way one develops signal processing systems with GNU Radio: You write blocks,
and you combine blocks to be part of a larger signal processing flow graph.

With that as a success story, we of course have faced quite a bit of change in
the systems we use to develop and in the people that develop GNU Radio. This has
lead to several changes that weren't compatible with 3.7.

**This is an excerpt only:**

### Fixed

#### Project Scope

- Roughly 36 dBBug, (ETOOMANYBUGS)
- Fixed .so versioning

### Changed

#### Project Scope

- C++11
- merged the wholeness of the `next` branch
- Dependency version bumps: CMake, GCC, MSVC, Swig, Boost
- New dependencies: MPIR/GMP, Qt5, gsm, codec2
- Removed dependencies: libusb, Qt4, CppUnit
- Python: Python 2 & Python 3 compatible. 3.8 will be the last Py2k-compatible
  release series
- gengen was replaced by templates (if you don't know gengen, don't do any
  research; save yourself that sorrow)
- Modern CMake (as far as feasible at this point)
- VOLK version updated to v2.0.0
- .clang-format file now dictates coding style
- clang-format'ed the whole tree.
- installed CMake files now tell about configuration

#### gnuradio-runtime

- reworked fractional tag time handling, especially in the context of resamplers

#### GRC

- C++ generation as option
- YAML instead of XML
- removed `blks2`
- much better canvas tooling
- consistent gobject usage
- ROUNDED ARROWS

#### gr-qtgui

- moving from Qt4 to Qt5

#### gr-utils

- `gr_modtool` now vastly improved

#### gr-vocoder

- improved versatility
- removed in-tree libgsm, libcodec2, use system-wide libs


### Removed

#### Project Scope
- Modules `gr-comedi`, `gr-fcd` and `gr-wxgui` are gone

#### gr-comedi

- nobody could remember who used this, or for what. It has seen 0 active code
  contributions in the 3.7 lifecycle

#### gr-digital

- python-based `packet_encoder` and related tools: Bugs that were sporadic and
  never fixed, so after long deprecation, we're removing it

#### gr-fcd

- since it's currently untestable by the CI, it's being removed, as there was no
  code contributions. Generally, we strive to include all batteries with GNU
  Radio. Re-integration within a more general SDR interface would be desirable.

#### gr-utils

- removed PyQwt (dead) based tools

#### gr-wxgui

- Unmaintained, breaks on increasingly many systems, always was slower than
  Qtgui. We've been starting to tell people to migrate to Qt since at
  least 2015. Now, we're finally removing it.