13
0
Commit Graph

701 Commits

Author SHA1 Message Date
90d45ac428
Merge branch 'ardour' 2024-11-15 16:57:30 +01:00
6f331be7ca
Remove deprecated 32C define 2024-11-15 03:54:17 +01:00
64cc01b0b1
Merge branch 'ardour' 2024-11-10 23:49:01 +01:00
Mads Kiilerich
3228a61e62 wscript: the i18n worker functions do not define the i18n waf commands
Rename the worker functions to make it clear that their name in this
case isn't magic. These functions "are not" the waf commands. It is the
custom build context class definitions that define the i18n commands ...
which will invoke these top level worker functions which in turn invoke
the others recursively.

The bare printing of the build environment in the top level i18n command
seems to be old debug code that safely can be removed.
2024-10-20 19:52:14 -06:00
Mads Kiilerich
27acda4ccd wscript: the i18n class names do not define the i18n waf command names
Try to make it a bit more clear what is going on with class and def for
i18n commands.

It was confusing that we first defined i18n classes in the beginning of
the wscript file, and then replaced them with "plain" functions with the
same name at the end. pyflakes also didn't like it.

It seemed magic. The i18n functionality easily broke if trying to touch
that. It deserves an explanation to make it maintainable ... and some
cleanup.

Simple functions in the top level wscript file are generally exposed as
custom waf commands. The command will have the same name as the function
and will get a plain Context. But that simple method doesn't work for
these i18n commands. They have to be declared in a different way with a
custom BuildContext, as seen and described in the comment.

The name of the BuildContext classes doesn't matter, so we change the
name to avoid the name collision and to give a hint how they actually
just are contexts for the commands - they are not the command itself. We
also place the classes next to the corresponding functions so it is more
obvious that they are related.
2024-10-20 19:52:14 -06:00
Mads Kiilerich
c78b3289d5 wscript: set PROGRAM_NAME in env early, not as set_compiler_flags side effect 2024-10-20 19:52:14 -06:00
d0a7cd00b5
Merge branch 'ardour' 2024-10-20 19:31:53 +02:00
Mads Kiilerich
38db12c9ea Drop Frontier Tranzport surface - it has been dead for a decade 2024-10-20 03:08:03 +02:00
2c503eab09
Remove old Coreaudio OSX 10.5 support 2024-10-18 22:52:34 +02:00
99906a2215
Drop build-system support for macOS < 10.13 (require C++17) 2024-10-18 22:52:29 +02:00
9c292ed6f4
Merge branch 'ardour' 2024-10-17 15:38:42 +02:00
2d0f472af5 Consolidate macOS version parser 2024-10-15 00:57:10 +02:00
3171eb5aa1
C++17 requires macOS 10.13 or later 2024-10-15 00:35:50 +02:00
be4dab7336
Merge branch 'ardour' 2024-10-10 03:18:28 +02:00
0357c12fcc Really fix compiler warnings for clang < v15
see also 57a6773833
2024-09-26 01:31:25 +02:00
57a6773833
Check compiler version to set compiler warning options
This cleans up nightly build log (clang 13) and allows
clean compiler output on dev systems (clang 16).
2024-09-18 15:27:37 +02:00
09e9d7781b
Add build support for macOS/Sonoma
At some point in the not-too-distant-future we need
to simplify this and simply require Mojave/XCode 10 or later.
2024-09-17 00:01:51 +02:00
f1b9f48c87
Merge branch 'ardour' 2024-09-16 14:46:05 +02:00
9f438aa773
Add workaround to compile w/c++17 and old boost on macOS
This can be removed once the buildstack has been updated to
use recent boost.
2024-09-15 00:19:26 +02:00
e299635511
Allow to explicitly request C++11/17 2024-09-15 00:14:51 +02:00
0ed82d2f50
Add C++17 option to force compiles to use std=cxx17 2024-09-14 16:08:58 +02:00
544b53f65c
Merge branch 'ardour' 2024-09-05 23:49:53 +02:00
961cf955d2
Hide warnings caused by glibmm/helperlist.h declutter build log 2024-08-19 00:22:38 +02:00
c2e01bc636
Merge branch 'ardour' 2024-05-16 00:52:37 +02:00
afb519cd84
Bump required boost version to 1.68 (for optional::has_value) 2024-05-15 18:50:17 +02:00
18115f5383
Merge branch 'ardour' 2024-05-02 22:16:18 +02:00
2af2df3516
Raise the butler's I/O priority
This likely won't make much difference on modern systems,
since it requires a kernel based I/O scheduler. which is
disabled (set to "none" for NVMe and SSDs).
2024-04-29 21:46:06 +02:00
b0718b62a0
Properly disable LV2 support for Trax 2024-04-27 21:02:47 +02:00
119698e371
LiveTrax: disable plugin support (exclude plugin scan tools) 2024-04-27 03:03:02 +02:00
2f8b1186ae
Exclude scripted frontends for LiveTrax 2024-04-27 01:22:35 +02:00
e8cfc595d2
Do not compile ACE LV2 plugin for LiveTrax 2024-04-27 01:21:04 +02:00
3767754ec9 fix default program name 2024-04-25 15:29:33 -06:00
bdb74ddfcb first steps towards livetrax prefs editor 2024-04-25 13:19:35 -06:00
GianfrancoCostamagna
338cd09a4a
Work around itstools bug #9648
Forwarded from debian.
Having stderr output inside the returned command tricks the system
to fail in detecting the version.
Ignoring stderr output looks fine to make it build properly:

```
itstool --version
/usr/bin/itstool:239: SyntaxWarning: invalid escape sequence '\s'
  if re.sub('\s+', ' ', text).strip() != '':
/usr/bin/itstool:337: SyntaxWarning: invalid escape sequence '\s'
  message = re.sub('\s+', ' ', message).strip()
/usr/bin/itstool:475: SyntaxWarning: invalid escape sequence '\s'
  return re.sub('\s+', ' ', self.locnote).strip()
/usr/bin/itstool:477: SyntaxWarning: invalid escape sequence '\s'
  return '(itstool) link: ' + re.sub('\s+', ' ', self.locnoteref).strip()
/usr/bin/itstool:891: SyntaxWarning: invalid escape sequence '\<'
  regex = re.compile('(.*) \<(.*)\>, (.*)')
/usr/bin/itstool:926: SyntaxWarning: invalid escape sequence '\s'
  if re.sub('\s+', '', prevtext) == '':
/usr/bin/itstool:1452: SyntaxWarning: invalid escape sequence '\.'
  _locale_pattern = re.compile('([a-zA-Z0-9-]+)(_[A-Za-z0-9]+)?(@[A-Za-z0-9]+)?(\.[A-Za-z0-9]+)?')
itstool 2.0.6
```
2024-03-04 14:27:23 +01:00
572a8b9cf8
Use G_ENABLE_DEBUG with default debug builds 2024-02-24 04:17:47 +01:00
85e361044b
Add option to enable G/YTK debug 2024-02-22 23:54:28 +01:00
1117f8d6cc
Always built at least the dummy backend
It is required by AAF import, various session-utils, unit-tests,
and it is generally useful for testing.
2024-02-18 16:43:10 +01:00
ad51c7c2ba
Localize stripped down gtk2
This is intended mainly for GNU/Linux distros who will remove
GTK2 support in the near future.
2024-01-06 21:52:48 +01:00
c7772c220e
Vapor: Bootstrap Atmosphere plugin 2024-01-06 20:21:13 +01:00
49256a37e8
NO-OP: clarify source-tree structure 2024-01-06 14:23:47 +01:00
e3128e932c
Compile and install libAAF 2023-11-27 23:25:04 +01:00
282be3d459
Nobody cares about GLIB_2_64 anymore 2023-11-25 13:08:12 +01:00
Nils Philippsen
a33514e270 Make option to disable VST3 support consistent
Previously, the option was called "--vst3", yet disabled building
support for VST3 plugins (as described in the help string).

Signed-off-by: Nils Philippsen <nils@tiptoe.de>
2023-11-17 22:23:58 +01:00
c7eb410d0b Make Xcode output readable again 2023-10-27 15:12:49 +02:00
3ce2cf937e
Revert "wscript: drop unnecessary str() on configure variables"
This reverts commit 615326be9b because it
breaks windows builds.

```
File "/home/ardour/ardour-w64/wscript", line 1462, in configure
  set_compiler_flags (conf, Options.options)
File "/home/ardour/ardour-w64/wscript", line 522, in set_compiler_flags
  if re.search ('x86_64-w64', conf.env['CC']) is not None:
File "/usr/lib/python2.7/re.py", line 146, in search
  return _compile(pattern, flags).search(string)
TypeError: expected string or buffer
```
2023-10-16 21:15:25 +02:00
Mads Kiilerich
2973e82a90 wscript: sanitize strings from fetch_*_revision_date early instead of applying str() all over
This can perhaps be simplified further when Python2 support is dropped.
2023-10-15 10:47:16 -06:00
Mads Kiilerich
615326be9b wscript: drop unnecessary str() on configure variables
CC is already set to a string. (And if it ever should be None, we want
to handle that explicitly.)

(And #autowaf.display_msg handle Booleans just fine.)
2023-10-15 10:47:16 -06:00
Mads Kiilerich
2bb311f99b wscript: don't add pointless -DHAVE_X for each mandatory subdir
The sub_config_and_use function recursed, but it also invoked
autowaf.set_local_lib , which however didn't do anything useful. The
HAVE_ defines are not used anywhere, and the AUTOWAF_LOCAL defines are
only used in autowaf.use_lib, which however isn't used anywhere.

Dropping these defines simplify the build environment and makes the
compiler command line half as long and thus makes debugging much more
manageable.
2023-10-15 10:47:16 -06:00
Mads Kiilerich
e1b9c8c4d9 wscript: handle libs/clearlooks-newer like all other children
Put it in the "children" list instead of recursing explicitly.

The only difference is that the (empty) options function in
libs/clearlooks-newer/wscript now will be invoked - before it wasn't.
2023-10-15 10:47:16 -06:00
0be8456d0d Add support for MacOS 14 (Sonoma) 2023-09-27 04:15:41 +02:00