9268d2e699
allow to auto-start dummy backend (needed for tests)
2014-12-03 18:24:56 +01:00
d45e19eda5
fix test segfault
...
whe XMLTree goes out of scope, the root-node is deleted
2014-12-03 16:03:57 +01:00
0d677f70b4
use Dummy backend for tests
2014-12-03 16:03:19 +01:00
0ec85a4164
Fix running libpbd unit tests when $TOP(in test scripts) is set as relative path
...
Add test for PBD::canonical_path that was used to track down silly mistake
in test code
2014-12-03 17:44:48 +07:00
dd976d8c7a
Add DEBUG_TRACE output to PBD::canonical_path
2014-12-03 17:31:05 +07:00
81d47f5b8c
Call PBD::init() from libmidi++ test runner
...
This is mostly just to be able to set PBD_DEBUG for testing if
necessary
2014-12-03 17:31:05 +07:00
d360bc3906
Add ability to set libpbd debug options via PBD_DEBUG environment var
2014-12-03 17:31:05 +07:00
9eb8b73b43
Call PBD::init from evoral testrunner
...
This is needed to set _fmode = O_BINARY for libsmf/fread and
fixes takeFiveTest on windows
2014-12-03 17:31:05 +07:00
4b751f9e87
Move setting _fmode flag on windows into PBD::init
...
This moves setting of _fmode slightly later in the initialization
of the application via Ardour::init.
2014-12-03 17:31:05 +07:00
dc4d76b1d4
Remove references in pbd test script to libs no longer in source tree
2014-12-03 17:31:04 +07:00
01a17f296f
Make pbd test script consistant with other test scripts
...
remove running script from pwd limitation
2014-12-03 17:31:04 +07:00
4fe6bf3530
export MIDIPP_TEST_PATH in midi++ test script so test files can be located
...
Modify formatting of script to be consistant with other test scripts
2014-12-03 17:31:04 +07:00
23c0452e05
Add test_search_path to find test files in MidnamTest
2014-12-03 17:31:04 +07:00
aed58d6f39
Use test_output_directory() in pbd filesystem test
...
This prevents directory being created in pwd and means deletion of
directory is unnecessary
2014-12-03 17:31:04 +07:00
61f9ca9f3c
Use test utility function to find evoral test files
...
either via EVORAL_TEST_PATH env var or on windows via directory relative
to dll
2014-12-03 17:31:04 +07:00
f6e15a3b6b
Use PBD::tmp_writable_directory to write evoral test file into
...
no need to remove directory as it is temporary
2014-12-03 17:31:04 +07:00
929fc3d235
Use PBD::tmp_writable_directory in libardour test_util.cc
2014-12-03 17:31:04 +07:00
376340a29f
Remove unnecessary comma in pbd wscript file
2014-12-03 17:31:04 +07:00
1c1eff8845
Use PBD::tmp_writable_directory in test_output_directory
2014-12-03 17:31:04 +07:00
1abe8f4e42
Add PBD::tmp_writable_directory utility function to pbd/file_utils.h/cc
2014-12-03 17:31:04 +07:00
2b03521ef9
Fixes for AudioEngine test
2014-12-03 17:31:04 +07:00
e83bb2f4e2
Remove unused test_common.h/cc files
2014-12-03 17:31:04 +07:00
f1b66b86b4
Move libardour test utility code into one source file
2014-12-03 17:31:04 +07:00
b9567b6d46
Use ardev_common.sh to setup environment for running libardour tests
2014-12-03 17:31:03 +07:00
7a3c54f031
Fix evoral test suite.
2014-12-02 23:30:16 -05:00
25efe5953c
Fix range of pan controls.
...
Set default range to [0,1] since [0,0] is problematic and useless anyway.
2014-12-02 19:02:03 -05:00
94b26cd9af
Make 'ARDOUR::ParameterDescriptor' exportable, so that it can get used outside of libardour
2014-12-02 18:59:43 +00:00
a13ff7b604
Remove accidentally committed file.
2014-12-02 12:15:54 -05:00
7aab4be5e4
alsa-backend: drop midi during freewheel processing.
2014-12-02 17:03:29 +01:00
db3d42d551
first pass at optimization for low-level canvas code
2014-12-02 05:32:34 -05:00
cf47d71761
Fix lost MIDI events due to broken range check.
2014-12-02 03:36:33 -05:00
07a381a8f5
Create discrete lists for recorded MIDI controls.
2014-12-02 00:21:52 -05:00
86de0386c9
Fix automation write/touch.
2014-12-01 23:35:24 -05:00
2b69fa6edf
Factor out copy-paste code.
2014-12-01 23:35:24 -05:00
767c0238a3
Replace half-baked param metadata with descriptor.
...
Among other things, this means that automation controls/lists have the actual
min/max/normal/toggled of parameters, and not those inferred from the Parameter
ID, which is not correct for things like plugin parameters.
Pushing things down to the Evoral::ParmeterDescriptor may be useful in the
future to have lists do smarter things based on parameter range, but currently
I have just pushed down the above-mentioned currently used attributes.
2014-12-01 23:35:24 -05:00
Julien de Kozak
cb8abbe8d2
Make it compile with C++11 support.
...
Reference : https://bugs.webkit.org/show_bug.cgi?id=59249
2014-12-01 23:18:11 +01:00
43cc295a58
Tidy up comments in Evoral::coverage
...
Remove old (already #if 0'ed) implementation of Evoral::coverage() and its
comments.
Tidy up the comment enumerating all the possible ways in which two ranges
can overlap, note the Evoral::OverlapType corresponding to each one, and add
comments to the if()s in coverage corresponding to the cases in the list of
overlap types.
Remove some commented-out assert()s that actually do happen, and re-instate
one that really shouldn't.
Fix a small typo (with -> within)
2014-12-01 20:13:04 +00:00
dd5fc9432f
Simplify Evoral::RangeList::subtract(), and make it pass amended tests
...
The various conditionals in Evoral::RangeList::subtract() appear to have
been there to work around
(a) coverage() not always returning the correct value, &
(b) the test suite assuming that the ->to point lies outside the range
Now that these are both fixed, the implementation of subtract() becomes
quite a bit clearer. I replaced the if()s with assert()s for now, but these
shouldn't trip if coverage() is working as I expect.
Also (attempt to) clarify the comments in subtract.
2014-12-01 20:13:04 +00:00
30a9c2d05b
Comments in various call-sites of coverage()
...
Comments in various call sites of Evoral::coverage() marking things I think
are dubious (with XXX). Also straightened up the alignment of some ASCII
art in libs/ardour/diskstream.cc
2014-12-01 20:12:51 +00:00
d7cc7775f3
Rework Evoral::coverage() to pass unit tests
...
Rewrite Evoral::coverage() to (hopefully) do what it's supposed to.
Return OverlapNone for invalid ranges: if either of the ranges passed to
Evoral::coverage() have negative length (i.e. start > end), return OverlapNone
- it seems reasonable to say that a negative-length range can't overlap
anything. Also return OverlapNone from the fallthrough case, though this should
never happen.
2014-12-01 20:03:22 +00:00
47d329ca40
Make tests of Evoral::Range functions include both endpoints in Range
...
Some of the tests for Evoral::RangeList::subtract() assume that ranges
don't contain their end (->to) point. This appears inconsistent with how
they are used elsewhere.
Add some ASCII art comments to the tests to try to clarify what they're
really testing for, and amend subtractTest1, subtractTest4, & subtractTest5
to incorporate the assumption that ranges include their end points.
2014-12-01 16:52:58 +00:00
0c7dd82239
exclude fader-reserve from range
2014-12-01 17:45:27 +01:00
0ffeaa6d6a
Remove braindead Parameter inheritance abuse.
2014-11-30 23:56:20 -05:00
7eb4e5d22b
Replace thinning static state with parameter.
2014-11-30 23:56:20 -05:00
b68fd1cc25
Remove dead code.
2014-11-30 23:56:20 -05:00
0b48eb68f8
Remove bogus test and noisy output.
2014-11-30 23:56:19 -05:00
2588b1cac9
Trim the include tree.
2014-11-30 23:56:19 -05:00
1693e57e0e
Move EventRingBuffer to libardour.
...
This is not used anywhere in Evoral and is just a wrapper around the PBD
RingBuffer anyway. Towards a (once again?) independently buildable/testable
Evoral and fewer cross-dependencies.
2014-11-30 23:56:19 -05:00
cf537b97f5
Add missing namespace qualifiers.
2014-11-30 23:56:19 -05:00
008bfceb77
Clean up state tracking of raw MIDI.
2014-11-30 23:56:19 -05:00
a0deb7c3bf
Save mute automation state.
2014-11-30 23:56:19 -05:00
272c93db12
CLA approved consistent fine-grained scale modifier
2014-11-30 21:27:31 +01:00
2996e677c4
Process 'default_ui_config.in' when building with MSVC
2014-11-30 17:57:44 +00:00
18efe8f446
consistent step+page sizes for fader+spinbox
2014-11-30 15:05:40 +01:00
fb3f9033cf
use update_steps() instead of custom function
2014-11-30 15:05:39 +01:00
0622fc9d4b
update_steps(), properly handle integer ranges
2014-11-30 15:05:39 +01:00
bd85c8ec43
fix ladpsa default value.
...
because get_parameter_descriptor() is const, also wrap
default_value() in a const function.
2014-11-30 15:05:39 +01:00
cf859270e0
fix LV2 default port value.
2014-11-30 15:05:39 +01:00
77d5fb7a35
Fix write for boolean automation.
2014-11-29 14:32:12 -05:00
c77f4b7a18
don't move faders on click.
...
for one, it can mess up the value when switching to
numeric-entry. It is also rather useless and not
Fader-like (faders are not scrollbars).
Besides, "stepvalue" is way to large and causes jumps
(sometimes step-value is even negative, see
"fast lookahead limiter", "release time"
-> some other bug)
2014-11-29 17:31:41 +01:00
58663db374
change ownership of processor window-proxy
...
fixes crashes:
* If the Editor-Mixer shows a channel with a plugin that
has been edited in the Mixer, double-clicking the plugin
will try to bring up a 2nd instance of the plugin-UI.
* When closing Ardour both the Mixer and the Editor-Mixer try to delete the underlying plugin, resulting in a double free.
2014-11-29 10:06:25 +01:00
784b96e255
Set discrete by default for boolean parameters.
2014-11-29 01:40:23 -05:00
ed1bbefd19
Somewhat working gain automation.
...
This probably isn't correct in several ways, but it works more than it did, so
I figure it's push worthy.
Still not working:
* Saving mute automation list
* Dragged control points are not snapped to model restrictions
(boolean, in this case, but general problem)
* Line goes funny if you record mute automation
(as opposed to drawing it which works)
2014-11-29 01:25:27 -05:00
f190e630ab
Show discrete/toggled automation as stepped line.
2014-11-28 21:15:28 -05:00
e55b68298b
Fix crash when deleting all points from a line.
2014-11-28 20:04:57 -05:00
358ebde106
Set the toggled default metadata of parameters.
...
This is so generic code looking at the Parameter for gain, solo, and
rec-enable will get the correct value for toggled.
2014-11-28 18:04:51 -05:00
ea35595370
Add some newly introduced source files to our MSVC project (cairocanvas)
2014-11-28 14:42:17 +00:00
16947f751b
Remove dead code.
2014-11-27 04:17:15 -05:00
3d4b1d324e
Fix calculation using uninitialized value.
...
Implicit "this" is the devil.
2014-11-27 04:15:30 -05:00
25e1eddc50
remove "port sources" management from MidiControlUI and Mackie support.
...
This was necessary due to a bug/design issue between Glibmm and Glib (see https://bugzilla.gnome.org/show_bug.cgi?id=561885 )
but the problem needs to be managed by the *creator* of the IOSource and that has now moved inside CrossThreadChannel.
2014-11-26 18:22:22 +02:00
6edd81e4f6
get correct ::distance() and ::delta() implementations for cases involving achromatics
2014-11-26 18:22:22 +02:00
318a9edab9
massive reworking of color selection implementation
2014-11-26 18:22:22 +02:00
c927eeb9ee
ensure that Ardour recognizes its own MMC device ID as "all-call" when set to 0x7f (127)
2014-11-26 18:22:22 +02:00
b9be8d2652
Replace individual waf options for enabling backends with one --with-backends option
2014-11-26 22:47:05 +07:00
4358b8db79
Add test for Evoral::coverage()
...
Add a test function to test Evoral::coverage() with all possible overlap
types. The first test (line 161) that expects OverlapExternal will fail
with the current implementation of coverage().
There's possibly still a discussion to be had about what the overlap type of
ranges with negative lengths should be: there are currently places in the main
Ardour code base where coverage() is called with ranges where start > end.
2014-11-25 17:49:51 +00:00
3a6c9e3bb9
Fix build of evoral tests
...
Fix compile errors in libs/evoral/test/, by explicitly calling
Evoral::MusicalTime::to_double() wherever a double value is required of a
MusicalTime.
Some of the double variables should probably really be made into MusicalTime
ones instead, but I don't want to mess with this too much.
takeFiveTest still fails for me after this, but a failing test is probably
more informative in the long run than a test that won't even compile.
2014-11-25 14:59:48 +00:00
1f58f81aa6
Fix order of parameters to stop_touch() in Mackie::Control
...
Reverse the parameters of Mackie::Control::stop_touch() to make them
consistent with AutomationControl::stop_touch(), and fix up the call to
AutomationControl::stop_touch() to have the parameters in the correct
order.
Unfortunately, I don't possess any devices that speak the Mackie protocol, so
though the patch seems logical and correct to me, I have no way of testing it.
If anyone has a device with touch faders that speaks Mackie, I'd be glad of any
confirmation that it at least doesn't break anything.
2014-11-25 14:59:48 +00:00
ea5876b836
Remove dead code.
2014-11-25 02:11:44 -05:00
78218e8c07
fix OSX open URL in browser
2014-11-24 20:35:38 +01:00
c7ddedc5b8
Simulate 'trunc()' which isn't available in MSVC
...
(needed by 'gtk2_ardour/editor.cc')
2014-11-24 12:56:07 +00:00
018f559aad
Remove a commented out line in 'pbd/stacktrace.h'
...
(was probably just needed for testing at some stage)
2014-11-24 12:53:24 +00:00
25745758aa
Changes needed to accommodate building libevoral with MSVC
...
Note that class 'MusicalTime' needn't be exportable (with a little work, this could be reverted if it affects the other builds)
2014-11-24 12:51:54 +00:00
c0e8647806
amend c4533939b9
2014-11-24 00:04:45 +01:00
cd91fa0046
hotfix windows-compilation (unfinished bfe49540b9
)
...
..at least it compiles and starts. Mackie support etc
is probably still broken.
2014-11-23 23:57:56 +01:00
8ddeb712f6
make debug message more useful
2014-11-23 23:57:55 +01:00
5aba4df252
fix/silence various compiler warnings.
...
* ifdef unused static functions
* brackets around assignment and comparision
* no return statement in function returning non-void
* boost concept_checks.hpp unused-local-typedefs
2014-11-23 23:57:55 +01:00
f1926dc863
Fix build with a certain awful broken compiler.
2014-11-23 16:31:59 -05:00
ec01682714
-Wunused-function on windows
2014-11-23 21:01:39 +01:00
40764f4cb3
use NULL not 0 for c-functions (even in c++ code)
...
4.3 KBytes less compiler warnings.
2014-11-23 21:01:39 +01:00
bfe49540b9
remove use of CrossThreadChannel::selectable() to allow cross-platform implementation of CrossThreadChannel via Glib::IOSource
2014-11-23 17:37:13 +02:00
57c1b6e261
Fix quantization and other time-related ops.
2014-11-22 21:49:42 -05:00
231e0009b4
Max is max and min is min and I am dumb.
2014-11-22 16:33:32 -05:00
42b5ed3904
Fix clang and MSVC build.
2014-11-22 15:26:41 -05:00
c1cfa12d6e
Wrap MusicalTime in a class.
...
This lets us get a more explicit handle on time conversions, and is the main
step towards using actual beat:tick time and getting away from floating point
precision problems.
2014-11-22 04:05:42 -05:00
cae74309a5
Fix assert failure in control list interpolation.
...
How we never hit this before is beyond me, it's in some ooooold code.
2014-11-22 00:46:15 -05:00
01493b14cf
Fix const violation warnings in tempo.cc.
2014-11-22 00:46:15 -05:00
4bc0d1c475
Fix lost MIDI note offs and controllers.
...
Fix initial read of discrete MIDI controllers.
Fix spurious note offs when starting to play in the middle of a note.
Faster search for initial event when cached iterator is invalid.
So much for dropping the cached iterator. The iterator is responsible for
handling note offs, so that doesn't work. This design means we have some stuck
note issues at the source read level, but they should be taken care of by the
state tracker anyway.
2014-11-20 15:36:11 -05:00
8bba63282a
Reduce unnecessary non-const MIDI buffer access.
2014-11-20 15:32:06 -05:00
8f8a4878f3
Clean up some MIDI code.
2014-11-20 11:31:31 -05:00
0c4869dc35
debug trace lost midi-messages
2014-11-20 02:46:30 +01:00
a9936205b1
Fix dropped MIDI events, especially with record enabled.
...
I am not precisely sure why the cached iterator was causing this problem, it
shouldn't be invalidated, and the times make sense. It may be some lock
related issue since the iterator holds a lock on the source.
In any case, this cached iterator was just to avoid repeated linear search of
the model, but since the model has a logarithmic search, instead just scrap all
this problematic persistent state and search for the appropriate start time
every read. No need to be careful about invalidating when anything changes.
2014-11-19 19:24:16 -05:00
e3e85a92db
some const'ness and hints for clang
2014-11-19 20:38:50 +01:00
0fb91e10b4
allow dummy backend for any product
2014-11-18 17:04:23 -06:00
bd577a0663
Avoid possible use of uninitialized values.
...
This could only happen with a broken widget, but it shuts up clang and doesn't
hurt.
2014-11-18 02:55:03 -05:00
976a150e6b
Remove obviously dead/redundant code.
2014-11-18 02:55:03 -05:00
07d4f50689
Clean up funny code and initialize variables.
2014-11-18 02:55:03 -05:00
881d77f1e5
Constrain variable scopen and avoid uninitialized values existing at all.
...
Note: dead assignments trigger scan-build dead code warnings, which
initializations do not.
2014-11-18 02:55:03 -05:00
6e482b3488
Remove unused variable and redundant arithmetic.
2014-11-18 02:55:03 -05:00
8dde26da7b
Gracefully avoid importing audio that somehow has 0 channels.
2014-11-18 02:55:03 -05:00
e5d9fd1aac
Fix potential use of uninitialized value.
2014-11-18 02:53:28 -05:00
fed4e8446b
Fix NULL pointer dereference on malformed Playlist XML.
2014-11-18 02:53:28 -05:00
afd67800b9
Avoid potential division by zero.
2014-11-18 02:53:28 -05:00
c4c7598adb
Fix broken logic and possible NULL pointer dereference for Bundle XML.
2014-11-18 02:53:28 -05:00
26f6f80a53
Fix incredibly unlikely NULL pointer dereference.
2014-11-18 02:53:28 -05:00
9aec0c627d
Fix Region::verify_length() so it actually does something.
2014-11-18 02:53:27 -05:00
b6db707989
amend 484e38053
2014-11-18 05:28:27 +01:00
484e380536
silence clang "dead assignment" warnings
2014-11-18 04:46:37 +01:00
37937d9c69
synchronize preferences and monitor-section
2014-11-18 03:49:20 +01:00
5e90b97c42
files should have at least one channel.
2014-11-18 02:15:07 +01:00
f2273dc3b6
remove dead assignment
2014-11-18 02:15:06 +01:00
c194e4df18
remove superfluous if-branch (_session is checked a few lines above)
2014-11-18 02:15:06 +01:00
f7ebae85a7
Fix "maybe" rounding bug.
...
Introduced in d63161426f
.
2014-11-17 19:53:30 -05:00
02d735ff00
waves/coreaudio PPC support
2014-11-17 09:07:39 +01:00
bd46de779f
check alsa midi ringbuffer
2014-11-17 06:50:48 +01:00
b4fcdbb100
Don't add origin value to MIDI controller regions on initial insert.
...
This should probably hijack the same modifier as the guard points and work the
same on all automation tracks, but I did it this way to not change behaviour of
track automation where a default is much more reasonable.
2014-11-16 22:35:45 -05:00
d63161426f
Add "maybe" rounding modes for rounding only if necessary.
2014-11-16 22:35:45 -05:00
fd9ccc7058
Use an enum for RoundMode instead of magic numbers.
...
No functional changes in this one (for easier auditing), but towards having
round up/down only if necessary modes, rather than kludging around that
situation with a double round as we do currently.
2014-11-16 22:35:45 -05:00
9d4f43fa7d
Make ControlList::paste const-correct.
2014-11-16 22:35:45 -05:00
df40782b62
prevent segfault if in/out property is missing
2014-11-17 04:05:00 +01:00
ad739cd018
add some assert() messages.
...
as hints clang static analyzer "Called C++ object pointer is null"
2014-11-17 03:49:19 +01:00
333c776e0e
Fix MIDI region keyboard editing (add key handling to canvas).
2014-11-15 02:04:28 -05:00
b52407043c
Fix crash when copying MIDI regions with midi-copy-is-fork is on ( #6002 ).
2014-11-14 23:53:59 -05:00
94954f852e
Fix out of order event errors on save with overlapping notes.
...
Another consequence of fuzzy Sequence timing, but if the difference is less
than a tick this should handle things correctly. If the difference is more
than a tick, something's wrong, and it might be okay to just bump forward
anyway, but I can't reproduce this and it could lead to corruption so I'm
leaving that case noisy.
2014-11-14 21:19:09 -05:00
6b3a8915f3
add abort() to non-reached code
...
This cleans up a lot of false-positives in static analysis
and also helps compilers to optimize code paths in general.
(tagging the fatal stingstream operator as ‘noreturn’ is
far less trivial)
2014-11-14 11:30:08 +01:00
8907477271
Fix crashes on various MIDI editing operations.
...
Assertions assumed old strict ordering, no longer reflecting reality of fuzzy
time comparison (introduced in 86f1b8).
2014-11-13 22:34:20 -05:00
4ba4677b45
yet another copy/paste typo :(
2014-11-13 22:19:22 +01:00
9301ffafab
fix typo in b9fc616538
2014-11-13 21:46:30 +01:00
b376f6bc45
auto-initialize weak-jack
...
(not needed with ardour - but keep code in sync with
https://github.com/x42/weakjack )
2014-11-13 21:31:09 +01:00
b9fc616538
prepare for JACK meta-data API
2014-11-13 21:30:13 +01:00
65fa38103d
make new jack thread API functions optional
2014-11-13 21:29:34 +01:00
7b39e0893a
add symbol debug information to weak-jack
2014-11-13 21:28:33 +01:00
cb38034334
install all additional binaries in $libdir
2014-11-13 19:57:40 +01:00
9a9a8f78ef
fix define
2014-11-13 19:05:58 +01:00
0689a2976e
Don't iterate in 'ARDOUR::Route::get_control()' unless a ReaderLock was successfully obtained
2014-11-13 13:07:34 +00:00
e59e8f6ab9
Make sure we can't execute 'ARDOUR::Route::meter()' unless a ReaderLock was successfully obtained
2014-11-13 12:13:16 +00:00
3e702c5520
yet another build fix
...
on x86_64 weak_libjack.c needs -fPIC which needs cflags which need 'c' :)
2014-11-13 01:02:09 +01:00
533b545295
add popup message and error log for AU crash reports
2014-11-12 13:35:15 +01:00
5624301a31
Prepare AU-discovery crashlog.
...
stopgap solution to allow a user to detect which plugin caused a crash during discovery.
2014-11-12 13:34:46 +01:00
1bc04ae351
failsafe AU scanning part one:
...
Don’t scan for AU plugins unless users explicitly starts a
plugin scan. Disable automatic scanning if Ardour crashes
during AU plugin discovery.
2014-11-12 04:58:53 +01:00
9254e80c39
refactor MIDISceneChange color property addition by moving it into SceneChange to anticipate other types of SceneChange objects (e.g. OSC)
2014-11-10 23:17:27 -05:00
e27651d315
when creating a new MIDI scene change, check existing scene changes for equality and reuse their color (which may or may not be set)
2014-11-10 23:17:19 -05:00
0f243d5f7c
serialize/deserialize MIDISceneChange color; put out of bounds color into static const
2014-11-10 23:17:08 -05:00
70960a5a00
add color property to MIDISceneChange
2014-11-10 23:16:59 -05:00
3d0efb8bf7
add operator== to ARDOUR::MidiSceneChange
2014-11-10 23:16:53 -05:00
12b715e6fa
tweak canvas on_expose_event() handler to avoid unnecessary save/paint/restore
2014-11-10 20:34:20 -05:00
fdcb792358
further work on new color manipulation code
2014-11-10 20:33:54 -05:00
929d3fe6ed
fix signed/unsigned compiler warning
2014-11-10 22:21:39 +01:00
a37ffd1f0c
improve random seed (for dummy generators)
2014-11-10 22:20:36 +01:00
1317b57221
add required header for compilation
2014-11-10 11:46:17 -05:00
ef56948d4a
add background color property to Canvas::Canvas ; rearrange expose handling and include background fill
...
Conflicts:
libs/canvas/canvas.cc
libs/canvas/canvas/canvas.h
2014-11-10 10:43:25 -05:00
9cf5b23392
Modifications needed to build ReasonableSynth with MSVC
2014-11-09 08:11:46 +00:00
373caeede8
Add some newly introduced source files to our MSVC project (cairocanvas)
2014-11-08 10:48:26 +00:00
bb12c750bb
Just rename problematic Variant::VOID type.
2014-11-07 17:21:43 -05:00
4260d0ca0e
use new canvas colors header as necessary
2014-11-07 14:24:27 -05:00
69fd25f6f4
move canvas colors code into its own file(s)
2014-11-07 14:24:13 -05:00
a441de1a18
when drawing only part of a Canvas::Rectangle outline, we only need to do the cairo-pixel alignment math for the axis perpendicular to the edge
2014-11-07 12:58:56 -05:00
87cd481a7d
VOID is a windows define (not MSVC specific)
2014-11-07 17:04:11 +01:00
0fedfab593
tentative fix for canvas background/image surface
2014-11-07 16:50:21 +01:00
159cb4d2f9
another hack for windows timers, DSP load calculation
2014-11-07 12:17:15 +01:00
7670e463cc
add API to extract color from ArdourCanvas::Text and tidy up member declarations
...
Conflicts:
libs/canvas/canvas/text.h
2014-11-06 20:38:02 -05:00
787f9748b6
changing the clamp_width of an ArdourCanvas::Text needs to potentially provoke an entire redraw
...
This fixes missing range marker text in sessions where the initial zoom value causes an initial clamp_width of zero
2014-11-06 20:38:02 -05:00
6cda13fad4
possible fix for DSP load calculation on windows
2014-11-06 17:29:35 +01:00
c9c4a5c29d
raising/lowering canvas items should trigger a redraw
2014-11-06 10:58:37 -05:00
ecd618c343
correct the rendering of Rectangle outlines, and the computation of their bounding box
2014-11-05 19:45:23 -05:00
857719f2e1
add Rect::shrink(Distance) even though it arguably should be handled by Rect::expand()
2014-11-05 19:45:23 -05:00
f3d349bc9a
add some more outline enums to ArdourCanvas::Rectangle
2014-11-05 19:45:23 -05:00
589e59d7a1
allow derived classes to access Item::{scroll,position}_offset(), mostly for debugging purposes
2014-11-05 19:45:23 -05:00
98b66b9f3d
catch up with waves/tracks version of MIDI scene changer code
2014-11-05 19:45:23 -05:00
b02ea3d435
alter implementation of ArdourCanvas::contrasting_text_color() to use gamma-adjusted luminance values rather than simple contrast
2014-11-05 11:13:28 -05:00
70e95f951d
add ArdourCanvas::Fill::set_pattern() to allow any filled canvas item to be filled using an arbitrary Cairo Pattern
2014-11-05 11:13:09 -05:00
4eac2a76da
fix insane design for Location::next_available_name() so that it has no effective limit (other than UINT32_MAX)
2014-11-05 07:41:43 -05:00
f018f3d37a
Add some newly introduced header files to our MSVC project (libs/ardour)
2014-11-04 15:44:27 +00:00
227732d15b
When building with MSVC undef type VOID if we're declaring an enum called VOID
2014-11-04 15:42:04 +00:00
04bda495c2
When building with MSVC, guard a declaration of 'strdup()' which conflicts slightly with the MSVC declaration
...
(maybe we should just #include <string.h> ?)
2014-11-04 13:23:15 +00:00
5e8c39ff47
When running an (MSVC built) VST scanner app, make sure we can utilise debuggable versions if we need to
2014-11-04 13:23:14 +00:00
f1e6b28ab7
use ArdourCanvas::TimeRectangle for regions, notes, markers
2014-11-03 21:48:08 -05:00
56994e785e
add new TimeRectangle to ArdourCanvas
2014-11-03 21:48:08 -05:00
a54f8141dc
Use format string even for units we understand (print Hz units).
2014-11-03 20:52:14 -05:00
9f3f017736
Check whether frequency controls are audible or low and provide appropriate controls.
...
Always clamp set value to controller range.
2014-11-03 18:41:57 -05:00
ef16fdcd89
Fix parameter descriptors for one control leaking into another.
2014-11-03 18:40:17 -05:00
6e16f69c26
Add menu to set frequency controls in beats, and half or double current value.
...
Test with e.g. fomp:cs_phaser1_lfo LFO frequency.
2014-11-03 17:29:11 -05:00
7204702c3f
Reduce coupling between Plugin and PluginInsert.
2014-11-03 13:40:50 -05:00
68507badd9
Fix compilation with old LV2. Maybe.
2014-11-03 11:54:35 -05:00
6e8e3856cb
Show 3 decimal digits when printing values.
2014-11-02 18:47:22 -05:00
bb4e3a2916
add Gtkmm2ext::event_inside_widget_window()
2014-11-02 08:57:47 -05:00
d02d8337a9
Support pretty display of arbitrary plugin parameter units.
...
Try e.g. fomp:cs_chorus1 which has Hz and ms controls.
2014-11-02 02:10:24 -05:00
47c4929bc2
Display gain and midiNote plugin parameters/properties nicely.
...
Show fancy values on generic GUI controls, automation lane controls, and automation lane verbose cursor.
Fix text display of midiNote values.
Make bigstep of midiNote parameters 12 (one octave).
Add ARDOUR::value_as_string() as a stateless one-stop-shop for value printing.
2014-11-02 02:10:24 -05:00
8a128b33d3
Automation of LV2 plugin properties.
...
Work towards ParameterDescriptor being used more universally to describe control characteristics.
2014-11-02 02:10:24 -05:00
6dfb11c2d0
Move ParameterDescriptor from Plugin to its own header.
...
This fixes circular dependency issues that arise when using ParameterDescriptor more widely.
2014-11-02 02:10:24 -05:00
bd6ba1717e
Decouple ControlUI from port index.
2014-11-02 02:10:24 -05:00
b6d8805961
properly initialize variable (amend cffd18d2d8
)
2014-11-02 02:50:28 +01:00
cffd18d2d8
implement spin-button internal<>interface bridge
...
Widgets were written to use "[User] Interface" values, but the
spinbutton in a Bar-controller should allow to directly access
the "Internal Value".
2014-11-01 18:24:31 +01:00
049c11de58
Slider Controller API Change
...
Prepare to allow BarController (with Spinbutton) to translate
internal/interface values for the SpinBox.
(The change is irrelevant for the fader/slider itself which
always use the [user-]interface value)
2014-11-01 18:22:29 +01:00
d948e4cc91
Add a newly introduced header file to our MSVC project (libardour)
2014-11-01 13:53:55 +00:00
660e8c4425
don't print error when announcing property values for plugins that have none
2014-11-01 07:27:36 +01:00
002ebcc4ea
increase fine grained control (Ctrl+Alt+[drag|scroll])
2014-11-01 07:27:35 +01:00
c855d17e09
Fix compilation against older LV2.
2014-10-31 21:03:12 -04:00
5de6c21ec1
More generic RT-safe implementation of LV2 properties.
2014-10-31 20:46:43 -04:00
324ab35abc
Use precise define names for LV2 version configure checks.
2014-10-31 20:46:09 -04:00
0a60077a61
Fix spelling error.
2014-10-31 20:46:09 -04:00
8d4cbc4cc1
fix Dummy DSP load on win32 & low-pass filter load
...
g_get_monotonic_time() has millisecond granularity,
use performance Counter if available
2014-10-31 22:51:47 +01:00
f02610b47c
LV2 patch amends:
...
* fix Label (user readable key)
* check if patch range is a file Path
* free some lilv-nodes (plug small mem leak)
2014-10-31 19:42:28 +01:00
90872c2b31
more changes for rectangle outlining
2014-10-31 10:56:08 -04:00
a942493565
comment fix and expansion re: Canvas::Rectangle outline
2014-10-31 10:56:08 -04:00
b8cea19b95
prototype [LV2]patch-change support for generic plugin UIs.
2014-10-31 03:26:47 +01:00
973d40e3ac
Add support for USE_CAIRO_IMAGE_SURFACE when building with MSVC
...
(new preprocessor directive - greatly improves CPU loading)
2014-10-29 17:36:30 +00:00
dc39448a22
turn pixfader into a cairowidget
2014-10-28 02:18:46 +01:00
b04414d170
allow to use cairo-image/software surface for canvas & cairowidgets
2014-10-28 02:15:10 +01:00
6dc3a4bf61
remove unused old API
2014-10-28 01:27:14 +01:00
e02efddb21
skip meter re-draw when no pixels are changed.
2014-10-25 17:28:14 +02:00
a6ceff61dd
proper solution for variable-args jack_client_open()
2014-10-25 17:28:13 +02:00
9239a49a32
Add some newly introduced source files to our MSVC project (gtkmm2ext)
2014-10-24 18:45:53 +01:00
85eda0a5f3
Add some newly introduced source files to our MSVC project (jack_audiobackend)
2014-10-24 18:45:52 +01:00
3d39689769
Some minor bits of casting to keep MSVC happy when building 'backends/jack/weak_libjack.c'
2014-10-24 18:45:51 +01:00
f900711136
port changes to ARDOUR::Location and ARDOUR::Locations APIs from Tracks to Ardour.
...
Fixes deadlocks caused by mutex on Locations list, and clarifies the purposes and uses of the class-level and
object-level change-related signals.
2014-10-24 12:18:46 -04:00
2f4a8cf693
avoid const cast warning
2014-10-24 12:18:46 -04:00
65036cf3eb
add #include that had gone unnoticed before
2014-10-24 12:18:46 -04:00
b52304661f
fix screwup in prev commit-diff
2014-10-24 17:19:28 +02:00
5e7f935a27
add explicit construction (for MSVC) and some debug messages
2014-10-24 17:15:03 +02:00
25a6aa57a3
fix jack transport return value
2014-10-24 16:19:44 +02:00
e025cc008c
use x-macros to simplify weak-jack implementation
2014-10-24 15:18:38 +02:00
ad09be8c78
fix libjack names
2014-10-24 14:36:04 +02:00
1e357b5109
fix 64bit compile with old gcc
2014-10-23 20:42:02 +02:00
2bc3ec6ab3
add new Gtk2mmext::EmScale class for measuring font-based sizing
2014-10-23 13:21:04 -04:00
9a60bb58b2
skip unavailable backends early on.
2014-10-23 16:53:14 +02:00
d991bb10ca
add 'available' interface to the AudioBackendInfo
...
If a backend can be loaded, it does not mean that it can be used;
e.g. weak-linked jack-backend if libjack is not available.
2014-10-23 16:53:14 +02:00
f3ff1b9669
weak/runtime jack linking: load libjack dynamically at runtime
2014-10-23 16:53:13 +02:00
730e09ce65
another round of compiler warning fixes
2014-10-23 05:31:40 +02:00
c6a3d6bc48
small round of compiler warning fixes
2014-10-23 03:43:04 +02:00
f7211ff8f7
fix jack/mingw compilation
2014-10-22 23:57:55 +02:00
679356c5ee
fix delivery of MMC events under split process cycle conditions
2014-10-22 17:07:11 -04:00
e32a8e8260
require a timestamp on MIDI::MachineControl::send()
2014-10-22 17:07:11 -04:00
a289947b4f
fix timestamps for immediate MIDI events to deal with split process cycle conditions
2014-10-22 17:07:11 -04:00
bd97450ef8
error/debug output when dropping MIDI event due to timing now shows MIDI bytes
2014-10-22 17:07:11 -04:00
aafd1d77e9
remove unused code
2014-10-22 17:07:11 -04:00
68458cde69
fix all 4 backends' failure to include the main "backend" thread when computing ::in_process_thread()
2014-10-22 17:07:11 -04:00
aa0effb4cb
ignore negative value locates and MMC locate commands
2014-10-22 17:07:11 -04:00
3ddd797e46
remove obsolete jack includes in libardour
2014-10-22 16:56:12 +02:00
fbaae45eb1
stopgap solution for VST plugins that call audioMasterWantMidi during effOpen
...
audioMasterWantMidi needs plugin _info, but _info is currently only
available after instantiating the plugin in LXVSTPluginInfo::load().
vstfx_instantiate() -> effOpen -> [plugin] -> vst_callback(..,opcode=6,..) -> crash
2014-10-20 23:30:49 +02:00
825e2c823c
add support for VST plugins without any parameters
2014-10-20 23:30:48 +02:00
ffde2fb5f9
OS 10.10 compile fixes
...
* fix clang pickiness regarding boolean
* ignore ‘verify’ macro
* tested with Apple LLVM version 6.0 (clang-600.0.51) (based on LLVM 3.5svn)
2014-10-19 23:15:13 +02:00
ab445a0e28
windows: don’t popup message box when libjack is not found
2014-10-17 10:58:49 +02:00
d34bd9e6a0
fix midi automation sliders
...
Allow controls to work without a list. see also 34c1465
and b469cd2
2014-10-16 21:21:45 +02:00
0477d11476
when cancelling an import, call DropReferences so that the Session forgets about the new sources.
...
Not doing this leaves the sources in the session list and then the session fails to open on next load
because the files are not there.
Arguably we should not announce the new files until they are complete, but this is a simpler fix for now.
2014-10-15 21:12:35 -04:00
49b9d22217
remove debug output
2014-10-15 19:08:46 -04:00
5d53335300
when flushing notifications/signals from an ARDOUR::Playlist, emit the RegionAdded/ContentsChanged signals first so that when LayersChanged is sent, the receivers know about all the new regions
2014-10-15 18:44:22 -04:00
ab658d7ca1
add mutex/lock to all Evoral::SMF methods that use _smf/libsmf, to avoid inadvertent corruption via multithreaded access.
...
Serialization of Session::save_state() will already protect against most of this, but there is really no
good reason why Evoral::SMF's API should require single-threaded/explicit serialization.
2014-10-15 18:44:22 -04:00
7e764ea405
allow internal generators to make noise, even if we are not rolling, and using auto-input
2014-10-15 11:09:00 -05:00
8dc4187f1a
allow zero-latency sessions
...
e.g. sessions with disconnected master-out,
or Dummmy (both jack and ardour) with no latency.
2014-10-15 15:49:49 +02:00
76edbdf5dc
make backend selection by product name more explicit
2014-10-14 16:58:11 -04:00
d3a4a92bd5
Containers should not consider the size of invisible items when computing their own bounding boxes.
...
This fixes issues related to the rulers container believing it was taller than it actually is, and possibly other issues also
2014-10-13 22:40:08 -04:00
b945fa69ba
remove default/testing tooltips from the canvas
2014-10-13 13:53:07 -04:00
7263877f06
tweak implementation of Session::audible_frame() to be a little more defensible, and correct problems with record align/playhead position while disconnected from physical outputs
2014-10-12 23:09:35 -04:00
848db8fbd5
fix scanning of VST shell plugins
2014-10-11 22:19:23 +02:00
250ea066ef
revert 8ca546e
- original version was correct (note added)
2014-10-11 15:00:52 +02:00
8525a0b258
add a Dummy Backend Loopback mode:
2014-10-11 14:47:35 +02:00
8ca546ec88
fix LTC generator latency compensation.
2014-10-11 11:15:08 +02:00
fa9780ba67
subtle changes to accomplish two goals (1) playhead should stop where the user pressed stopped (2) captured regions should end where the playhead ends
2014-10-10 13:22:50 -04:00
d1e303247b
ensure registration of all SessionEvent types
2014-10-10 13:22:50 -04:00
446a8b69d9
make output of SessionEventManager::dump_events() a bit more readable
2014-10-10 13:22:50 -04:00
472900f4ff
cleanup unused nested regions:
...
This and two previous commits fixes #5979
2014-10-10 14:12:48 +02:00
5b25dbba03
Properly count nested regions
2014-10-10 14:11:43 +02:00
9df1d1ef2e
re-reference regions during uncombine - #5979
2014-10-10 14:05:49 +02:00
b4cf5e279b
fix dup free in waves/midi:
...
The PortMidi queue just holds a pointer to the midi events.
The same Midi events are free()d during WavesMidiBuffer::clear()
which is called at cycle start for every buffer and in the dtor
of the WavesMidiBuffer.
2014-10-09 09:48:54 +02:00
4af4938c25
fix wavesbackend midi timing
2014-10-09 09:43:53 +02:00
aac3ca5521
initialize some uninitialized vars - closes #5974
2014-10-08 19:16:21 +02:00
8d12ebd972
MSVC - add support for 'HAVE_NEW_LV2'
...
(after updating to the latest versions of serd, sord, sratom and lilv)
2014-10-08 15:11:27 +01:00
50ab3db0ca
When building 'libs/pbd/cpus.cc' (with MSVC) make sure we know what 'PTW32_VERSION' is
2014-10-08 15:11:26 +01:00
2d04ac9796
Add a newly introduced header file to our MSVC project (dummy_audiobackend)
2014-10-08 15:11:25 +01:00
000609901b
add correct copyright statements to all files in Waves backend except those derived from portaudio/portmidi
...
This follows the letter sent from Waves Audio Ltd. to Paul Davis dated February 20th 2014 agreeing to release this code
under the GNU Public License, version 2, with copyright owned by Waves Audio Ltd
2014-10-07 16:17:34 -04:00
118adc286b
fix crashes for plugin using maxBlockLength during latency compute.
2014-10-07 21:48:04 +02:00
ce818efe37
avoid a set of calls to gettext() and/or cousins from global constructor scope
2014-10-07 14:27:12 -04:00
a7da906a79
NOOP, proper label for flat carbon header includes
2014-10-07 16:53:10 +02:00
7bfa25777b
OSX 10.9.5 compile fixes - part one
2014-10-07 14:46:06 +02:00
94e7015c70
64bit windows VSTs
2014-10-05 18:11:11 +02:00
114296d958
fix VST user cache typo: info/blacklist
2014-10-05 02:57:14 +02:00
adefca9dd2
use single newline char in VST info files.
...
windows writes \r\n to non-binary files and the VST
info parser does not handle that case.
2014-10-04 18:01:43 +02:00
29d89f0128
update VST scanner lookup (prepare for moving to BINDIR)
2014-10-04 17:57:05 +02:00
0af091edf3
fix ming32 build&install for asio backend
2014-10-03 00:46:34 +02:00
b376042166
Add UK English translation for libardour
2014-10-02 18:03:19 +01:00
44f71ce599
Text tweaks and typo fixes
2014-10-02 18:03:19 +01:00
8fe3360c35
fix clock_gettime detection
...
fixes compilation with mingw64 gcc-4.9.1
2014-10-02 02:53:35 +02:00
aeda6cbea9
add --with-wavesbackend config option (disabled by default)
2014-10-01 18:28:47 +02:00
3d12a4da82
fix deep obscure problem with loading multiple backends on OS X
...
all backends export a symbol declared as "extern "C" ARDOURBACKEND_API ARDOUR::AudioBackendInfo* descriptor ()".
dlopen'ing the backend apparently pushes the symbol "descriptor" into the single flat global namespace that we use
to be like other unix-like systems. this means that if a backend calls its OWN function named "descriptor", it
is indeterminate which one it will be, since the symbol will refer to the function first loaded by the runtime
linker. If the backend is not the first one discovered, this call to its own "descriptor" function will invoke
the function defined by another backend, even though these are supposed to have local scope only according
to our arguments to dlopen().
This fix doesn't try to fix the linker or namespace - it just makes sure that the WavesAudio backend doesn't
invoke its own descriptor() function, which it never really needed to do anyway.
2014-09-30 22:57:05 -04:00
cd60fd9dfe
untested fix for compiling waves audiobackend on case-sensitive FS with mingw.
2014-10-01 04:02:30 +02:00
9838616679
NOOP, re-indent
2014-10-01 03:25:03 +02:00
32e05fa640
update backends for API change in 8d59afb048
2014-10-01 03:24:54 +02:00
a927b79c26
Implement the new pthread macros (for the Waves audio backend)
2014-09-30 21:00:35 -04:00
3033c1cc54
Re-introduce our pthread macros (after fixing an earlier typo)
...
these are so we can differentiate between 'libpthread' and 'libpthread-win32' (whose implementation is subtlely - though annoyingly! - different)
2014-09-30 21:00:35 -04:00
8d59afb048
copy over current Waves version of their backend, along with minor changes in libs/ardour and libs/backend/jack to fit with API changes
2014-09-30 21:00:34 -04:00