add_conan_lib cmake function is defined, that allows to add a dependency using Conan with two possible system fallbacks:
1. pkg_check_modules is invoked, if `PGK_CONFIG ...` is present
2. find_package is invoked if `FIND_PACKAGE_OPTIONS` is present and `pkg_check_modules` has failed
If `ALWAYS_ALLOW_CONAN_FALLBACK` is present - `obey_system_dependencies` will be ignored for the package
Currently, the following dependencies are retrieved using Conan:
* zlib
* expat
* wxwidgets
* libmp3lame
* libid3tag
* libmad
The last three libraries are included in this commit, as they depend on zlib.
Properly pass **arch** and **os.version** to Conan
A special mode `audacity_obey_system_dependencies` is added, so the build will fail if we could not build against a system package for some reason.
The following packages are marked for the local build now:
* **wxWidgets**: Ubuntu lacks support for 3.1.3. We can't build against 3.0 branch.
* **portaudio**: there are issues, that prevent using a system version of portaudio.
* **sqlite3**: Ubuntu package is very dated; we care about the performance and stability.
* **nyquist**: Ubuntu has no package available.
* **vamp**: Ubuntu has no development package available.
* **portmixer**: Ubuntu has no package available.
* **sbsms**: Ubuntu package is very dated; we care about the performance and stability.
We use docker to create a clean build environment. Currently, `pkg-config` is used to locate the system libraries. There are few issues with `pkg-config` on Ubuntu:
* It does not work with `lame` and `portmidi`.
* The packaged files for `id3tag` and `mad` have wrong version.
We fix such cases by copying the
into `/usr/local/lib`.
... It will be used as a target link library of some lower level Audacity
libraries, to be separated from the executable.
These libraries will have link time dependency only on the wxBase subset of
non-monolithic wxWidgets builds.
More, they are restricted to use only a proper subset of wxBase functionality:
they are prevented, at compile time, not only from using any graphical user
interface, but also from using the main event loop or global application object
-- though these things are also in wxBase.
They may still use things like strings, files, threads, atomics, and other
utilities that may have modern C++ standard library equivalents.
It would be preferable to use those, but it is not a priority to make those
rewrites.
... not assuming that it is invoked from cmake-proxies/CMakeLists.txt
This will allow modules to use lib-src libraries, and even for a module to be
the sole user of one of them.
So addlib lines may be moved out of cmake-proxies/CMakeLists.txt, when no
longer needed to build the reduced executable without the modules.
... So mod-foo defines FOO_API properly on the command line, for use in its
public header files; on Windows, appropriately expanding one way when compiling
the library, another way when compiling other code that uses it, so that all
will link correctly.
... This makes it impossible to forget to include the EXPERIMENTAL definitions
(such as when cutting and pasting code) and so get unintended quiet changes of
behavior.
The EXPERIMENTAL flags are now specified instead in new file Experimental.cmake
Several improvements in determining how much actual disk space a
sampleblock uses. This allows us to provide how much space will
be recovered when using File -> Compact Project.
In addition, the History window now provides better space estimates.
This is highly experimental. It's defers most checkpointing
to a separate thread to see if we get better throughput and
less choppiness when applying effects.
This reenables synchronous mode by default. However, for processing
where a power cut or crash can be tolerated, synchronous and journaling
are disabled. Once the processing is complete, they are reenabled.
Types of processing that are like this are "Save As", "Backup Project",
and "Vacuuming". They all write to a separate project file while
running, so the real project file is safe.
Unfortunately, effects are back to be slow and sluggish.
I believe I've address all of the weird file corruption issues and
I'll continue to continue testing for these.
It now uses VACUUM INTO instead of the SQLite backup API
in hopes that the copies will be smaller. And VACUUM INTO
is "supposed" to be faster, but time will tell. It's easy
to put the backup API usage back in.
This also fixes a bit I missed with redoing the orphan block
handling that was reported by Paul.
And finally, it renames the AutoRecovery.cpp/.h files and AutoSaveFile
class to ProjectSerializer since the AutoSaveFile class is being
used for regular project documents now and it doesn't write to a
file anymore.
If anyone has a better idea for a name other than ProjectSerializer
feel free to change it. I hate naming things.