sys::basename now matches boost::filesystem behaviour so if a session name
had a dot in it the session name used in the interchange directory would be
truncated.
git-svn-id: svn://localhost/ardour2/trunk@2408 d708f5d6-7413-0410-9779-e7cbd77b26cf
This makes this function equivalent to PBD::basename_nosuffix, sys::basename
takes a path as an argument rather than a string but that is ok as a path can
automatically constructed from a string.
git-svn-id: svn://localhost/ardour2/trunk@2404 d708f5d6-7413-0410-9779-e7cbd77b26cf
This is basically equivalent to Glib::path_get_dirname but returning an
empty string(path) if a path doesn't contain any directory components rather
than "."
git-svn-id: svn://localhost/ardour2/trunk@2403 d708f5d6-7413-0410-9779-e7cbd77b26cf
This is equivalent to the fix applied as rev 2244 to the 2.0-ongoing branch
git-svn-id: svn://localhost/ardour2/trunk@2402 d708f5d6-7413-0410-9779-e7cbd77b26cf
Defining it using std::numeric_limits<layer_t>::max() would of avoided
this inconsistancy but as it unused it is doesn't really matter I guess.
git-svn-id: svn://localhost/ardour2/trunk@2392 d708f5d6-7413-0410-9779-e7cbd77b26cf
Replace one line C style comments with C++ equivalent in Session::remove_state
git-svn-id: svn://localhost/ardour2/trunk@2378 d708f5d6-7413-0410-9779-e7cbd77b26cf
Basically moving PBD::copy_file implementation to pbd/filesystem.h/cc.
The implementation itself looks like it could be improved to use
much less memory when copying big files by reading and writing in
chunks but I don't think that is an issue at present.
git-svn-id: svn://localhost/ardour2/trunk@2376 d708f5d6-7413-0410-9779-e7cbd77b26cf
API is intended to be indentical(apart from the string type) to boost::filesystem::remove
git-svn-id: svn://localhost/ardour2/trunk@2372 d708f5d6-7413-0410-9779-e7cbd77b26cf
Extremely broken in several ways.
This commit brought to you by the letters D, R, and my need to switch machines.
git-svn-id: svn://localhost/ardour2/trunk@2323 d708f5d6-7413-0410-9779-e7cbd77b26cf
Read from MidiRingbuffer directly into model, don't read MidiRingBuffer into a new midi buffer, then into the model.
Pass rec data to UI via model instead of a separate buffer.
Read MIDI CC data into MidiModel (though not actually used yet).
Made quantization toggle edited flag so model is saved.
git-svn-id: svn://localhost/ardour2/trunk@2308 d708f5d6-7413-0410-9779-e7cbd77b26cf
Shave a few bytes off each ControlPoint if not used by a Curve (24 bytes on 64-bit).
git-svn-id: svn://localhost/ardour2/trunk@2294 d708f5d6-7413-0410-9779-e7cbd77b26cf
Seems to be a pretty random problem with note duration restoring though...
git-svn-id: svn://localhost/ardour2/trunk@2290 d708f5d6-7413-0410-9779-e7cbd77b26cf
Fix shutdown crash (one of many, namely the least nasty).
Failed attempt at getting MIDI edits to apply ASAP (Diskstream::playlist_modified isn't the magic I'd hoped it was).
git-svn-id: svn://localhost/ardour2/trunk@2265 d708f5d6-7413-0410-9779-e7cbd77b26cf