annotate CHANGELOG @ 44:aa521baace07

* Add test for output destination when input filename has more than one dot
author Chris Cannam
date Mon, 18 Oct 2010 14:16:17 +0100
parents 6e41464262f9
children 6c87f6443fe6
rev   line source
Chris@36 1
Chris@36 2 Changes in Sonic Annotator 0.5 since the previous release 0.4:
Chris@36 3
Chris@36 4 Build changes:
Chris@36 5
Chris@36 6 - Remove unused audioio library and avoid its dependencies
Chris@36 7
Chris@36 8 Front-end changes:
Chris@36 9
Chris@36 10 - Avoid leaking file descriptors on exceptions
Chris@36 11 - Fix embarrassing bug that caused failure to mix down to mono
Chris@36 12 properly for single-channel plugins with multi-channel input files
Chris@36 13 - Fail sooner if the output file is not writable (i.e. don't wait
Chris@36 14 until the first data is available for writing)
Chris@23 15
Chris@27 16 Changes in Sonic Annotator 0.4 since the previous release 0.3:
Chris@27 17
Chris@27 18 Build changes:
Chris@27 19
Chris@27 20 - Sonic Annotator now requires the Vamp plugin SDK v2.1 or newer.
Chris@27 21
Chris@27 22 Front-end changes:
Chris@27 23
Chris@28 24 - Frequency-domain plugins now use the PluginInputDomainAdapter's new
Chris@28 25 ShiftData processing method, ensuring that the first block received
Chris@28 26 by the plugin is the one centred on 0 rather than starting at 0.
Chris@28 27 Unfortunately, the old behaviour omitted a timing compensation step,
Chris@28 28 and this fix actually changes the results from some plugin processes:
Chris@28 29 the previous behaviour was not always in line with the Vamp plugin
Chris@28 30 specification.
Chris@28 31
Chris@27 32 - More useful diagnostics are now available when a plugin fails to
Chris@28 33 load or run.
Chris@27 34
Chris@27 35 CSV back-end (feature writer) changes:
Chris@27 36
Chris@27 37 - The CSV writer now closes its output files properly when it
Chris@27 38 finishes writing to each one, rather than leaving them all open
Chris@27 39 until the end and risk running out of file descriptors. This was
Chris@27 40 already the behaviour of the RDF writer, the CSV one has just been
Chris@27 41 updated to match it.
Chris@27 42 - The behaviour of the CSV writer has also been changed to match
Chris@27 43 that of the RDF writer in handling file write failures (it now
Chris@27 44 continues processing only if --force is given).
Chris@27 45
Chris@27 46
Chris@27 47 Changes in Sonic Annotator 0.3 since the previous release 0.2:
Chris@23 48
Chris@23 49 Front-end changes:
Chris@23 50
Chris@23 51 - A new --force front-end option (distinct from the --csv-force and
Chris@23 52 --rdf-force back-end options) has been added, which makes Sonic
Chris@23 53 Annotator continue to process subsequent audio files instead of
Chris@23 54 exiting after an error, if multiple audio files have been specified.
Chris@23 55
Chris@23 56 RDF back-end (feature writer) changes:
Chris@23 57
Chris@23 58 - The RDF writer now conforms more correctly with the Music Ontology
Chris@23 59 in the way it relates signal, track, and track metadata. Signals
Chris@23 60 that have available metadata now gain a Track resource to associate
Chris@23 61 that metadata with, instead of hanging it directly from the Signal
Chris@23 62 resource; also the audio file now encodes the signal rather than the
Chris@23 63 signal being available as the audio file (matching the domain
Chris@23 64 specification of the Music Ontology). Note that this new resource
Chris@23 65 structure will not be properly read by versions of Sonic Visualiser
Chris@23 66 prior to 1.6 (should importing the data into Sonic Visualiser be of
Chris@23 67 interest to you).
Chris@23 68
Chris@23 69 - The --rdf-signal-uri option has been removed and replaced with the
Chris@23 70 more meaningful set of options --rdf-audiofile-uri, --rdf-track-uri,
Chris@23 71 and --rdf-maker-uri.
Chris@23 72
Chris@23 73 - A new --rdf-network option is available, to cause Sonic Annotator
Chris@23 74 to try to retrieve RDF descriptions for plugins from the network
Chris@23 75 where those descriptions are not available locally. Use of this
Chris@23 76 option is recommended, but it is not the default because of the
Chris@23 77 possible performance implication (even though the results are mostly
Chris@23 78 cached, there may be some network access involved).
Chris@23 79
Chris@23 80 - The RDF writer now writes the computed_by property for signal
Chris@23 81 features.
Chris@23 82
Chris@23 83 - Plugin and output URIs in the resulting RDF are now percent-encoded.
Chris@23 84