Category Archives: Development

About next stable Picard release

Picard team decided to release a new stable version of Picard before the end of the summer.
To help us, advanced users, translators and developers are encouraged to:

Exact release date is not yet known, and features aren’t yet frozen, but the only important change we’ll try to get in before final release is OAuth support (PICARD-305), we still accept minor patches until feature freeze is announced.

Since Picard is now using some translation resources from MusicBrainz project, translators should make sure following resources are up-to-date:

A simplified list of changes made since 1.2 can be read there.

Be aware that downgrading from 1.3 to 1.2 may lead to configuration compatibility issues, better save your configuration before using 1.3 branch if you intent to go back to 1.2.

 

 

MHD San Francisco 2014

MusicBrainz is represented at Music Hack Day San Francisco 2014 this weekend!

If you’re a hacker who is participating, this page collects a number of interesting things about MusicBrainz that lets you get started with MusicBrainz easily.

If you’d like some help with MusicBrainz, come see Rob in the the back of presentation area. Rob is the guy with the crazy hair and the bright orange t-shirt.

Have fun!

Our embedded RDFa will be going away in two weeks

We’ve been trying to find anyone making use of the RDFa embedded in our pages, but we’ve been unable to find anyone.

Given that the RDFa makes our templates much harder and cumbersome to edit, we’ve decided to go ahead and remove our RDFa support in two weeks time.

Since we could not find anyone who actually uses our RDFa, this shouldn’t be a problem. But we suspect that someone, somewhere will be angry with us for removing the RDFa support without any warning or without being asked for feedback. :-/

Announcing python-musicbrainzngs, release 0.5

We’ve just release a new version of python-musicbrainzngs, a library for accessing the Musicbrainz webservice from python.
Version 0.5 comes with a lot of new features, including:

  • Support for URL entities
  • Support for Area entities
  • Support the experimental JSON webservice
  • Support fuzzy disc lookup by TOC
  • add a -count element to browse and search requests
  • added support for link type UUIDs
  • deprecated puid and echoprint support

And a number of bug fixes and other small changes. See the CHANGES file  for more information.

For consistency with online repositories and distribution packages, we’ve renamed some URLs. You can now find the library on github and readthedocs at:

You will be automatically redirected if you access the old urls.

Thanks to Johannes Dewender, Ian McEwen, Wieland Hoffmann, Simon Chopin and Ryan Helinski for contributing to this release.

The new version can be downloaded from github, pypi, or installed with pip

python-discid 1.1.0 (and 1.0.3) released

New versions of python-discid are available.

Changes in 1.1.0

  • feature: #36 add Disc.toc_string
  • #38 remove DiscId (deprecated since 0.5.0)

Changes in 1.0.3

  • fix: #37 test_emptyness: Assertion disc->success failed

Python-discid 1.0.3 is a bugfix version for the recent libdicid 0.6.0 release and python-discid 1.1.0 adds the TOC string, which is also part of new libdiscid versions. The TOC can be used for fuzzy searches in the web service.
The toc string feature was backported and can be used together with any version of libdiscid >= 0.2.2.

The bugfix for 1.0.3 is only relevant for running the test suite (which is done at install time usually). It shouldn’t affect normal operation. If you don’t have any problems, you don’t need to upgrade.

The DiscId class was already deprecated in the 0.5.0 beta, but still works in 1.0.x releases.
The correct usage is to create a Disc object with read() or put().
See also basic usage of python-discid.

Downloads are available at pypi. Packages are available for some linux distributions.
Recent stable and git versions of python-discid and other MusicBrainz client libraries and tools are now available for many Ubuntu versions at the MusicBrainz Stable and https://launchpad.net/~musicbrainz-developers/+archive/daily PPAs.

libdiscid 0.6.0 (and 0.6.1) released

This is a feature release of libdiscid.

It took more time than planned, but 0.6.0 is done now with lots of changes.
Main changes are the TOC string api (handy for fuzzy lookup) and a much better default drive detection on all platforms.
An example to use libdiscid together with libmusicbrainz (including fuzzy lookup) is also provided now.

Unfortunately I found one major bug only while merging changes to another branch. Libdiscid 0.6.1 fixes reading of ISRCs on Windows again (introduced in 0.6.0).

The full set of changes:

Bug

  • [LIB-48] – Fix asserts/return values on empty object
  • [LIB-59] – windows: unable to read ISRC (0.6.1)

Improvement

  • [LIB-52] – validate parameters for put()
  • [LIB-53] – don’t use old redirect for submission_url

New Feature

  • [LIB-28] – allow using numbers as devices on Mac OS X
  • [LIB-41] – Add TOC (id) to API
  • [LIB-43] – Use first CD drive letter on Windows as default device
  • [LIB-45] – Linux/BSD/Solaris: search for an existing default device from a list of devices
  • [LIB-54] – Add example: libmusicbrainz / query for metadata
  • [LIB-55] – Allow using numbers as devices on Windows
  • [LIB-56] – Allow using device numbers on Linux

I want to thank Philipp Wolfer for the Windows drive detection, Laurent Monin for the TOC string API and also Sebastian Ramacher and Lukáš Lalinský for smaller changes.

Testing wanted for raw ISRC reads

Reading ISRCs raw from the device with SCSI commands ([LIB-11, duplicate ISRC problems) is already implemented for Linux and Windows and works quite well on most drives. There is no code for support on Mac, yet. The Windows and Linux implementation still need some testing (with different drives).

If you are willing to test this or really need this feature you can build the code in the “isrc_raw” branch from the github repository.
There is also a windows binary available.
We do want to know about problems (some drives are already reported with strange behavior, possibly due to Alignment), but also a short note about working systems is helpful to get a bit of insight on how rare or frequent failures are.

We also happily accept patches to send SCSI commands on Mac OS X. There is some information in the ticket how this could work, but this is probably the most difficult of the platforms to implement.

Information, documentation and other links are at:
http://musicbrainz.org/doc/libdiscid
That includes builds for Windows and Mac OS X.

libdiscid 0.5.2 released

This is a bugfix release of libdiscid.

There are already lots of changes (features) ready for 0.6.0, but I did want to release these fixes separately, since they have a high impact on the platforms involved.
0.6.0 is probably ready in a couple of weeks.

Bug

  • [LIB-50] – Solaris: segfault in mb_disc_load_toc
  • [LIB-51] – Mac: Unable to read ISRC for track

New Feature

  • [LIB-26] – disc access test suite
  • print time information in the discid example (zas)

Information, documentation and other links are at:
http://musicbrainz.org/doc/libdiscid
That includes builds for Windows and Mac OS X.

.NET / Mono bindings for libdiscid

There has been quite some activity regarding libdiscid in the last few months, here is some more: I have released dotnet-discid 0.2 which provides .NET / Mono bindings for libdiscid. The source is available at GitHub. For easy usage Windows users might want to use the NuGet package. Alternatively you can download source and binary releases at http://users.musicbrainz.org/~outsidecontext/dotnet-discid/. To use the assembly you will also need the native libdiscid release (e.g. discid.dll on Windows or any other native package for your platform). On Unix-like platforms you can just use the system wide installation of libdiscid.

No API docs yet, but the examples on Github will get you started. The library should be straight forward to use. This is an early release but it supports all the features provided by the latest libdiscid 0.5 (while still supporting earlier versions).  I don’t expect many changes in the API but I will provide better documentation soon.

ruby-discid 1.0

This has already been announced on the mailing list back in May: ruby-discid, the Ruby bindings for libdiscid, have been released in version 1.0. This libdiscid binding for Ruby is the successor of mb-discid. As libdiscid has gained quite a few features in the past I took the opportunity to simplify the API and do some fundamental changes to ruby-discid.

The most important features of this release are:
  • Full support for libdiscid 0.1 to 0.5, including ISRC and MCN reading and feature detection.
  • Uses FFI to load libdiscid, so no compilation is required.
  • Support for Ruby 1.8.7 – 2.0, jRuby and Rubinius.

The probably easiest way to install and use ruby-discid is by using the Ruby gem. For Ubuntu users there is a PPA and for Arch Linux users an AUR package available.

The source code is available on Github, there you will also find more detailed installation and usage instructions.

If you have previously used mb-discid you can continue using it, but you should consider updating to ruby-discid. The API is quite similar and simple enough to make that a painless process.

python-discid 1.0.0 (and 1.0.2) released

In februar I announced a beta version of the first python binding of libdiscid. A couple months have passed and I can now announce the first stable release of python-discid.

The main purpose is the calculation of MusicBrainz Disc IDs from a CD-AUDIO disc or a TOC of such a disc, but reading of ISRCs and the MCN (EAN/UPC) are also supported.
Disc IDs can be submitted without further dependencies. For lookup and ISRC submission it is recommended to use it together with python-musicbrainzngs.
Python 2 >= 2.6 or Python 3 >= 3.1 as well as libdiscid >= 0.2.2 are required. Newer libdiscid features need a newer libdiscid, but there are methods available to check for these conditions.

If you still have a program using python-musicbrainz2, I recommend upgrading to the combination of python-discid and python-musicbrainzngs, since python-musicbrainz2 uses a deprecated web service and is not actively maintained anymore.

The main website and API documentation is at readthedocs, the repository and bugtracker are at GitHub.
Official source tarballs are available on pypi and releases are announced in this blog. There are several linux packages available and listed in the documentation.

If you have been using a pre-release version of python-discid you will have to update your code.
Compared to python-discid 0.5.0 you now have to use discid.get_default_device(), rather than discid.DEFAULT_DEVICE. If you were using 0.4.0, you should also stop using DiscId() directly or a DeprecationWarning is displayed.

There is also a patch release 1.0.2 available, which has no code changes compared to 1.0.0.
There was an intermediate 1.0.1, which tried to make some convenience changes for beta users, but failed. 1.0.1 was never available as a tarball, but the tag and changelog exists.