Linked data

Why link, not just type

A name is a label. An address is a fact. The difference shows up the moment there is more than one of something — which, in music, is always.

• Same title, different song. “Amazing Grace” is eleven distinct works in MusicBrainz: the hymn, a Boy George song, several arrangements that are their own compositions. Typing the title says what you called it. Attaching the identifier says which one you played.
• Same name, different thing. Search Kalotaszeg and you get the folk region, a periodical of the same name, and an embroidery style. Two composers share a name more often than you would like. The identifier tells them apart, permanently, without you writing “(the Hungarian one, 1882–1967)” in a field meant for a name.
• Your library becomes answerable. Once charts carry identifiers rather than spellings, “everything we play from Mezőség” is a question with an answer — even when one chart says Kodály Zoltán and the next says Zoltán Kodály.
• It survives you moving on. The address resolves whether or not GigCharts is still around, in any app, in ten years. That's the point of using someone else's public identifier instead of inventing our own.
• It's free provenance. A chart that can point at where its song and its credits come from is a better citizen than one that can't — especially for material you share, publish or hand to a choir.

The two databases

Both are open, both are free, both are built by volunteers, and neither needs an account or an API key. GigCharts queries them straight from your browser.

MusicBrainz — for the music

MusicBrainz is an open encyclopedia of music — songs, the people who wrote and played them, the records they came out on — built since 2000 and run by the non-profit MetaBrainz Foundation. Its core data is public domain. It's the reference a lot of the music world quietly runs on, from streaming services to libraries to Wikipedia.

Wikidata — for the world around it

Wikidata is Wikipedia's structured sibling: the same movement, the same volunteers, but facts instead of articles — a hundred million things, each with an identifier, described in hundreds of languages at once. Its data is public domain (CC0) too.

It's here because a chord chart is not only about music. A song comes from a place, is sung at a time of year, in a language — and MusicBrainz, quite reasonably, has never heard of Kalotaszeg. Wikidata has: Q832218, described in Hungarian as tájegység Romániában, részben Erdélyben. Mezőség is Q1058238. Hungarian the language is Q9067.

That multilingual habit is the practical reason as much as the philosophical one: Wikidata searches and labels in whatever language you are working in, so a Hungarian user searching Karácsony gets Hungarian descriptions to choose between, not English ones.

Which field asks whom

Every one of them behaves identically: type and the matches appear; click the link glyph for the full search panel; the panel names the database it is asking, and so does the list of captured identifiers at the foot of the editor.

How it works

• The “By” box does the heavy lifting on song searches. A hymn title alone can return several hundred works; adding a writer's name usually cuts that to a handful.
• You see before you commit. A result shows its writers, its type, its ISWC — or, on Wikidata, the one-line description that separates the region from the periodical. That is the reason this is a panel and not a longer dropdown.
• It finds the song, not a recording of it. A chart is of a composition. Search recordings — the escape hatch for songs MusicBrainz has performances of but no song entry for — and a pick resolves through to the underlying work, or stores nothing.
• Credits are offered, never imposed. The writing credits are listed with a button each: put this name in Music, in Lyrics, in Arranger. Where MusicBrainz only says “writer” without saying who wrote what, we show you that rather than guessing.
• One identifier per field. Edit the field by hand afterwards and its identifier is dropped, because an address that no longer matches the name beside it is worse than none at all.
• Some lists are shipped, not searched. Genres and languages come bundled with the app — MusicBrainz publishes no genre search, and language names come from your browser's own data — so those two are instant, work offline, and still carry real identifiers.

Where the link travels

The identifiers you attach are listed at the foot of the Song editor, grouped by the database they came from and each one clickable. From there they go with the chart:

• Into the .gigchart file — mbTitleUri, mbArtistUri, mbMusicUri, mbLyricsUri, mbArrangerUri, mbInstrumentUri, mbGenreUri, mbAlbumUri, regionUri, holidayUri, languageUri.
• Into the exported PDF's document metadata, so provenance survives the one format everybody actually emails around.
• Into the published vocabulary as JSON-LD, where they line up with schema.org terms — so a .gigchart file is linked data, not just our JSON.

The mb… prefix marks the ones that are MusicBrainz by definition. The other three are named for the field instead, because a place could as reasonably be identified by GeoNames or the Getty Thesaurus one day — and the address itself always says which authority answered.

When it finds nothing

Often. Traditional material, folk songs, most hymnody, anything regional — a great deal of it is simply not in MusicBrainz, and that is not a fault in your song or in the search. A chart with no identifiers is a perfectly good chart; leave the field as plain text and move on.

Or don't. Anyone can add an entry to either database. The song you add is the one the next person finds — and that is the only reason there is anything in there to search in the first place.

Thank you

Every result this feature shows you was typed in by a volunteer who wanted the world catalogued properly and got nothing for it. GigCharts queries that work, credits it, and pays nothing for it — so the least we can do is say plainly where it comes from: the MetaBrainz Foundation and the MusicBrainz community, and Wikimedia and the Wikidata editors. Thank you, both.

If this is useful to you, consider supporting MetaBrainz or Wikimedia — or adding the entries you find missing, which costs nothing but your afternoon and helps everyone after you.

Open the app