Do We Still Need Ross?

Elaine Gould's Behind Bars is here. Ted Ross's The Art of Music Engraving and Processing isn't, because there's nowhere to order it from. The print edition has been dead for decades. Hansen Books no longer functions in any meaningful commercial sense. Ted Ross himself is dead. Second-hand copies surface occasionally on AbeBooks or eBay, usually in poor condition, usually at absurd prices, usually spoken for within hours. The Internet Archive lends a scanned copy in twenty-four-hour slots. A Californian outfit called npc Imaging has sold a searchable CD-ROM edition since 2001, though whether it still runs on a modern operating system is an open question.

Which provokes the obvious one: do I still need Ross in 2026?

The temptation to answer no was real. The book was published fifty-six years ago. It describes a workflow of punches and gravers, of metal plates and photographic reproduction, none of which has survived the move to digital notation. Gould arrived forty-one years later, covers what publishers currently expect, and is in print.

But no. I need it.

Gould and Ross overlap, but their centres of gravity differ. Gould's focus is editorial: contemporary conventions, edge cases in modern repertoire, the negotiated practices that settle disputes between composer and copyist. Her domain is current, and her vocabulary includes extended techniques and the Ferneyhough generation's demands.

Ross's focus is geometric. Stem lengths with actual measurements. Beam angles. Optical spacing tables. Ledger line dimensions. The precise relationships between noteheads, stems, beams, and staff lines, specified with the exactitude of someone producing real plates. His specifications are tabular and implementable rather than descriptive. Ross becomes data in Ooloi; Gould has to be read and decided.

The invariance argument settles the age question. The five-line staff hasn't changed since 1970. A half note still looks the same. The geometric relationships Ross specifies have the same validity now as when he wrote them, because notation itself hasn't shifted underneath them. What dates in Ross is the production workflow, which I can skip entirely without losing anything I need.

LilyPond's spacing algorithms reference Ross directly. Gourlay's seminal paper on music spacing builds on him. The rendering pipeline I'm about to close – spanner geometry, optical adjustments, beam slope selection – sits more in Ross's territory than Gould's. Not consulting him would be working blind on problems he addressed head-on half a century ago.

So I need the book, and I can't buy it.

There's something grimly appropriate about this. Ross's book is the kind of reference work whose commercial half-life expires decades before its technical half-life even begins to decay. The market that would sustain a reprint is smaller than the importance of the content. LilyPond's continued citation of Ross is, in a sense, more preservation than any library has managed. The notation community keeps him alive by quoting him, not by reading him.

It's also, not incidentally, one of the reasons Ooloi's documentation looks the way it does. The normal fate of a specialist reference work is exactly what happened to Ross: the author dies, the publisher dissolves, the copies scatter, and fifty years later someone trying to do the work has to borrow the book in twenty-four-hour slots from a scanner in San Francisco.

None of which amounts to preservation.

And even if it did, neither book would be the final word. They don't fully agree with each other, because the engraving community and major publishers don't fully agree with each other. Ross, despite devoting some fifty pages to beaming, contradicts himself more than once within that span. This isn't a failing. Engraving is a living craft with regional traditions, house conventions, and genuine disagreements about what looks right; two books documenting it honestly will reflect those disagreements rather than paper them over.

Which means the rendering pipeline isn't a matter of looking up The Answer. It's a matter of taking positions in ongoing debates, and making the positions configurable wherever taste genuinely differs.

Ooloi's formatting is data-driven wherever it can be. Engraving rules live as editable values, not as switch statements buried in source code. Hard-coding them treats them as invariant when they aren't, and turns house styles into an act of patching rather than configuration. Which is how the real disputes get handled: Boosey & Hawkes flat beams against sloped ones, whether grace note beams separate from or share the main beam, whether a slanted beam group shears or rotates. Ooloi will rotate.

I'm looking forward to that.