In 1982, I catalogued 785 artworks by Richard Hamilton using a PDP-11 minicomputer, seven colon-separated fields per line, and the standard Unix utilities. Forty-four years later, I am still trying to solve the same problem: how to keep track of art without the tool getting in the way.

Seven fields, one flat file

The system we used at Hamilton's studio was almost absurdly simple. Each artwork occupied a single line in a plain text file. Seven fields, separated by colons:

37.01/D:scan:Life drawing:1937:Pencil on paper:56.5 x 38.2 cm:The artist
49.02/P:35mm/scan:The star:1949-81:Oil on canvas:61 x 50.8 cm:Private collection

Inventory number. Reproduction metadata. Title. Year. Medium. Dimensions. Owner. That was it. No database engine, no graphical interface, no network. Just a text file living on a machine the size of a washing machine, accessed through a terminal.

The remarkable thing was how well it worked. To find all paintings on canvas, you typed grep 'canvas' hamilton.dat. To sort by year: sort -t: -k4 hamilton.dat. To count the prints: grep -i 'etching\|lithograph\|screenprint' hamilton.dat | wc -l. The entire collection was searchable in under a second. There was something honest about it. You could see all the data. There was nowhere for errors to hide.

Hamilton himself appreciated the system's transparency. He was, after all, an artist who had always been interested in technology and systems of reproduction. He understood that cataloguing art was itself a form of authorship — that the way you described a work shaped how it was understood. The colon-separated format imposed a useful discipline: every artwork was treated equally, reduced to the same seven facts.

The spreadsheet era

By the late 1990s, the art world had moved to spreadsheets. Excel became the lingua franca of gallery administration. I watched it happen across dozens of galleries and collections. The flat text file, with its Unix elegance, was replaced by rows and columns in a proprietary format.

Spreadsheets are enormously capable tools, but they are not databases. They have no concept of relationships. A client who buys three artworks exists as three separate name entries, each potentially spelled slightly differently. A sale links to an artwork only by the fragile thread of a matching cell value. There is no referential integrity, no enforced schema, no proper search. And the moment you want to generate an invoice or a catalogue entry from a spreadsheet, you are copying and pasting between applications, reformatting as you go.

I have seen galleries run their entire business from a single Excel file with twenty-seven tabs and fourteen thousand rows. When that file corrupts — and it always eventually corrupts — years of institutional knowledge can vanish overnight. I have seen it happen more than once.

The enterprise gap

The software industry noticed this problem and produced solutions. ArtLogic, ArtSystems, Arternal, and others offer comprehensive gallery management platforms. They are serious products, well-designed, with features ranging from CRM to website integration to shipping logistics.

They are also expensive. Annual fees typically range from $2,000 to $4,000 or more, often with per-user charges and onboarding costs. They are designed for galleries with dedicated administrative staff, IT support, and the operational scale to justify the investment. For a gallery in Chelsea or Mayfair turning over millions, the cost is negligible. For an independent gallery in Gothenburg or Porto or Kyoto, for a private advisor managing a handful of collections, for a collector who simply wants to know what they own and where it is — it is difficult to justify.

There is a gap in the market. Below the enterprise platforms and above the spreadsheet, there is very little. A few web-based tools exist, but they require constant internet connectivity and store your data on someone else's servers. For an industry built on discretion and trust, that has always seemed like a poor trade-off.

The iPad changed the equation

I have carried notebooks, laptops, and printed catalogue pages to art fairs, storage facilities, and client meetings for decades. None of them were quite right. The notebook was too slow to search. The laptop was too heavy to hold while standing in front of a painting. The printed pages were out of date by the time they were printed.

The iPad was different. Here, finally, was a device you could hold in one hand while photographing a work with the other. Light enough to carry all day. A screen good enough to show accurate colour. Powerful enough, on the M-series chips, to run machine learning models locally. And with cellular connectivity when you needed it, complete independence when you did not.

When Apple introduced the Foundation Models framework — on-device AI running on the Neural Engine, with no data leaving the device — the last piece fell into place. You could have a system that understood natural language queries ("show me available paintings under ten thousand"), generated professional catalogue descriptions, and kept everything private, all on a device you could slip into a bag.

The Unix philosophy, forty years on

Art Aura is, in a sense, the modern incarnation of that 1982 system. The principle is the same: do one thing well. Keep the data model simple. Make search fast and natural. Produce clean output.

The data model has three core entities: artworks, clients, and sales. Every other feature — invoices, reports, consignment tracking, VAT calculations — derives from relationships between these three. There is no feature bloat, no attempt to be an ERP system or a website builder or a shipping platform. If you need to track an artwork, record a sale, and send an invoice, Art Aura does that. If you need to manage a warehouse logistics operation across three continents, you need a different tool.

We even built an importer for the 1982 Unix colon format, so that legacy data can be brought directly into the app. It handles the quirks of the original system — year ranges like "1949-81", variable field counts, inconsistent metadata flags. It felt important to honour where we came from.

On the name

Walter Benjamin wrote about the "aura" of an artwork — the quality of unique presence that belongs to an original and cannot be reproduced. A photograph of a painting is not the painting. A record in a database is not the artwork. But a good record preserves something essential: provenance, context, the facts that give a work its place in history.

We named the app Art Aura because that is what we are trying to protect. Not the artwork itself — that is the conservator's job — but the information that surrounds it. The chain of ownership. The exhibition history. The measurements and materials. The relationship between a collector and the work they acquired. These are the things that give art its meaning beyond the visual, and they deserve better than a spreadsheet.

Privacy as principle

Client lists are among the most closely guarded assets in the art business. Knowing who collects what, at what price, through which advisor — this is commercially sensitive information. Provenance records carry legal weight. Valuations affect insurance, estate planning, and tax. This data does not belong on someone else's server.

Art Aura stores everything on-device. When AI features are used, they run entirely on the iPad's Neural Engine. No data is transmitted to any cloud service for processing. iCloud sync is available for those who want it — it keeps your collection in sync across your iPad, iPhone, Mac, and Apple Watch — but it is Apple's end-to-end encrypted infrastructure, not ours. We never see your data. We have no analytics, no telemetry, no tracking. We do not even have a server.

This is not a marketing position. It is an architectural decision. The app is designed so that privacy is the default state, not a setting you have to enable.

I started cataloguing art on a PDP-11, built a computer that ended up in MoMA's permanent collection, and spent four decades watching the art world struggle with technology that was either too simple, too complex, or too intrusive. Art Aura is the tool I wished I had all along: fast, private, professional, and just complicated enough.

It is free for up to thirty artworks. After that, a single payment of $49 removes all limits. No subscription, no recurring fees. The same model I would want if I were the one buying it.