Summary

  • SIMBAD's first web interface arrived in November 1996. Its November 1999 snapshot held about 2.7 million objects and 7.5 million identifiers.
  • Name-format rules handled familiar variations, while ambiguous abbreviations needed context. Bibliographers, specialists and a correction desk kept the associations useful.

The universe had plenty of room. Astronomers kept running into each other over the letter N.

In their 2000 account of SIMBAD, Marc Wenger and colleagues offered the query N 1992 as a problem requiring more information from the user. N could stand for NGC in one context, Nova in another, and occur in several other naming schemes. An impressively short name had omitted the bit that would tell the database what it meant.

This was awkward territory for a service built to bring names together. In November 1999, SIMBAD held about 2.7 million astronomical objects and 7.5 million identifiers. M 87 was among the objects with more than fifty identifiers. One galaxy could arrive with the paperwork of a small delegation.

A web form on top of a long memory

SIMBAD organised information around the astronomical object. Papers and catalogues could call it different things; cross-identification brought those names and references into the same record. Searching an alias could then lead to the object's coordinates, observations and bibliography.

The service's interactive online history reached back to 1981. The CDS team's companion history dates its first web version to November 1996. The browser provided an easier doorway into work that had been accumulating for years. Other archives could also use client/server access to turn an object name into a position.

Names came with grammar. SIMBAD's identifier guide describes a catalogue label followed by a designation within it. The internal sesame module described in the 2000 paper used rules to convert common variations into a recognised form. Tidying a label and establishing that two observations belong to one object were separate jobs.

Somebody had the correction duty

The paper describes daily bibliographic work by CDS and collaborators in Paris and Bordeaux, covering roughly a hundred astronomy journals. Large catalogue integrations could take months or years. Each new name had to find its proper relationship to what was already recorded.

Users also sent corrections. A few dozen messages a week reached the service, according to that contemporary account. A rotating duty person handled them each day or passed specialised questions to an expert. The neat search box had a rota behind it.

That work mattered beyond the immediate result. A wrong association could attach the wrong literature to a query; a useful association could unite observations scattered across names and wavelengths. Once another archive relied on the answer, the quality of the link travelled with it.

SIMBAD's small request for more context was therefore part of its competence. The universe could keep its aliases. The person typing N still had to say which N.