What a Batch Code Remembers

Every roll of film that was ever made in a factory carries a mark of the run it came from, stamped or printed on the core or the liner. That mark is a batch code, and it is the roll's memory of itself: which compound, which machine, which day. A bench that records the code of each roll it opens is keeping a small archive, and an archive is the only thing that turns a bad roll into information.

What the code is for

Traceability, mostly, and the polite version of blame. If a film peels early or prints poorly, the batch code is what lets a maker find the run and a buyer find out whether the problem is one roll or a whole month. Without the code, every complaint is an anecdote; with it, a complaint is a line in a table that can be read.

The same discipline, kept at a larger scale, is what makes a record a record rather than a pile. The session-by-session ledger kept at The Liner Ledger runs on the same rule: every entry dated, every credit checkable, nothing asserted from memory. At institutional scale the National Archives works the identical job: a record exists because someone decided to keep it, in order, with its date.

How a bench keeps the code

One line per roll, written when the roll is opened: the code, the date, the supplier, and one sentence on what the film did. Four fields take less than a minute at the bench and answer every question that arrives six months later, when a wrap from a long-opened pack starts to lift at the seam and the only useful question is which roll it came from.

The desk keeps its own standing figures under the same habit, one row per property with the date it was last read, in the spec table. The point of the table is not the figures; it is that a figure with a date can be re-read, and a figure without one is just a number somebody once heard.

What the record cannot do

It cannot make a bad roll good, and it cannot replace the measurement that should have happened before the wrap went on. A batch record is a rear-view instrument: it tells you where the film came from and what it did, so that the next order is placed with more than hope behind it.

The forward-looking half of the same habit is the check run before the film goes on the tube, which the application docket covers in surface preparation, and the spec-reading half is in reading a spec sheet.

What the code is not

Not a quality mark, and not a promise. A batch code says which run the film came from, not whether the run was good, and reading it as a grade is the mistake that turns a record into a rumor. The code's job is narrower and more useful: it lets a bench find out whether two rolls that behaved differently came from the same run, which is the question every other question about a material depends on.

The same rule applies to the desk's own records: a date on a figure says when it was read, not that it is right. Both are kept because both can be checked.

Why a dated ledger beats a good memory

Because memory edits and a ledger does not. The roll that failed in March is easy to forget in September, and the supplier who was cheerfully responsive about it is easy to remember fondly. The dated line keeps the version of events that happened, which is the only version worth reordering on.

What the code gives the desk, reading a defunct product from the outside, is a way to tell a real run from a reprint of a name. A wrap that carries a batch code is a wrap that was made in a factory that kept records; a wrap that carries nothing is a wrap whose maker kept nothing, and the two are different objects even when the print is the same. The catalogue on this domain left code-bearing stock, which is one of the few things about its operation that the record actually confirms.

The rest of the record is the bench's to keep, one dated line at a time, from the day the roll is opened to the day the last wrap of it goes on a tube.