# The Marrow Tape

A DC300-format quarter-inch cartridge, grease-pencil label, no case. It has been
in the hands of fourteen people since 1995 and is currently held by the person
who received it at the 2019 meet, whose handle is not recorded here at their
request. Twelve blocks of it have been read. The thirteenth has not, and the
reason is the only technically interesting thing about the object.

## The label

```
   ┌─────────────────────────────────────────────┐
   │                                             │
   │   M A R R O W                               │
   │                                             │
   │   blk 0-11  ....  index, partial            │
   │   blk 12    ....  hdr says 4096, is 512     │
   │   blk 13+   ....  ---                       │
   │                                             │
   │   DO NOT RETENSION                          │
   │   DO NOT COPY                                │
   │   ONE YEAR. HAND IT ON.                     │
   │                                             │
   │                             — hand mark —   │
   │                                  \|/        │
   │                                 --+--       │
   └─────────────────────────────────────────────┘
```

Three hands are visible in the lettering. The `DO NOT RETENSION` is in a
different pencil from the rest and was added after 2003, following the incident
at the eleventh custodian's, which is described in [[lore/sources]] and which
nobody enjoys discussing.

## The header

The volume header declares a length that the medium does not have. A DC300
cartridge of this vintage holds what it holds; the header asks for roughly four
times that. Every custodian who has run a drive at it has stopped at the same
place, in the same way: block 12 begins, reports a size of 4096 bytes, delivers
512, and the drive waits for the rest until it is told to stop waiting.

There are three readings.

1. The header was written by a program that expected a different medium and the
   tape is simply short. Dull, likely, and unpopular.
2. The tape is one volume of a set and the header describes the set. This is the
   reading that produces the interesting question, which is what happened to
   volumes two through four.
3. Block 12 is not truncated but padded, and the declared size is the padding —
   that is, the object is a claim about how much there was, made by someone who
   no longer had it.

## The contents

Blocks 0 through 11 are card records in a fixed 96-byte layout: a class, a
status, up to four citations, and one contradiction field. The layout matches
the specimen card that [[lore/sisken]] printed in 1988, field for field,
including the separate `contra` line.

This is the whole basis of the claim that the Marrow Tape is a dump of the
Sisken Index. It is a good deal weaker than it is usually stated. What has
actually been shown is that somebody who knew the card format wrote a file in
it, some time before 1995, and that the file is incomplete. Roughly 2,100 cards
are present against an Index reputed to run to four thousand.

## The custody rules

Three, and they are stated as a single sentence when the tape changes hands: one
year, hand it on, do not copy.

The first two are observed. The third is not, and this is understood by
everybody involved. At least three copies exist. Two are acknowledged. The rule
survives anyway because its function has drifted: it no longer protects the
contents, which are widely available, but it keeps the *object* in circulation,
and the object is what the meets are actually for. A cartridge that everyone has
copied and that must still be physically carried to a room once a year is a very
efficient way of ensuring the room keeps happening.

```mermaid
graph TD
  S["Sisken, 7 shoeboxes<br/>(4 unaccounted)"] -.->|claimed, unproven| W["writer of the card file<br/>pre-1995, unidentified"]
  W --> C1["1. Ossuary, 1995"]
  C1 --> C2["2. Halberd, 1996"]
  C2 --> C3["3–5, Fallow meets 97–99"]
  C3 --> C6["6. the retensioning, 2003"]
  C6 --> C7["7–13, meets 04–18"]
  C7 --> C14["14. current holder, 2019–"]
  C1 -.->|copy| K1[copy A, acknowledged]
  C6 -.->|copy| K2[copy B, acknowledged]
  C7 -.->|copy| K3[copy C, denied]
```

The first hand-off is the one that matters and is the one least explained.
Ossuary received the tape in 1995 and gave it to Halberd in 1996, in the middle
of the feud, ten months before the Armistice. Neither ever said why. See
[[lore/halberd-and-ossuary]], where it is treated as the turn.

## Related

- [[lore/index]] — the collection
- [[lore/sisken]] — the card format, and the shoeboxes
- [[lore/halberd-and-ossuary]] — the 1996 hand-off
- [[lore/armistice-banner]] — the other object that circulates by being copied
- [[lore/lineage]] — the custody chain as part of a larger graph