Problems with a chunk-based file

[noeska]

Have a read on my second example:

- multiple records per chunk
A chunk can contain (for example) 600 bytes. When you write a 400 byte record in a chunck you need to specify its lengt infront of it. Now the next record can be written in the first chunk after the first 400 bytes + size (e.g. 2 bytes). and continues in subsequent chunks

| c1 | c2 |c3
| r1 | r2| r2 | r2 | fs

Also a list of records e.g. a table can be considere one filestream with
(size)(record)(size)(record)
When a record is added a new chunk is added when needed. When deleting a record an investigation should be made what chunks are involved.

This has multiple records per chunk.