Blocks, Attributes, and External References
Blocks
A block definition lives inside the drawing and is placed by INSERT with an insertion point, scale and rotation. Redefining a block with BEDIT updates every insert, which is the main reason to use blocks for repeated content. Dynamic blocks add parameters and actions (stretch, visibility, lookup) so one definition covers many configurations. Exploding blocks is a last resort: it breaks the link to future redefinitions and converts attributed inserts into loose text.
Attributes
ATTDEF creates a tag (the data name), a prompt and a default value; an insert becomes an attribute reference. ATTDIA controls whether the insert dialog appears and ATTREQ whether prompts run. After redefining a block that contains attributes, existing inserts keep their old attribute set until ATTSYNC is run. BATTMAN reorders and edits attribute definitions, and DATAEXTRACTION writes attribute data to a table or CSV for schedules.
External references
Attach and Overlay differ in nesting: an attached xref travels with the host when the host is itself referenced, an overlay does not. That makes overlays the right choice for background files shared across a team. Path type (relative, absolute, no path) determines how the reference resolves after a folder move; relative is the safest inside a project folder. UNLOAD keeps the definition but drops the display, DETACH removes it entirely, and BIND converts the xref into a local block while INSERT binds without the $0$ name prefixes.
Worked example
A structural grid xref is needed in your plan, but the architect who references your plan must not inherit the grid. Attach the grid as an Overlay; when your drawing is attached to the architect's file, the overlay is ignored.
Traps
Expecting a redefined block to refresh existing attributes without ATTSYNC; binding xrefs and then hitting duplicated layer names; absolute paths breaking after a server migration.