Disk file capabilities

From the the ECS system's viewpoint, a capability for a disk file was a user type capability. The capability-creating-authorization for these disk file capabilities was retained by the disk/directory system and not made available to the general user. To open a disk file, a user program presented to the disk system a capability for a disk file. The disk system read the data part of the capability to obtain a disk address and unique name. Since there was no way user programs could fabricate such a capability, the disk system knew that the disk address and unique name had been put there by itself. As a final check during the open operation, the disk file header was read from the given disk address and checked for the given unique name. This was necessary because the file might have been destroyed, and that disk address used for some other purpose. These disk system unique names were distinct from the ECS system unique names. A different one was associated with each disk file. The contents of the files were dumped to tape each night and reloaded the next morning. During reloading, new disk files were created to hold the old contents, and consequently unique names were reassigned. Thus the unique name of a file differed from day to day.