Two types of files, BINARY files which could be either music, graphics or sound etc... or just a TEXT file.
A BINARY file contains; records; fields and values.
Data is not stored in a particular order in a serial file. Bank transactions is a good example, though time would be the ORDERING field. Theoretically the data in the files is not ordered by data (apart from time).
Similar to a database in that files are structure by a unique primary key field (very similar to a database). The difference is with a database the key can be accessed directly where as with sequential files the key field is read sequentially until the required value is found. A subtle difference! The sequential files could be customer records(profile) for a bank, transaction details are kept as serial files which are read periodically and then the relevant customer details, sequential files, updated in a batch. The KEY FIELD is the one which is used to find records. Sequential files are found, using a LINEAR SEARCH process for the key-field, which can be time consuming.
These types of files are the most time efficient to access, sometimes referred to as 'random access files'. Randomness comes from the lack of sequential file access. Direct access to its unique record isnt actually possible, due to limits of the data storage references. A hashing algorithm is often used to calculate the position of the start of the file, example usually a nearby file is located at the beginning or end of the current block and then serially incremented up or down to reach the required record. A NEW version of the file being edited is created, records are processed until the one being edited or deleted is found in the copy process. This record is then edited or deleted as required and then rest is written to the new copy,
Day-to-day file access normally uses DIRECT ACCESS of files (hard disk), SERIAL files would be most suited to a backup medium where data is stored in no particular order, batch processing.