ELI5: What is a blob?

Tech terminology varies from fairly clear, like “machine learning” (letting machines learn) to the more opaque, like “intelligent process automation”. Or, it varies from completely serious, like “convolutional neural network”, to colloquial and weird, such as “dogfooding” and “rubber ducking”.

One of the latter terms is “blob”. It’s fun to say, but what is a blob, exactly?

Here’s an “explain like I‘m 5” style overview.

Blob: binary large object

A blob is a category of data. It’s a general term for the storing of a long string of information/ data as a single object.

Blobs are characterised by:

  • Large size
  • Binary data

As such, the term ‘blob’ is now understood to be short for ‘binary large object’. It’s a large cluster of binary data stored as a single object.

What is binary data?

To understand the answer to ‘what is a blob’, it’s helpful to refresh your understanding of binary data.

Binary data is a data unit with only two possible states. These states typically take the form of 0 or 1, which mirrors the binary numerical system.

Before it was known as binary data, it was called quantal data.

The history of the blob

The term ‘blob’ came before the definition ‘binary large object’. That is, large sets of binary data were called blobs before blob meant ‘binary large data’.

This is because this type of data would be referred to as blobs, and a backronym was created to suit the name by marketers. (Who believed it wasn’t professional enough to be left only as ‘blob’.)

Blobs were first invented by Jim Starkey at DEC. At first, they were nothing more than big, vague chunks of data.

Later, Terry McKiever, an Apollo marketer, decided blob needed to be an acronym for something, and invented the backronym “Basic Large Object”. Following this, IBM Informix created the alternative backronym “Binary Large Object”, which we understand as the meaning of ‘blob’ today.

What is a blob used for?

Blobs are often used to hold multimedia information. For example, images, video, and sound data. They can also be used to store programs, though this is less common.

Blobs are used like a file, but blob storage doesn’t have a user interface. So, you must create a label or ‘path’ to retrieve the data. Additionally, you need software that will interpret the binary data for you. Otherwise, you’ll only see a series of 0s and 1s.

Blob storage features in NoSQL databases. (That is, non-relational databases.) They are particularly useful in key-store-based data storage. Images and documents can be difficult to store in databases — blobs allow this to happen by storing the data in binary form, retrievable with a label/path.

Blobs are also used in some programming languages, like JavaScript. Particularly, in the creation of application programming interfaces (APIs).

What is a blob?

So, what is a blob? To summarise:

A blob refers to the storage of large data strings as a single entity.

Blobs are just one of the ways that we can store and retrieve data. It’s useful in niche places, from NoSQL key store databases to JavaScript APIs.

Yes, blobs are now considered binary large objects. But their first name was blob — and it’s a much more fun and efficient one at that.

Further reading