Do we need to provide pagination for a directory structure?

Question

In a typical web UI, pagination is usually provided for lists. However, I would like to know whether it makes sense to provide pagination for a tree structure like a directory structure since it may contain files as well as other directories.

My dilemma is that we are combining two different entities i.e a directory and a file to show this list. Moreover a directory is equivalent to a container of files. Each time we enter a directory, the context changes and the pages displayed are relevant to the current directory. That means a user who has navigated to the inner most directory must be able to switch to any page of its parent directories and this may require multiple pagination rows to be show.

For example:

Folder A > Folder B > Folder C > Folder D

Folder D: 1 2 3 4

Folder C: 1 2 3 4 5 6 7

Folder B: 1 2 3 4

Folder A: 1 2 3 4 5

Answer 1

It doesn't seem you need a pagination control: most traversing of a tree can be done with loading more instances as the user scrolls to the bottom of the viewport.

Is there anything to be gained from the pagination, or is this because of implementation? What are users looking to do here?

An alternative: Invest in reducing the foraging costs by more robust search and filters.

You might gain more value by giving users:

• more powerful search
• filtering capabilities (file types, file owner, date ranges for creation)
• an ability to 'pin' resources
• a way to view recently accessed resources

These features save users time and cognitive costs by allowing them to access files as quickly as possible.

Breadcrumbs will be valuable for showing users where they are in the hierarchy, and are standard in applications like Box, Dropbox, and Google Drive where large amounts of files and folders mix together freely at multiple levels.

Answer 2

Sorry to be so direct: No, you should not combine hierarchy and pagination in this way. Both are difficult to use by themselves (although they are ubiqitious).

Why do we use pagination instead of scrolling: This is an implementation hack, invented for web pages, to reduce the traffic required to load the page, and thus speed up time to first interaction. Pagination does not originate in the user's mental model, it's a technical workaround for a technical problem (limited bandwidth).

Directory hierarchies are a technical construct as well. Although they are also called "folders", they are drastically different from folders, which in reality can never be recursive (folders containing folders containing folders... you get it). It is an easy way to organize the data bytes in this way, but it places the burden of organization and recall on the user.

If you want to stick to these technical workaround, use breadcrumbs to represent the directory hierarchy, and pagination to represent the list of files in the current directory, separately.

If you want to serve your users better, get rid of the directory hierarchy altogether, and use a tagging system and a powerful search (in line with Mike M's answer). Tagging a document, instead of filing the document in a specific folder, has the added advantage that several tags can be assigned to the document, while a document can only be saved in one folder. Then, make the file/folder structure invisible to the user.

You might want to read Alan Cooper's "About Face 3" for more about implementation models vs. mental models, and the file/folder structure shortcomings.