This data structure has two strong properties:

1- guaranteed uniqueness of its elements. 

2- preservation of insertion order of its elements.

The combination of these properties will allow simplification of a lot of algorithm. 

Expl:

While creating a dependency graph. You want to memoize the elements of the current branch you are traversing, and at the same time catch cyclic dependency. Very useful.

‐---------

hashed_stack | depends on

[main]               | fileB , fileC, fileD

[fileB]                | file1 , file2, file3

[file1]                | file# , file@, fileB <-- cyclic dep

Instead of maintaining two data structures, a hash and stack, you have both in the same container.

And you can adopt this implementation to other ordered containers,  like a queue:.

std::hashed_queue