I do believe that this is not a bug, but an intended "feature", of sorts.
It appears that all lists are really the same, and whitespace separation is allowed (it simply increases vertical spacing)
* this is
* a spaced out
* unordered list
this is
a spaced out
unordered list
What determines the character of the list? The first item does. This is why:
1. This is number one
15. This is number two
7. This is number three
turns into:
- This is number one
- This is number two
- This is number three
Essentially, markdown determines the character of the ordered list by the character of the first item, and basically "ignores" the list "identifiers" subsequently. If you read the official Markdown specifications, this is expected behavior. (it is even defined behavior that 2.
will start an ordered list at index 1)
Now, by the virtue that Markdown essentially ignores subsequent list identifiers after the first one, it makes sense that
1. This is the first thing
* This is the second
9. This is the third
renders to:
- This is the first thing
- This is the third
and, subsequently:
* Hello, world
1. I am unordered
turns into:
- Hello, world
- I am unordered
Again, note that whitespace is only implemented by <p>
's in the rendered list, and does not separate list items
See more of the list implementation here
This doesn't explicitly say "lists are defined by their first item*, but I believe that the behavior of the unordered list numberings being ignored after the first identifier is an obvious case to show this.