Project

General

Profile

Forums » Open discussion »

What is a canonical use case for subtask?

Added by Max Moroz over 12 years ago

Note: I posted a similar question at SO.

To clarify my question:

I think of subtask relationship as a type of issue relationship, along with related-to, blocks, precedes, etc. The related-to relationship is the most basic (with no semantics at all). The subtask seems to be equivalent to related-to plus:

- tree-like representation in the UI
- certain parent's attributes are auto-calculated from from children attributes

At least this is what this definition of subtask implies (I believe it hasn't yet migrated from the issue tracker to the wiki).

Now, this issue raised several very serious concerns with the auto-calculation rules. (Until I saw that issue, I thought I'm crazy and just don't get something very obvious.) In particular, the very fact that parent task attributes are not free to change after a subtask is added, is shown to be somewhat questionable.

If, for the sake of the argument, a subtask is redefined to remove the auto-calculation of parent attributes, would it become simply a related-to relationship? (The extra subtask visualization should arguably be available for other relationships anyway). Then why is subtask even there?

If the current auto-calculation is so useful as to make subtask worth adding to Redmine, then there must be a really powerful canonical use case where this precise (and somewhat arguable) semantics is a great fit. What is it?


Replies (4)

RE: What is a canonical use case for subtask? - Added by Hung Nguyen Vu over 12 years ago

If you really mean canonical use, that usage of subtasking should be divide-and-conqueror where we devide a big task into smaller tasks to make management easier.

RE: What is a canonical use case for subtask? - Added by Max Moroz over 12 years ago

But isn't this better achieved by blocking tasks? (By default, subtasks are not blocking the parent task.)

And also, why is the priority of small tasks determine the priority of the big task? Shouldn't it be the other way round?

RE: What is a canonical use case for subtask? - Added by Hung Nguyen Vu over 12 years ago

For me, I don't care about the relationship between the parent task and its children as well as their priorities.

Redmine meant to be simple and I follow that.

May someone will show you and me some a smarter way on how to use subtask and grouped tasks.

RE: What is a canonical use case for subtask? - Added by @ go2null over 12 years ago

Main benefit of Subtasks that I see is that they are presented as a tree c
in the UI.

    (1-4/4)