Question: Referenced children tasks' accumulated attributes do not contribute to parent task

[Fitnerd]

Hi Dan!

I am not sure if this is by design, or a bug. Recently I started using referenced tasks (pasting task in some other task as a reference). I noticed that those attributes that can be accumulated (summed) and are defined (have clear values) on referenced tasks do not contribute to the values (are not added to) of the attributes in the "real" (non-referenced) parent task that contains them:

In the screenshot above, all 3 numbers are for custom attributes, and all are defined as accumulated. Yet, they do not "accumulate" in the parent "SSD" task.

[.dan.g.]

There are two main reasons:

1. Including a referenced task's time/cost/etc in more than one place would greatly inflate the actual time/cost/etc required to complete the task.

2. It is impossible to satisfactorily prevent 'circular' references eg. Task A references Task B which references Task C which references Task A

Think of references like Windows' shortcuts: They are convenient 'pointers' to files/folders that exist elsewhere but they don't contribute the target file's/folder's size.

[Fitnerd]

Thank you for your explanation.

As for point 1, I agree but it has a use case. For me, for example, I was trying to see how best to group "tasks" (in my case they were actually folder with files). Hence I wanted to define "base" tasks with all attributes set, and then have 2 or 3 "trees" with task references representing several possible hierarchies. What I was hoping for is that I would then be able to read summary (accumulated) values of the parent tasks, thus getting a quick overview of the pros and cons of each hierarchy (tree). I think it's a valuable feature in general - when you are doing planning and trying to compare 2 or more alternatives as far as resource allocation and thus would like to see both versions at the same time.

Point 2 is definitely an issue though. I suppose you could avoid that by adding entries (task ids) to hashtable why walking the reference path and stop whenever you encounter a task that's already been seen before, but that would add complexity.