Feature: Time LImits (testing the curriculum)

[SallyMcGrath]

What is the work that needs to be done?

prep

Given a prep view
And a time_limit_prep set in hugo.toml or an override time_limit_prep set in the page front matter

When the total time of blocks set on prep exceeds the time limit
Then the build should fail
And the error should explain too much work has been assigned

GIven a prep view
And no time_limit_prep assigned
Then the build should not fail

day plan

Given a day plan view
And a time_limit_day set in hugo.toml or an override time_limit_day set in the page front matter

When the total time of blocks set on the day plan exceeds the time limit
Then the build should fail
And the error should explain too much work has been scheduled and the day is too long

GIven a day plan view
And no time_limit_prep assigned
Then the build should not fail

backlog

Dependent on time being extracted from issues, which does not yet exist, so possibly this could be done separately

Given a backlog view
And a time_limit_backlog set in hugo.toml or an override time_limit_backlog set in the page front matter

When the total time of issues showing on backlog exceeds the time limit
Then the build should fail
And the error should explain too much work has been assigned

GIven a backlog view
And no time_limit_backlog assigned
Then the build should not fail

Note that some backlog items are labelled as stretch, and don't have time allocated to them. If the time allocated section contains the (case-insensitive) string "stretch", we should skip it.

Why is this work important to do?

https://common.codeyourfuture.io/common-theme/front-matter/#time

Time is very important to our learners, as many people are fitting learning around other commitments, including multiple jobs, childcare, and other responsibilities. If we are asking someone to do an activity, we must be clear about how long it will take, so they can find that time.

Additional context

#738

Who might need to know about this change?

[illicitonion]

I would have thought there would be a single backlog+prep time limit, rather than separate ones for each?

(Basically I model this as: There's an implicit backlog item every sprint which is "do the prep", but there isn't a fixed proportion of backlog vs prep, so we should treat sum(prep) as a single component of sum(backlog))

[SallyMcGrath]

I considered this, but then I thought there is this problem with overloading the prep and not leaving enough time for projects. But maybe I'm wrong and the total time not exceeding 20 hours is better.

But in this case the backlog time extraction blocks this, instead of as an improvement on this, and that's a much bigger task that might not happen. What do you think?

[illicitonion]

I considered this, but then I thought there is this problem with overloading the prep and not leaving enough time for projects. But maybe I'm wrong and the total time not exceeding 20 hours is better.

If we're saying people don't need to spend more than 20 hours a week, I think "all the work we ask them to do" should fit in that time. It feels fine to label things as optional/stretch if we don't think they're required and then not count them.

If prep + projects takes >20 hours, and we think the projects are important, it feels like the solution should be to have less prep that week, or to stretch out the work over more weeks?

But in this case the backlog time extraction blocks this, instead of as an improvement on this, and that's a much bigger task that might not happen. What do you think?

Incremental improvements are good! At a first cut, we could do something like "Assume all backlog items take less than 12 hours" which may not be true, and improve our accuracy as/when we can?