As we developers work on our software, we get ambitious. We want to add features that are cool that will make life easier for our users. This may seem like a good thing at first, but, in the words of Admiral Ackbar, it’s a trap.
I have come to such point working on Outside Tasks. Having added dependent tasks, I had two tasks that depend on each other. The first task was to ask someone for some information, and the second task was to use the information. The problem is that, in keeping with my original vision for Outside Tasks, the second task shouldn’t be available until it is relevant. However, the second task is not relevant until you receive the information. The challenge is knowing when this task is blocked blocked and when it is no longer blocked.
Currently, Outside Tasks has two ways to defer a task:
- Defer by date. An example of this is my recent interaction with customer service where they said they only do certain things on Tuesdays and Thursdays which meant that my task was deferred until Tuesday.
- Defer by task. This is how dependent tasks are created. If task A is blocked by task B then task A won’t be available until task B is done.
The problem is that neither one of these methods of deferring a task is sufficient.
There are a few solutions to deferring the task of acting on the information:
- Add a manual task is deferred flag. This allows users to mark the task as deferred and then unmark it when they receive the information. However, this is problematic because the task is no longer in their relevant tasks so they might completely forget about it.
- User defer by date. Deferring the task by a few days will do the trick but it is far from perfect. You may still not have the information by the chosen date in which case you have to defer it again. You may get the information before the chosen date in which case we are back to the problem of the first solution.
- Defer by person. This solution involves determining if the person has been communicated with. This provides the most accurate results but the most amount of work. Imagine having to interact with APIs for all popular methods of communication and supporting new ones. We’re talking email, Jabber, Discord, WhatsApp, Viber, SMS if an API is available, Facebook Messenger, Snapchat, Instagram, etc… But wait, there’s more! You don’t want the task to be unblocked by the first message that the user receives from that person; they could have a conversation about something entirely different, or the person you are waiting for may ask for clarification. This means that we need some kind of trigger word that marks the task as unblocked. What if the person doesn’t use that word?
Unfortunately, I haven’t figured out how to best solve this issue. My best inclination is to use a combination of defer by date and some kind of view that allows users to quickly see which tasks are blocked and unblock those when appropriate. Additionally, GTD involves performing weekly reviews. My initial goal was to reduce the need for weekly reviews. That is why I created dependent tasks since they allow for planning tasks ahead of time. I believe that, as a compromise, that solution will have to do for now.
Decisions like this slow down development, however, they do not slow it as much as implementing a cool feature just because it is the best one that provides the most convenience to the user.
Since the last update I have added dependent and sub tasks which are displayed in a tab that only shows the tasks that are relevant right now. I have also added projects to the app to help organize tasks a bit more. It is still a long way to go until I have something I am comfortable releasing but I should be able to have a stable alpha version in the next month or so.