2019-2020 Retrospective

In the past year, Dev. Pri. accomplished the following:

1. We created and administered a prioritization survey for the ArchivesSpace user community.

2. We established a Ready for Community Developer queue to support engagement and code contributions by Community Developers. We publicized it with a presentation at the 2019 Member Forum in Austin, TX, and in a blog post in March 2020.

3. We continued to review and prioritize bug reports and feature requests.

Things to consider for the future:

1. While we also experimented with creating “swimlanes” for shuttling tickets to other subteams such as Metadata Standards and Usability, it’s not the easiest to figure out a workflow for getting them back into the Development prioritization queue. For example, if someone applies the “usability” tag, it goes into the Usability subteam queue, which may increase the response time for actually fielding it, especially if the subteam is unable to field it earlier. Additionally, once the subteam has looked at it, it’s not obvious how to route it back to Dev. Pri. Perhaps we should explore using an additional “subteam” tag to further distinguish it from tickets that don’t require the additional subteam review.

2. We began experimenting with having Dev. Pri. members volunteer to assign tickets each month. This could potentially be explored further. Benefits are that it will reduce the time commitment for the subteam leader(s) and that other Dev. Pri. members will have greater holistic insight to issues in the Jira system, potential challenges might be that it can be a lot of work and a learning curve for whomever is assigning tickets in that month.

3. Dev. Pri. might want to consider budgeting time or having a “communications point-person” to compile a list after each meeting to share on the member list of tickets that need community input, to prompt greater participation and feedback in Jira.

4. Dev. Pri. might want to consider additional approaches to close outdated, irrelevant tickets. Graphs of opened versus resolved tickets show that there’s a large and increasing backlog of tickets. While assigning three tickets per person per monthly meeting ensures a large chunk of tickets are being fielded, there hasn’t been an obvious path forward to systematically and aggressively reduce the backlog.

   1. An approach we piloted this year was an “old ticket blitz” using an older ticket view, assigning 6-10 tickets each for rapid-fire assessment of whether tickets are relevant and should remain open or if they can be closed. Perhaps this approach can be employed 2-4 times per year.