Improving Bugzilla: People, Bugs, Search, and Planning

Advertisement:

POWERED by FUSION

The Firefox UI team recently got together with the Bugzilla UI team to brainstorm future features and user experiences for Bugzilla.

We scoped our brainstorming session to be between trivial fixes and doable-in-a-couple-months features. That is, pragmatic, nearish-term sorta-big thinking. Our brainstorm revolved around four major themes: People, Bugs, Search, and Planning. Of course, during the session most of the themes blended together — real-life is never as orderly as our a priori taxonomies would have them.

Before we delve into our thoughts, what would make Bugzilla better for you? For a beginning contributor? For project management types? For a one-time bug submitter? Most of the Firefox contributors spend most of their day in Bugzilla, so any improvements made are dramatically multiplied.

Many thanks to the folks on both teams who took the time to sit down for the better part of an afternoon. In particular thanks go to Max Kanat-Alexander, Alex Faaborg, Jennifer Boriss, Madhava Enros, Dave Miller, and an especially large thanks for Guy Pyrzak (who took notes and guided the session).

Thoughts on Bugzilla

People

* Bugzilla is fundamentally about issue tracking. But Bugzilla is also scaffolding and structure for the people who report and fix those bugs. Make people first-class citizens in Bugzilla (yes—a social platform…)

* Help users determine “who is this guy”, when looking at bug reports and commenters. Dashboard for people: what bugs they’ve filed, what patches they’ve submitted, what groups they are in, how active they are, etc. See   Github for an example.

* Social rankings by participation (bug submitting, duplicate marking, patch submitting, patch reviewal, high-valued comments, etc.). Social incentives and social standing.

* Help create a “team” feeling. Should be able to become a part of a virtual working group (I’m interesting in UI, Graphics, Networking, etc.). There are some open questions here: Are built in user groups good enough? Is fluid definitions of groups better than restrictive (probably, yes).

* Autocomplete for user names, ordered by social ranking.

* Extensible, RESTful APIs for accessing the social graph and other social features. Support Open Social API.

Bugs

* Allow tagging comments with meta data (useful/troll/repeat/etc.) to make it easier to read, parse, and participate.

* Display users and groups in comments (in the classes/style area). This way ambient information about which comments are important can be visual indicated.

* Understanding how “Hot” a bug is. How many times has it been viewed? How many times has it been commented on? How many times has it been marked as a duplicate of other bugs? How often people start to create a bug that they find out is a duplicate?

*A five-minute undo/edit ability for bug comments. So that when you’ve got that bother-I-forgot-to-spellcheck feeling just after hitting submit, you can do something about it.

* Google Mail style shortcuts to help with efficiency handling bugs (would require some more info on most used actions)

* Improve Bugmail work-flow with a more human readable format. Better fonts, better diffs, built-in threading.

Search

* Make search not horrible! It should be as painless and as easy to find the bugs you need to find (or the ones you didn’t know you need to find) as a web search.

* Make automatic finding of duplicate bugs a zero-cost action for users.

Planning

* Create a dashboard overview of projects, groups, and sets of bugs. See Trac. Questions the dashboard should answer: When are we going to release? What needs to be done for the next milestone? How can I get involved, and at what level? What’s happened since the last time I checked? Which things are at risk? What’s the big picture? What’s the long term trending?