BSDCan2008 - Final - we hope
BSDCan 2008
The Technical BSD Conference
| Speakers | |
|---|---|
|
Mark Linimon |
| Schedule | |
|---|---|
| Day | Talks - 1 (2008-05-16) |
| Room | B |
| Start time | 15:00 |
| Duration | 01:00 |
| Info | |
| ID | 77 |
| Event type | lecture |
| Track | Hacking |
| Language | en |
| Feedback | |
|---|---|
|
Did you attend this event? Give Feedback |
Busting Bugs on FreeBSD
Where We Are, Where We Ought To Be
FreeBSD has an active user community that submits hundreds of Problem Reports (PRs) every week. This paper discusses the problems faced in trying to handle this large number of submissions.
Traditionally, FreeBSD has used the GNATS problem tracking system. This is a flat-file "database" with limited search and browsing capability. Its main strength is its simplicity and ability to accept submisssion via email, which is why it still remains in use. However, it has become inadequate for FreeBSD's needs, as we fall further and further behind handling new submissions.
Rather than discuss available problem tracking systems, this paper attempts to define what a modern workflow model for PR handling should look like. Topics include:
what kind of PRs are there, and which do we think we are able to handle?
can we create a culture where non-committers can help with certain types of PRs? If so, what method(s) do we use for that?
what can we do to make it more attractive for developers to work on PRs?
what kind of search and browsing capabilities do we need?
what kind of notification of incoming PRs, or PR state changes, is appropriate?
Once we understand the answers to these questions, then we can finally start looking at what software can be made to be more suitable for our needs.