git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: Add a bugzilla website

From
Andrew Ardill <andrew.ardill@gmail.com>
Date
Nov 15, 2013, 17:27 UTC
Message-ID
<CAH5451=+N1vYoNeweQXe9vavjmRvkEzP=7U+mDf-0zH9OSpZPA@mail.gmail.com>
In-Reply-To
<20131115135132.431d3e344dadee64e2be5127@domain007.com>

On 15 November 2013 01:51, Konstantin Khomoutov <flatworm@users.sourceforge.net> wrote:

> But there was an announcement that an experimental JIRA instance has
> been set up for Git [1].  I'm not sure what its current status is, but
> you could look at it.
So!

The biggest concern has always been that any bug tracking system needs to complement the existing workflow of many developers. For bugs and feature requests, they are raised, discussed, and fixed on the list. Replacing this process is not in scope for a bug tracker.

In that framework the main value a bug tracker has is keeping track of what bugs exist, what versions they affect, and when they are fixed. Unfortunately, at the moment, collecting and curating this information is entirely manual.

The JIRA attempt [1] looks to pull in every conversation and thread it for us (adding replies as comments to existing tickets), but hasn't tried to anything beyond that. Automation may be possible, to do things like parse What's Cooking and the release notes, but that is a harder problem :)

The tools are there, and you should be able to log in and add/update issues. Feel free! (Definite value would be derived from a dedicated curator who updated the bug tracker manually)

I had a look over the set up (I hadn't in a while) and realised we were dropping some emails, so I'll try and fix that, but the bigger problem is that simply creating tickets to track conversations is not enough.

We need to then identify those conversations that we care about and capture some metadata about them - did they resolve the reported issue, and when did that happen? Junio is the source of truth for this, and so ideally we would use his communications to understand it, but that just shifts the problem to linking the issues he writes about to the conversations that started them.

In any case, adding value to the existing process is hard (because it works quite well!) and probably requires significantly more work to even understand what that value might look like. This, I think, is the key reason it is hard to truly get started with any bug tracking solution; the solution is not obvious, and the current (very customised) workflow is not supported directly by any tool.

Regards,
Andrew Ardill
[1] https://git-scm.atlassian.net
Previous: David AguilarNext: Jonathan Nieder
Message 8 of 13 in “Add a bugzilla website”
  1. ycollette.nospam@free.frNov 15, 2013
  2. Damien WyartNov 15, 2013
  3. ycollette.nospam@free.frNov 15, 2013
  4. Konstantin KhomoutovNov 15, 2013
  5. ycollette.nospam@free.frNov 15, 2013
  6. Matthieu MoyNov 15, 2013
  7. David AguilarJan 8, 2014
  8. Andrew ArdillNov 15, 2013
  9. Jonathan NiederNov 15, 2013
  10. brian m. carlsonNov 15, 2013
  11. ycollette.nospam@free.frNov 18, 2013
  12. Matthieu MoyNov 18, 2013
  13. Fwd: Add a bugzilla websiteStefan Saasen, Jan 13, 2014

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.