Papercuts remixed (was Re: What's the status of papercuts?)
kent at caspia.com
Thu Jul 12 15:42:44 UTC 2012
On 7/12/2012 8:04 AM, Joshua Cranmer wrote:
> really high-vote bugs also end up being "really complicated
> things"--of the top 49 bugs by vote counts, there are maybe 5 bugs
> which wouldn't require major invasive changes somewhere.
One good example of that is bug 105169 "Filter for attachments". MIME
information (such as attachment name or even existence) is not available
at the time that filters are executed. It's a major backend change (with
possible negative performance implications) to provide MIME information
at that point.
My first impulse was to toss those out of the "papercut" list as too
complex, but as I think about it maybe what we should do is to tie these
high-vote bugs to specific backend rework projects that we have talked
about doing. (and actually "filter for attachments" is really an
enhancement and not a failure to meet design specifications ("true
bug"). I really think we need to focus on "true bugs" first).
More information about the tb-planning