On 2015-06-11 6:43 PM, Chris Hofmann wrote: >> Furthermore, since bugs with lots of votes also have lots of CCs (see an >> earlier post of mine), if we want to just acknowledge that a bug is >> popular, we can just use CC counts above a certain threshold. >> Admittedly there's no way to search for that, but it wouldn't be hard to >> add. >> >> > yeah, it seem reasonable that there might be correlation between > votes and cc lists at some steady state of the bug. > > one interesting question is which one is the leading indicator > and which one is the lagging indicator.
Unfortunately we don't keep voting history, so that would be hard to tell. Here are a few more stats though: For the combined products Core, Firefox, and Firefox for Android, there are currently * 81 234 open bugs * 337 open bugs with greater than 25 votes. * 4 of those were opened within the last year. * of the top 10, the most recent was opened in 2005. 6 were opened sometime in the year 2000 or earlier. In the last 6 months, there were * 20 129 bugs resolved. * 11 154 of those resolved fixed. * 11 of those had 25 or more votes, although one was reopened. Of the other 10, half were fixed, two were duped, and 1 each invalid, wontfix, and incomplete. * the creation dates of the 10 resolved bugs ranges is scattered, with two opened within the last year, and they were resolved invalid and duplicate. Of the fixed bugs, they were opened in 2011, 2011, 2010, 2004, and 2000. I get the point about acknowledging votes. But what value does that really provide when (a) the number of highly voted bugs is actually rather small, (b) the number of highly voted bugs that we close in a 6-month period is miniscule, and (c) the closed bugs range from 4 to 15 years old. In that light, these acknowledgements seem disingenuous. Sure, this feature gives a little insight into a small number of issues that a small number of users care about, but how much weight should we give that compared to other forms of feedback, user research, and marketing? The presence of this feature in Bugzilla implies it is weighted much higher than it really is. I think we've outgrown its usefulness. Mark (Anecdote: bug 35168 was one of those 10 bugs resolved in the last 6 months. It was opened in 2000. It seems that what finally prompted it to be fixed was that Chrome does it properly.) _______________________________________________ dev-platform mailing list [email protected] https://lists.mozilla.org/listinfo/dev-platform
