Number of bugs as a function of River position

CPAN Riverbugs Thu 31 December 2015

How would you expect the number of open bugs / issues to vary as you look at distributions at different positions on the CPAN River? I expected the count to increase as you go up river, and it does. There, you don't need to read the rest of this post now :-)

For each CPAN distribution I just added the number of open RT tickets to the number of open github issues. I know there are various flaws with that, but it's good enough for my purposes. Here's the summary by stages of the CPAN River:

Number of downstream dependents
10k+ 1k - 9999 100 - 999 10 - 99 1 - 9 0
# dists 45 195 570 1589 8210 21250
# issues 613 1869 3427 5103 9259 10395
issues per repo 13.6 9.6 6.0 3.2 1.1 0.5

A distribution moving up river means that it is being used by more and more code, and thus by more people. Which is likely to result in more issues being found, more wishlist items, etc.

Those distributions that aren't used by any other CPAN distributions are least likely to have open issues, which kind of makes sense.

A better metric would be rate of issues being raised, but that would be a lot more work to come up with. If someone fancies coming up with that ...

And really here I'm interested in whether issues are getting resolved, either by fixing, or closing them with "sorry, not going to add/change that", etc.

There are only 45 distributions at the head of the river, which makes it feasible to look at the outstanding issues, see which ones have in fact been closed, and if there are any which are never going to be addressed (and could be closed).

comments powered by Disqus