On Thu, 14 Mar 2013 22:07:39 +0100 Stefano Zacchiroli <z...@debian.org> wrote:
> For example, LWN's regular stats on contributions to the > various Linux Kernel releases are based on email domains. So it seems > unfair to compare the two worlds on the basis of which email domains are > used when contributing, given that in only one case (Linux Kernel) there > is an explicit incentive to use the company's email. Just for the record, LWN's tracking is a bit more sophisticated than that. An awful lot of kernel developers post code from something other than a corporate email address, but we (try to) get it attributed correctly anyway. Not everybody thinks that the LWN statistics are a good thing, BTW. It is easy to measure things like "changesets contributed" or "lines of code changed," It is far harder to measure "actually useful work contributed." So we publish the things we *can* measure, at the risk of slighting the developer who slaved for a week over a five-line scheduler fix and encouraging companies to spew out a lot of relatively useless code-churn patches. But it's the best we can do and is still useful, so we persevere. jon Jonathan Corbet / LWN.net / cor...@lwn.net -- To UNSUBSCRIBE, email to debian-vote-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130318080736.5696d...@lwn.net