Hi So Apache Camel is a project that was created before github, and as such its a bit harder to count how many patches we have received from unique contributors, as back in the old days, we would write a "thanks to" in the commit message, (if we remembered).
So I wanted to see if there was some bash script that could potentially count "in a ballpark" range how many contributors there has been over the years. I am no bash ninja / perl expert (no I didn't jump on that old Camel, only our Camel). But here is a rude script that can count a bit. I opted to pick only the first name of a user, as sometimes we would just say thanks to Joe, and not Joe Washington etc. There is a bunch of commands chained together in a nice pipes and filters style. git log | grep -Eoi "Thanks to (.*)" | cut -d ' ' -f 3-3 | tr '[:upper:]' '[:lower:]' | sed 's/[^[:alnum:]]/ /g' | sort -u | cut -d ' ' -f 1 | uniq | wc -l Its the cut -d ' ' -f 3-3 that cuts the first name only. If you change it to 3-4 you get 2 words. Also this assumes that we wrote "thanks to" as the phrase. But there is likely other phrases in use. But the ones I usually did was as "thanks to" so that is a lot of them committed by me on their behalf. I am pretty sure some better bash hacker can come up with a better script. Also if we cut down to a reasonable size of log lines, then a human could run through it and filter out duplicates / invalid names etc. Also this may count people who contributed before they do PRs. But the output is 364. So I guess a conservative guess could be around 250+ people contributed before github. In github we have around 160 people. There would be some duplicates among those, so maybe we have 350 unique users, that contributed *code changes*. There is a lot of other users who contributes JIRA tickets, help on mailing lists / write blogs / give talks / and so on. -- Claus Ibsen ----------------- http://davsclaus.com @davsclaus Camel in Action 2: https://www.manning.com/ibsen2
