to add some metrics to this discussion. we have ~3,500 unique email addresses subscribed to at least one hadoop mailing list.
subscribed to: 1 list : 2431 2 lists: 784 3 lists: 238 4 lists: 118 5 lists: 56 6 lists: 66 7 lists: 102 (I didn't look at the various 'issues or commits' mailing lists). So I'm not really sure adding another list will get more focused discussions, or further split up the community. On Jan 12, 2011, at 1:08 PM, Owen O'Malley wrote: > > On Jan 12, 2011, at 9:47 AM, Nigel Daley wrote: > >> >> On Jan 11, 2011, at 8:35 PM, Nigel Daley wrote: >>> >>> On Jan 11, 2011, at 9:38 AM, Owen O'Malley wrote: >>>> We do try to move the questions off of general. It would be great if >>>> someone >>>> updated the website with the intended usage of each of the lists. >>> >>> Ok, I'll work on a page. >> >> Done. See the revamped http://hadoop.apache.org/mailing_lists.html >> Corrections welcome. You must now click on general@ to see the subscription >> instructions (it's a new page). > > I'd suggest segregating the list into 4 tables based on audience: > > user questions: > common-user > hdfs-user > mapreduce-user > security - only for notifying the project of security vulnerabilities > > project level announcements and discussions: > general > > developer questions: > common-dev > hdfs-dev > mapreduce-dev > > jira and subversion tracking for developers: > common-issues > common-commits > hdfs-issues > hdfs-commits > > mapreduce-issues > mapreduce-commits > > -- Owen
