Brian Lamar wrote: > I love the idea of having an openstack-common project. However, the prospect > of creating such a project is daunting and quite difficult. > [...]
Thanks for bringing up the subject ! I think there are two types of benefits from this: The first is, like you said, to lower the barrier of entry for new projects. Having a set of proven ways of doing logging, configuration etc. would definitely help in spawning a new project "the OpenStack way". The second is to better establish "OpenStack" as an integrated product made of cooperating components. If people that want to deploy multiple OpenStack components end up having to learn 4 different config file formats and teach their tools to recognize 3 different log formats, it appears like the whole "OpenStack" thing is non-professional. Choosing the "best practice" way among the different projects implementations and making it the "common" way would allow to quickly reap the first type of benefits. The cost of separating it is not so large. Then we can encourage existing projects using different implementations to migrate to the "common" way, in order to reap the second type as well. There would certainly be delays and exceptions to the common rule, since this part would probably be a bit more painful: the cost of replacing a current (working) system is a lot larger. Cheers, -- Thierry Carrez (ttx) Release Manager, OpenStack _______________________________________________ Mailing list: https://launchpad.net/~openstack Post to : [email protected] Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
