I think we approach these with a disable, then delete approach. Disable for one or two releases and if no one steps in to revive it in that period, we delete it.
On Mar 14, 2017 9:10 PM, "Rafael Weingärtner" <rafaelweingart...@gmail.com> wrote: Dear ACS fellows, Recently there have been two threads asking and discussing the “midonet” integration with Apache CloudStack (ACS) [1-2]. After quite some discussions, we noticed that despite having some people willing to use it, the plugin has never been fully developed by its vendor (Midokura). Further, nobody else has put the effort on fully testing and finishing its implementation. It seems that the plugin was incorporated into our code base without being fully finished. Moreover, I have asked around at the Midonet community, and the java client they use has changed quite a bit from the one we use. It begs the question, if it does not work, why do we advertise such integration? [3]. In my opinion, it would be great if we had such integration; however, we as a community of individuals cannot bear the burden with the cost of such task by ourselves. It seems we have three options; (i) disable the build for the plugin and let the code create its own mystic throughout time in ACS code base; (ii) remove everything; or (iii) someone that may benefit from this plugin jumps in and concludes the integration with Midonet using their new client. There maybe other solutions that I am not seeing. So, @Devs yours thoughts and comments are welcome ;) [1] http://cloudstack.markmail.org/thread/qyedle5jb2c34gsc#query:+page:1+mid: xn2zq2v3eim5vl2q+state:results [2] http://cloudstack.markmail.org/message/rewzk4v7dgzpsxkm? q=midonet+order:date-backward&page=1#query:midonet%20order% 3Adate-backward+page:1+mid:i563khxlginf6smg+state:results [3] http://docs.cloudstack.apache.org/en/latest/networking/midonet.html -- Rafael Weingärtner
