\summary{2009}{12}{7} \agendaitem{Actions} \index{council!votes!by mail} \begin{itemize} \item Discuss voting by e-mail post-meeting in case of absence \item ulm to talk with Zac Medico about defining current portage mtime preservation behaviour for documenting in EAPI-3 \item dertobi123 to check what day of the week is best for council members nowadays and what week to hold the next meeting, possibly rescheduling the next meetings date \item lu_zero missed the meeting while having a slacker mark, figuring out what happens next (via e-mails) is necessary before the 18th December, 2009. \end{itemize} \agendaitem{EAPI-3 status} The ETA of EAPI-3, see \bug{273620}, is roughly 3 months. The currently implemented EAPI-3 items have no significant benefits to warrant a new EAPI with only the items that are already implemented. Because prefix support will be EAPI-3 (see below), the EAPI items referenced here will be referred to as EAPI-4 in the future. \agendaitem{Prefix support} \index{prefix} References: \begin{itemize} \item \agoref{gentoo-dev}{62b5df924d6e9e74c94149e7e7f17d23} \item \agoref{gentoo-dev}{2a62689c71f95e4de5699a330b8b5524} \item \url{http://sources.gentoo.org/viewcvs.py/portage/main/branches/prefix/} \end{itemize} The council accepts (4x yes, 2x abstain) the technical proposition about Prefix support made to it by now, in the form of a PMS patch, answers by the Prefix team members to the discussion thread and the portage branch implementing this. The council majority voted for a quick prefix-specific EAPI bump, from here-on known as EAPI-3. The previously intended collection of EAPI changes for a EAPI-3 will likely be referenced from now on as EAPI-4 instead. \agendaitem{mtime preservation} \index{mtime preservation}\index{EAPI!3} The council majority voted to document precisely the current behavior of portage and what can be relied upon as part of the upcoming EAPI-3 (prefix support EAPI), so that since EAPI-3 the current portage behaviour can be relied upon from all compliant package managers. The exact behaviour needs to still get documented, however.