On 24 May 2013 20:08, Benny Pedersen <m...@junc.eu> wrote: > Simon B skrev den 2013-05-24 18:32: > >> In an unscheduled maintenance window next week, I will have the >> opportunity to upgrade to 2.x should I wish to do and provided I can >> get it working on stage first. > > > +1, i would have installed 2.x if it was first time install of dovecot, i > would keep 1.x until i need a new server, since 1.x is all i need, and wiki > page for 1.x still exits so all is fine imho :=)
Thanks for the response Benny - the opportunity is that it's a new server :) >> My questions: >> >> I've seen a lot on the list about the rock-solidness of 1.2 but also >> some people saying that some versions of 2.x better than others. Is >> there a recommended version - I don't need bleeding edge, I'd prefer >> stability, or one most of you can agree on? > > > imho its not just a version change, its more then that, mailstore and > backend and out support and whole new config layout keeps me away from > migradeing it, well when i migraded from curier-imap to dovecot i have both > running the same time binded to diff localhost ips, then it was simple to > use imapsync to migrade over storages for all mailboxes, but now with > dovecot 1.x to 2.x its not that simple anymore > >> What am I missing by not upgrading? > > > if 1.x is working now, then you miss nothing, no matter that dovecot 1.x is > nearly not supported in any distros anymore, so i keep my 1.x ebuild on > gentoo, just in case i still really need to build it again Yeah, 1.2 is working and I never have to worry about it. The problem is I don't really see a feature list to give me an idea of whether the reward is worth the risk. >> A few months ago I tried to convert a Dovecot 1.2 config into 2.1 and >> wasn't very successful. Any tips on how to go about it? > > > its dangoryous to ask that here, most people would just say read the docs or > do "dovecot -n >new.conf" with the new dovecot installed, not there fault it > ends with single conf like dovecot 1.x had Yes, that's what I did - failed spectacularly. And I've just reread it again. > suggested keep 1.x for now Cheers. Unless anyone has anything to add, that is probably what I will do. Simon