On 12 April 2012 22:24, Anthony Liguori <anth...@codemonkey.ws> wrote: > The end goal shouldn't be s/qdev/object/g. That doesn't make things better > by just using a new shiny infrastructure. Splitting device initialization > into two stages is the fundamental reason for introducing QOM in the first > place.
It might be yours, doesn't mean it's everybody else's :-) My main desire from shifting to QOM is named connections between devices and killing off the strict hierarchy of qbus buses; I haven't been particularly bitten by device init issues. -- PMM
