+1 for 2 interfaces Jeff
On Mon, Dec 22, 2014 at 11:36 AM, Emmanuel Lécharny <[email protected]> wrote: > Le 22/12/14 11:26, Emmanuel Lécharny a écrit : > > Le 22/12/14 11:05, Emmanuel Lécharny a écrit : > >> Le 22/12/14 10:28, Jeff MAURY a écrit : > >>> Given the fact that we introduced some kind of incompatibility (with > 2.0.9 > >>> or 2.0.10), don't you think it would be valuable to witch to 2.1.X ? > >> We probably should. My mistake would then be fixed. > >> > >> > > The real pb is that MOINA 2.0.8 broke the API, and if we switch to 2.1.0 > > - the move we should have made back then - that would mean we should > > create a new branch 2.0.10 where we remove the change I intriduced in > > 2.0.8. Here is the > > > > > > 2.0.7 -+-> 2.0.8 (whith API change, wrong) --> 2.0.9 (still wrong) --> > > 2.0.10 (dead branch) > > | \ | \ > > | \ | \ > > +-------+------------------------------------+----------------> > > 2.0.10 (with a revert on the changed API) > > | > | > > > +---------------------------------------------+ > > > > +--> 2.1.0 ( with the API change) > > > > So if we cut 2 releases : > > * 2.0.10 which contains the change done in 2.0.8 and 2.0.9 but rollback > > the changes in teh API done in 2.0.8 > > * 2.1.0 which is a branch done based on 2.0.10 but with the change done > > in the API > > > > do you think that is a viable solution ? > > > > > Another option would be to add a new Interface in 2.0.10 that those who > want to benefit from the /inputClosed() event will have to implement. We > roll back the previous interface to what it was. > / > -- Jeff MAURY "Legacy code" often differs from its suggested alternative by actually working and scaling. - Bjarne Stroustrup http://www.jeffmaury.com http://riadiscuss.jeffmaury.com http://www.twitter.com/jeffmaury
