Yeah... I could have *swore* that we allowed this before with the scoreboard, but can't for the life of me find it... Must be dusty neurons.
Agreed that it breaks ABI to change the struct. :/ On Sep 4, 2014, at 3:44 PM, William A. Rowe Jr. <[email protected]> wrote: > No... only if the patch is restructured to preserve all existing structure > members at their current offsets. New struct members at the end of an > existing structure is the definition of a minor mmn bump. If third party > module authors allocate ap structs, it is their job to track against mmn > minor revs. > > > Jim Jagielski <[email protected]> wrote: > > I think, in this case, a minor could be justified. > > On Sep 4, 2014, at 1:57 PM, Plüm, Rüdiger, Vodafone Group > <[email protected]> wrote: > > > But IMHO that would be a major bump and not a minor one. And we cannot do > > major ones in stable branches. > > > > Regards > > > > Rüdiger > > > >> -----Ursprüngliche Nachricht----- > >> Von: Jim Jagielski [mailto:[email protected]] > >> Gesendet: Donnerstag, 4. September 2014 19:55 > >> An: [email protected] > >> Betreff: Re: svn commit: r1622429 - /httpd/httpd/branches/2.4.x/STATUS > >> > >> I think we can, as long as we bump the MMN... > >> > >> On Sep 4, 2014, at 7:22 AM, Rainer Jung <[email protected]> wrote: > >> > >>> Am 04.09.2014 um 12:13 schrieb Ruediger Pluem: > >>>> Can we really backport this? > >>>> > >>>> We are increasing the size of proxy_worker_shared and changing > >> offsets inside the struct. > >>> > >>> Bummer, I guess you are right. mod_proxy.h seems to be part of the > >> public API so we can't backport like this. Will revoke the proposal. > >>> > >>> We could think about adding new larger name fields to the end of the > >> struct and keep a truncated copy in the original struct mebers. But that > >> means 3rd-party modules using the old original fields would only see > >> part of the names. > >>> > >>> Rainer > >>> > > >
