Re: BC break in 5.4.29 and 5.5.13

On Tue, Jun 17, 2014 at 10:37 PM, Stas Malyshev <smalyshev@sugarcrm.com> wrote: > Hi! > >>> It is definitively not OK. I cannot imagine any feature that cannot >>> x.y+1, which happens less than a year later. > > I think we've been through this a number of times and most of the people > agree that waiting for the next release cycle (at least a year) for a > self-contained minor function or feature addition makes little sense. > People add functions because they have need for them now, not years > later when they're ready to make a major version upgrade. We're not > talking something huge like new syntax or language feature, we're > talking about this: > >> OPCache: Added function opcache_is_script_cached(). > > Why one should wait for a year or more to be able to check if the script > is cached? It makes little sense to me. Adding it to active branch does > not hurt anything, so I see no problem with it. And where is the limit? Add a full extension is fine too? No, it is not. It reintroduces the pain to add dozen minors version checks for php version dependencies. -- Pierre @pierrejoye | http://www.libgd.org