Hi Sergio, This sounds like a bug in the code generator. The intent is that supplemental interfaces can be disabled using the Conditional attribute, just like other interfaces. Would you be willing to file a bug at bugs.webkit.org and CC Kentaro and myself?
Thanks! Adam On Thu, Feb 23, 2012 at 7:44 AM, Sergio Villar Senin <[email protected]> wrote: > Hi, > > I've been trying to build the GTK port without geolocation support and > found that is no longer possible (after r108610). The reason is because > that commit removes an #ifdef guard that was hiding the actual issue IMO. > > Thing is that NavigatorGeolocation.idl specifies a conditional interface > to be added to the Navigator interface. It seems fine as the conditional > clearly specifies that it needs the GEOLOCATION feature. The generated > code properly has the #ifdef guards to ensure that the generated code is > only built if geolocation is enabled. The problem is that it generates a > function whose signature is something like > > WebKitDOMGeolocation* > webkit_dom_navigator_get_geolocation(WebKitDOMNavigator* self); > > It does not matter that the implementation is correctly surrounded by > the proper #ifdefs as the doc says[1] because the compiler will always > complain due to the lack of WebKitDOMGeolocation object. > > So my question is, shouldn't the platform specific bindings generators > (resolve-supplemental.pl maybe?) skip the addition of supplemental > interfaces to their parents if the specified conditions are not met? > > BR > > [1]: http://trac.webkit.org/wiki/IdlAttributes > _______________________________________________ > webkit-dev mailing list > [email protected] > http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev _______________________________________________ webkit-dev mailing list [email protected] http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev
