On 3/23/2011 8:25 PM, Nick Kew wrote: > > On 24 Mar 2011, at 00:10, William A. Rowe Jr. wrote: > >> So if this is not in the .h's (and without namespace protection it has >> no place in our .h's), can we just 'not do that' and fully declare all >> our 'enum XML_Error' results, please? I think that's the better patch. > > No. "enum XML_Error" is an expat-ism and breaks compile with libxml2. > Hence typedefing it to int in apr_xml.c (the common code) and to two > different enums in the lib-specific stuff. > > We can clean it up (and switch to a neutral name) or drop it, but I didn't > see that as a priority. But if builds are breaking then I guess it moves up > the priority list.
Thanks for clarifying!
