On Mon, Oct 1, 2012 at 5:28 PM, Federico Schwindt <[email protected]> wrote: > On Mon, Oct 1, 2012 at 4:02 PM, David Coppa <[email protected]> wrote: >> On Mon, Oct 1, 2012 at 4:59 PM, Stuart Henderson <[email protected]> >> wrote: >>> On 2012/10/01 15:08, Federico Schwindt wrote: >>>> > The diff below should do the trick (not tested, qt4 is such a beast to >>>> > build)... >>>> > >>>> > Grepping for "QLibrary ", finds other occurrences that maybe we should >>>> > also fix. >>>> >>>> I'm not sure about this. It seems to me that this: >>>> >>>> >> dlopen: loading: libicui18n >>>> >>>> Should really be libicui18n.so. I think this is better but untested >>>> and straight from the source: >>> >>>> --- qlibrary_unix.cpp.orig Mon Oct 1 14:59:57 2012 >>>> +++ qlibrary_unix.cpp Mon Oct 1 15:00:28 2012 >>>> @@ -143,9 +143,8 @@ >>>> #endif // Q_OS_AIX >>>> if (!fullVersion.isEmpty()) { >>>> suffixes << QString::fromLatin1(".so.%1").arg(fullVersion); >>>> - } else { >>>> - suffixes << QLatin1String(".so"); >>>> } >>>> + suffixes << QLatin1String(".so"); >>>> #endif >>>> # ifdef Q_OS_MAC >>>> if (!fullVersion.isEmpty()) { >>>> >>>> f.- >>> >>> I haven't tested yet, but this seems a much better approach than >>> patching the callers. >> >> Sure, I also think Federico got it right. > > Ok, attached. Ttested with keepassx and a demo for the cups patch removal. > The gtk patch removal still needs testing. Anyone? > > f.-
i just applied your patch and lawrence's original kdiff3 submission. kdiff3 works fine.
