I can describe the concrete problem I will have with LOGCXX-319 given that all the code for which I am responsible has been written using log4cxx 0.10
In 0.10, you have to know the macros are "blocks" not "statements" as any conditional logging macro has to be written without a trailing semicolon. For example: if (someCondition) LOG4CXX_INFO(m_log, "Condition 1 message") // Note a semicolon here will cause a compile error else LOG4CXX_INFO(m_log, "Alternate message) As a result of knowing the macros are "blocks", most LOG4CXX_ XXX() code does not have a trailing semicolon. I would like to be able to compile my systems with both 0.10 and 0.11 for a transitional period (i.e. to avoid a 'big bang' switch over). If I was forced to use a 'big bang' change, I would probably just change to log4cpp which is already supported by package managers (conan.io and vcpkg). I would be surprised if any unix distribution would change to 0.11 log4cxx if its API is incompatible with 0.10. Regards Stephen Webb On Tue, Mar 3, 2020 at 4:28 AM Thorsten Schöning <tschoen...@am-soft.de> wrote: > Guten Tag Ralph Goers, > am Montag, 2. März 2020 um 16:34 schrieben Sie: > > > There is a difference between a user’s compile failing vs the build > > having changed. > > And which? Things don't work in the worst case either way and need to > be adopted. Why exactly is getting rid of build support by ANT > acceptable for users relying on that, but applying LOGCXX-319 might(!) > not be? > > If I remember correctly, the concrete changes could even be adopted > using automatic search&replace. > > > Given how old log4cxx is I would expect it to be > > used in a fair number of places despite its version number. > > And a fair number of users applied either the available patches > already since the last release or simply work with master already > anyway. I can't remember anyone complaining about the changes > introcuded by that concrete issue in the last years as well. > > > I > > haven’t looked at the code myself but is there no way to keep it > > backward compatible while also keeping the new changes? > > In my opinion this is an unnecessary meta-discussion until a concrete > problem has been described introduced by LOGCXX-319 or other changes. > So at least I won't reconsider each and every change since the last > release. > > Mit freundlichen Grüßen, > > Thorsten Schöning > > -- > Thorsten Schöning E-Mail: thorsten.schoen...@am-soft.de > AM-SoFT IT-Systeme http://www.AM-SoFT.de/ > > Telefon...........05151- 9468- 55 > Fax...............05151- 9468- 88 > Mobil..............0178-8 9468- 04 > > AM-SoFT GmbH IT-Systeme, Brandenburger Str. 7c, 31789 Hameln > AG Hannover HRB 207 694 - Geschäftsführer: Andreas Muchow > >
