On Friday 10 November 2017 07:59:05 Michiel Beijen wrote: > I'll stick with my earlier proposal - I'll propose to go back to the > *current* latest DBD::mysql release which does not break backcompat > for our users; add the patches that we discarded when we rolled back > one by one, such as testing on many different lib/db options, memory > leaks and so on, and make a new release so we can be on the move > again.
Haven't we already done it? Haven't we already tested those Unicode changes under different libraries and MySQL/MariaDB servers, including releasing two trial versions on cpan with a long testing period without any reported problem? Answer is yes, we already did it. But people who started complaining 1) fully ignores any development and discussions; 2) fully ignores any testing release; 3) do not have own integration tests; 4) update everything after release to last version; 5) complain if something is not how they imagine, even if they misuse internals. As already stated, if you want to have those users happy, then is no other option as stop doing active development. If they still want to misuse internals, you cannot do low level changes. Or at least you would never be 100% sure that new changes does not break misusing of internals. And because for using new version of libmariadb or libmysql is needed more robust changes, it has very high rate that something which misuse internals would be broken. So conclusion is: Either new version of MariaDB/MySQL or no guarantee that misusing internals would lead to same results. And beware of fact that C code of DBD::mysql misuse libmysqlclient structures which leads to problems with compilation and also cause segfaults for libmariadb. Very similar problem like between perl application and DBD::mysql, but here it is between DBD::mysql and libmysqlclient. And if you are asking, then yes, MariaDB and MySQL decided to changes their internals and application (like DBD::mysql) which misuse it, starts crashing or not compiling. And they (MariaDB and MySQL) do that, even without fact that DBD::mysql (as application) stop working. So DBD::mysql is now in the same position as people who want to have old behavior of DBD::mysql for their applications... Should DBD::mysql maintainers now starting complaining to Oracle and MariaDB, that they must revert their changes in MySQL and MariaDB, just because DBD::mysql (as libmysqlclient application) stopped working, because is misusing internals of C structures?