Hi DBI Developers, I need a nudge to the right direction for my ideas how to get DBD::Sys more extensible. My current way is, each plugin can define tables and each table has to define it's column names and a way to retrieve the data.
This is simple, works fine as long the extensions are controller by a group of maintainers but has one (from my point of view) nasty limitation: Table PROCS - Unix implementation uses Proc::ProcessTable, Win32 implementation will use Win32::Process::Info. Of coursem they will reside in different plugins. When one fine day Proc::ProcessTable will support Win32, too - I get a conflict. So I thought the best idea would be to separate the table definitions from the data providers. And there I need a nudge how to do it the "best" way for future extensibility. Of course, other hints how to improve some things are very welcome, too :) Best, Jens