Victor Mote wrote:
Your meaning here is, at best, ambiguous. Please clarify.
If you've looked into a fair number of open source projects, and add projects from your work environment, you'll probably see certain abstractions over and over again. Counting the number of reincarnations, logging certainly comes into the top ten, I guess even at position three after configuration services and i18n. The tendency to have a project specific abstraction, however small, isn't new, check out the history part in the syslogd docs. If you are interested in a list of other recurring themes beside the three named above: - service discovery, often including loading code or data from a directory or some other repository - URL resolving - URI, URL, pathname and search path handling as Strings - command line argument parsing, maybe as part or complement of a configuration service - object pooling, in particular network connection pooling and multiplexing - XML creation - Java object persistence The list isn't complete of course. J.Pietschmann