Hello again, (@non-developers: this post will not interest you in the slightest. @devs: it may or may not interest you in the slightest.)
Regarding the subclassing model of sqlite4, where one subclasses by creating an over-sized struct and padding it with their fields... About a year ago i wrote a detailed article which demonstrates an alternate approach to OO in C which is, IMO, an improvement over the more conventional approach used by sqlite. The current model coincidentally runs on all modern compilers (apparently) but officially has undefined behaviour. C does not require that the pointer address will be the same for both types when a base type pointer is cast to a derived type this way (at least that's my understanding, and i apologize for not having a link handy). In the off chance that i might influence such a design decision in sqlite4, here's the article: http://wanderinghorse.net/computing/papers/DoingOOInC.pdf Happy Hacking! -- ----- stephan beal http://wanderinghorse.net/home/stephan/ http://gplus.to/sgbeal _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
