Richard, > As the desire for dynamic pointer type strings seems to be a recurring > theme, I have added lots of extra documentation that attempts to > explain why the pointer-passing interfaces deliberately use a constant > static string for the pointer type: > > https://sqlite.org/bindptr.html#ptrtyp
As I already wrote in my response to Gwendal, I admittedly haven't read that page carefully enough. Sorry. Nevertheless, as I explained in the above mentioned post this restriction doesn't prevent a bad guy from developing a SQL UDF that can construct arbitrary pointers from SQL. That is, IMHO the restriction gives a false sense of security. Exposing pointers in SQL is potentially dangerous, although there are certainly useful applications. BTW, for my SQLite wrapper I already implemented the required housekeeping (which even avoids allocating duplicates of already registered pointer types) in less than 40 lines of code in the meantime. Regards, Ulrich _______________________________________________ sqlite-users mailing list sqlite-users@mailinglists.sqlite.org http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
