I know Simon, just asking, nope an empty string is a valid string, else you say no NULL allowed for strings, might be a backend option 8)
On Mon, May 5, 2014 at 8:27 AM, Simon Slavin <slav...@bigfraud.org> wrote: > > On 5 May 2014, at 4:18pm, mm.w <0xcafef...@gmail.com> wrote: > > > "an empty string should be false" strongly disagree, a NULL string should > > be solely false, now in this case, the question is: comparisons should be > > handled as bin or by; 'literal' values? or equality/comparison must not > be > > eval'ed and strictly made on type? > > > It's a consequence of some other requirements by SQL and of weak typing. > Sorry. > > It is a requirement that an empty string must evaluate to TRUE or FALSE or > NULL. > > If you take a function that requires a number and feed it an empty string, > it understands the string as meaning zero. > > If you take a test that requires a boolean and feed it zero, it > understands that value as meaning FALSE. > > Therefore an empty string must evaluate to FALSE. > > Simon. > _______________________________________________ > sqlite-users mailing list > sqlite-users@sqlite.org > http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users > _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users