Sorry if this is a bit off topic, but I read from http://www.sqlite.org/testing.html
that your test code is 647 time bigger than the database code. I've often pointed developers of the Sage maths software to your link. I feel more attention should be paid to testing Sage, and less to adding features, but I feel it's an uphill struggle some times. But luckily the situation is improving, and now there is even commercial sponsorship for having days where people get together just to sort out bugs. Someone recently said he felt that a test:code ratio of 1:1 is about optimal, so I asked him where he got this figure from. He rather jokingly said "I only need to find one person who agrees with me" and pointed to the link http://blog.flipbit.co.uk/2009/06/what-code-coverage-percentage-should.html Of course there are practical issues with having a lot of test code - if you spend a lot of time writing that, you have less time to add functionality. I'm just interested on what your rationale was, and how the developers consider the 1:1 model? Was all your test-code written by hand, or was some of it automatically generated? I've recently wrote some code for the Sage maths project that randomly generates polynomials, integrates them, then differentials the integral. One should get back to the same as the polynomial one started with. Clearly in cases like this, I could generate millions of lines of test code very quickly, with a C program that is only a couple of hundred lines long. So I've not written a million lines, but only 200. Dave _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
