BFG Python web framework 1.0 released(lists.repoze.org)
lists.repoze.org
BFG Python web framework 1.0 released
http://lists.repoze.org/pipermail/repoze-dev/2009-July/001936.html
1 comments
Exactly the kind of nonsense that makes me not want to do TDD. Testing = good. TDD = religion.
Ok, maybe a bit of an exaggeration. But not much.
Everything in moderation.
Ok, maybe a bit of an exaggeration. But not much.
Everything in moderation.
I'm the primary author of said nonsense. I didn't do TDD on this project, for the record.
So was my mistake in actually writing these tests or was it just advertising them? ;-)
So was my mistake in actually writing these tests or was it just advertising them? ;-)
I think it's great that your framework is 100% tested. I hold my iPhone toolkit (ZergSupport) to the same standard.
The part I wasn't happy about was that you need 8000 lines of tests for 3000 lines of code. Without having actually read your code, these numbers make me think that your framework makes testing difficult. I strive to keep my tests smaller than the code they're testing, and I'm sad when that fails.
Perhaps it's better to advertise 100% code coverage, rather than say that tests are 72% of your codebase?
Re: the previous TDD comments... TDD is supposed to make you think about tests, so when you write the main code it's more testable. TDD projects supposedly have smaller tests. Disclaimer: I haven't grasped TDD yet.
The part I wasn't happy about was that you need 8000 lines of tests for 3000 lines of code. Without having actually read your code, these numbers make me think that your framework makes testing difficult. I strive to keep my tests smaller than the code they're testing, and I'm sad when that fails.
Perhaps it's better to advertise 100% code coverage, rather than say that tests are 72% of your codebase?
Re: the previous TDD comments... TDD is supposed to make you think about tests, so when you write the main code it's more testable. TDD projects supposedly have smaller tests. Disclaimer: I haven't grasped TDD yet.
Got it. "Advertising" problem, not an actual problem then.
I would have thought that having a 100% tested framework should be a good advertising point. Where is the nonsense in that?
It's just statistics.
And this is a win?