Re: Credits for tests

Hello, I don't think you have to lecture me about testing. And one thing is probably important here. 100% test coverage is better than 50%. Because 50% test coverage means that for half of our code we do not have the slightest clue what it is doing or whether it still does the same when we change something. And btw, I never said we stop when we gained over 90%. And I never said run-tests is all we should do. It is just one effort that can help a lot. marcus Monday, May 26, 2008, 9:12:17 AM, you wrote: > On May 24, 2008, at 3:23 PM, Marcus Boerger wrote: >> Hello all, >> >> I just had an idea for the new credits section in the tests. Since >> only >> the new once have them, we could add them for all tests by adding >> them from >> CVS, by submitter. Anyone interested in taking this on? >> >> The other idea was about using these tags more. Since each tag >> consists of >> easy parseable lines, we could generate bar charts. Or we could >> allow to >> search for them on gcov.php.net. Heck we could even allow to >> 'sponsored' >> links in the section. So company XYZ could get a link >> http://xyz.com >> if we >> submit their tests. That would look like: >> >> --CREDITS-- >> Foo Bar <foo.bar@xyz.com> >> XYZ (http://xyz.com) >> --... >> >> To distinguish between them I chose () for links as opposed to <> for >> mails. However that is not really necessary as we can detect >> whether it >> is a link or a mail address easily. >> >> Any thoughts on this? >> >> In case this is not clear, I want to make writing tests more >> attractive to >> get up in test coverage. Up as much as possible. In fact I want to >> raise >> the coverage from 0%-50%,50%-75%,75%-100% to 0%-75%,75%-85%, >> 85%-100%. That >> is I especially want to show green only for 85%+ instead of 75%+ right >> now. And imo anyone thinking 50% test coverage is any usefull is damn >> wrong, so I changed from 50%+ to 75%+ as well. >> > I can understand the desire to upgrade the level of what might be > considered acceptable > test coverage by increasing the requirements for a green status. So I > do not argue the > merits of this change. But I also agree 100% with Zoe. Code coverage > is a superficial > measurement of how well tested some code is. A piece of code with > tests designed to > specifically touch all the code is an artificial security blanket. It > certainly does not mean > that the code is future-proofed from bug introduction via testing. > Of course it's impossible to determine proper test coverage in an > automated fashion, so > code coverage at least gives us an indicator to work from. I'm just > hoping that all the > focus on achieving high code coverage doesn't cause testers to write > tests for the sake > of coverage instead of analyzing the domain and testing based on that > analysis. > As for what might attract more testers. I can only tell you why I'm > doing it. I've been > wanting to contribute to Open Source for some time. But I've shied > away till now for > probably the same reason a lot do in the beginning. Didn't think my > skills were good > enough and had a hard time finding something I could do with a low > barrier to entry. > I was introduced to the idea of TestFest at a local user group and the > presentation included > enough detail that I could get started doing something without > spending weeks on end > just researching and putting the tools together to do the job. > The additional side benefit of being able to say that I do contribute > to Open Source on my > resume is certainly an additional motivator. And lastly, even though I > already have more > email addresses than I need, I would love to have one ending in > php.net. :) >> Best regards, >> Marcus >> >> >> -- >> PHP Quality Assurance Mailing List >> <http://www.php.net/> >> To unsubscribe, visit: http://www.php.net/unsub.php >> > Eric Stewart > e.stewart@mac.com Best regards, Marcus