Hi Thorsten, On Wednesday 19 May 2010, Thorsten Ziehm wrote:
> > The fact that we differ in the development style (everyone commits the > > patches directly to the main repository, no CWSes, and the possibility > > for bugs coming from integration/merges of various CWSes), does not mean > > that we do not have a QA process. > > > > We have a tinderbox that notifies committers of wrong commits, we release > > betas in timely manner, have quite an extensive test plan (in our > > Testopia), run qatesttool during the beta phase & analyze the results, > > have a large set of documents for regression testing, etc. Oh yes, I > > must not forget the bug triage of the new bugs which guarantees the > > initial treatment of the new bugs in up to a (working) day or two. > > Mechtilde and you are talking about different QA processes. What you > are writing is developer orientated testing. This is done for OOo often > before a CWS goes to a QA community member. So the manual testing and > the regression testing on UI level is missing in your processes. Of course the manual testing is not missing :-) - the test plan (Testopia) I wrote about are all just manual tests, done internally by our (Novell) QA. And of course, our openSUSE unstable repository is always updated with betas and RCs, this is where the community is involved in improving the go-oo quality. > And > most of the community members in the QA project are doing only manual > testing on UI level. So one major part of Quality Assurance (or QA > management) for Software is missing in your process. Having a different approach to the testing does not mean that it is worse. It is all it is - different, as our development style differs too. Regards, Kendy --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
