I believe the intention is to continue testing with a minor security hole. Then, fix all the problems uncovered while testing and start a new cycle. However a major problem would make all further testing pointless making it worth it to stop the cycle, fix the problem, and start over. This means we aren't testing a moving target. Once there are no more release critical problems (old or uncovered in the last testing cycle), then we can release.
I've seen the opposite in practice and it's not pretty. Basically regression testing is started, a flaw is found and fixed. However, they continue regression testing from the middle instead of starting from the beginning. The result is patches which fix problems introduced in other patches. Brandon Mitchell <[EMAIL PROTECTED]> SRA International 703-227-8326 (W) or 888-358-6065 (P) > -----Original Message----- > From: Wichert Akkerman [mailto:[EMAIL PROTECTED] > Sent: Thursday, April 20, 2000 8:55 AM > To: Richard Braakman > Cc: Martin Schulze; debian-release@lists.debian.org > Subject: Re: Preparing for first test cycle > > > Previously Richard Braakman wrote: > > No. Any such change means aborting the test cycle. This may > be reasonable > > if a security problem is big enough, but I'm not going to decide that in > > advance. > > I'm certainly very much against releasing with known security holes. > At this moment we already have to do 2.2.15+1patch :( > > Wichert. > > -- > ________________________________________________________________ > / Generally uninteresting signature - ignore at your convenience \ > | [EMAIL PROTECTED] http://www.liacs.nl/~wichert/ | > | 1024D/2FA3BC2D 576E 100B 518D 2F16 36B0 2805 3CB8 9250 2FA3 BC2D | > > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with a subject of "unsubscribe". Trouble? Contact > [EMAIL PROTECTED] > >
