This release is really bad for Padre. The old branch was stable besides one crash bug which was covered at the same day the branch was frozen. Adding this fix would have been a great release.
The last 200 commits introduced a number of critical bugs breaking important features of Padre (like find in files, entering one blank line into the editor) - but noone cared. The bugs were reported as "blocker" state tickets - but noone cared. Padre was released containing with noone caring that it was really unusable for productional work at this time. Today, the new blocker bugs were fixed and it turned out that all of them plus another one from #105 had been typos, logic errors and simple things like this. I'm aware that not everybody could run the unit tests before each commit - including me (with 30 to 45 minutes for a full test run on my hardware), but all these bugs (which burned about two or three hours of Padre development time) were really easy detectable by running "dev" and trying out the chanced feature. The newline issue wasn't easy to try out, because it was unexpected, but it caught me within 30 seconds after downloading the bug using "svn up", it must have been the same for the guy committing the bug. Why was it not fixed? Anyway, we got a release out there which is heavily broken and the release notes doesn't even flag this release as "unstable/testing", so we're going to annoy many users and maybe loose some of them. I suggest two new global Padre development rules: 1. Before committing anything, run your local Padre trunk copy and do a quick test of what you changed. It's ok to commit broken things and maybe even syntax errors if it's written down in the commit notice and either fixed within a very short time (commit to switch from work to home PC) or a "blocker" state ticket is being opened/the ticket dealing with it changed to "blocker" state. 2. No release is being done as long as there is any "blocker" ticket open. The release manager has to walk though all open blockers and either downgrade them to a lower priority or clearly ensure that the bug was introduced after branching and hasn't been merged to branch. Of cause, he may also call for checking blockers like calling for translators, I'm pretty sure bowtie, zenog and others will jump in and take care of the blockers. #1268 is no longer blocker, this should have been downgraded shortly after upgrading it. #1270 is an enhancement, I doubt it could be a blocker. Should be downgraded to a real value. Your comments, please. Sorry for being that aggressive, I don't like wasting dev time :-( Sebastian _______________________________________________ Padre-dev mailing list Padre-dev@perlide.org http://mail.perlide.org/mailman/listinfo/padre-dev
