Re: Minutes for PEAR Group Meeting 2008-10-05

> > Second problem is that all tests are run in a row. When on test makes a > fatal error, the whole suite suffers. This might change with PHPUnit4 > and test forking, but it does not work now. > Yeah, that is a pain. If we had a neat way to tell when the build was broken; which email'd pear-qa or pear-commit; that would probably reduce the problem. > > What I'm a little proud of is the phpt integration into phpunit. > Yeah it's pretty neat :) > > The suite could and should be made better, but I'm too occupied with > the peardoc transition at the moment. If someone would take it over, I > can give you ssh access to my server. > Well; in terms of available tests; I've done: http://pear.php.net/bugs/bug.php?id=14766 http://pear.php.net/bugs/bug.php?id=14746 ... and hassled people who's code breaks the suite a bit. As for the actual pear testrunner; though I've hacked away at work on PHPUnit, I haven't really done too much fancy. This might be a silly question, but have we looked much at PHPunderControl? I have cobbled together a similar CI environment for work to what you are doin; and know how much of a pain it can be to keep it functioning - whereas phpundercontrol seems a reasonably polished tool. The only thing is its use of cruisecontrol instead of phing; which I have 0 experience with (though I know phing is a PITA at times too!) > I planned to convert all my shell scripts to phing, but didn't get > really far. The scripts are in pear's svn, sandbox/QA_PEAR_CI. > The doc coverage code is in PEAR itself, QA_PEAR_Coverage or so. > Some files could also lurk around on my own svn server, svn.cweiske.de > in "tools" or "php" (public read access). > > -- > Regards/Mit freundlichen Grüßen > Christian Weiske > > -= Geeking around in the name of science since 1982 =- > -- Looking for a new php job? See what you can do with https://vx.valex.com.au/tests/season/