Re: Re: PFC-RFC Update

From: Date: Wed, 26 Feb 2003 22:31:57 +0000
Subject: Re: Re: PFC-RFC Update
References: 1 2 3  Groups: php.pear.dev 
Request: Send a blank email to pear-dev+get-13948@lists.php.net to get a copy of this message
"Tobias Schlitt" <toby@php.net> wrote in message news:20030226195935.11254.qmail@pb1.pair.com... > > Either you propose a new PEAR-QA-Group or there's a missunderstanding > > about my proposal. My proposed PFC-core-group (aka PFC-gc) should not > > have to do QA for all those packages out there. That would be a hell > > of a job... PFC-gc is merely there to control, if a package fullfills > > all requirements. They base this desicion on some kind of report the 2 > > independent peer-reviewer provide about the package. If they don't > > trust the peer-reviewer, they still can dig deeper into the proposed > > packages. > > I understand the difference, but I'm still of the opinion that > introducing PFC will divide PEAR in to classes of packages. Perhaps this is a good thing. Basically, the PFC status divides packages that are well-written, testable, and well-documented from those that aren't. The user benefits 100% from this. PEAR developers suffer a bit, but only at first. Everyone knows that a well-written, testable, well-documented package is simpler to maintain and upgrade in the future :). PEAR can only benefit from this. The better the code is, the more likely it is that companies will pay developers to develop PHP, continuing the process of making it better for all who use it. Having said this, there are some packages that may have difficulty with one or more of these requirements. For instance, how does one unit-test a PHP parser like phpDocumentor? I have been inspired by this thread to add unit testing to the 2.0 release, if possible. phpDocumentor has several extremely complicated processes in its design, and it will be very good to re-evaluate how these things are done. As an added bonus, if I can figure out or get help doing the unit testing, I'd be happy to attempt to write extensive docs for PHPUnit to augment the current docs. In any case, the process of making phpDocumentor unit-testable will be an extraordinary effort, and take a long time, and in that time, phpDocumentor won't qualify for PFC status, and that's fine. People will still use it because it has been tested rigorously by many users and passed that test. However, phpDocumentor isn't really a class, it's an application, so should we have a separate measure for PFA? Remember, Stig envisions implementing the "A" part of PEAR sometime in the future too... Can of worms, open. Enjoy. Greg -- phpDocumentor http://www.phpdoc.org

« previous php.pear.dev (#13948) next »