Re: Something about PEAR policy

In article <3B5A24ED.9754AA7E@fast.no>, Stig S. Bakken wrote: >nloboek@e35203.upc-e.chello.nl wrote: [forking thoughts] >> Why would it be considered a failure? If one thing leads to another good >> thing it should make the original coders proud, even if the forked project >> thinks the original code is bad or lacks x, y or z. > >_I_ would consider it a failure because of my own personal goals, which >are more about creating a community for sharing good, reusable code. >This community should of course be able to replace bad code, but if it >has to replace itself, well... :-) I can understand :). >> I think a lack of documentation is the real problem. Most coders are not >> familiar with the api documentation that is currently used. > >I agree totally, but lack of (easily accessible) documentation is a >bigger problem for "end users". The situation I was trying to describe >was the one among code contributors. I understand. But consider the following. I started a project where developers from around the world can contribute. We started off with a codebase that lacked a lot of documentation (within and outside of the classes). Drews suggested we would use phpdoc, which does give the users a very good api documentation, and so we did. It took me about two months to install and document all of the the classes. I was hoping it would help people to jump into the project and start working with this api. It all ended up in very large files that nobody seem to understand. A month ago I decided to go back to the original 'comment what you are coding' style and from that day on developers are sending me solutions, patches, ideas, etc... that didn't seem to work with the phpdoc api. This doesn't state that phpdoc is bad, I just think that a good documentation for both phpdoc and PEAR would give a real boost to the project. The PEAR repository is a very good one, people with a lack of time should just find a good set of docs to read about it. It might help people to jump into using PEAR. Regards, Hans -- http://phpreview.nl.linux.org > > - Stig