March 23, 2004
Removing author names from their work

Recently the ASF board issued a recomandation that author tags in java code are "deprecated". They believe this is better for the community and there are legal issues with having them.

I personally believe removing the name of the authors from their work is morally wrong and unacceptable in an open source environment. Even if ASF holds the copyright ( because all code authors assigned it and donated their work ), and even if the consensus of the current community is to remove the names - it still can't justify such act.

Ant project implemented the board recomandation, and ( with a lot of regret ) I had to quit the ant PMC and project. There are already voices in jakarta supporting the board's position, so I'm already preparing myself to leave Jakarta as well if they chose to do it.

For me it is a moral issue - it doesn't matter if an author is no longer in Apache ( or is now working for SCO or competing projects, dead or doesn't care ) - removing the author names from the code they wrote is as bad as removing it from an published article or book or research paper. I am willing to remove my name from Ant/Jakarta - if the community believe this is better for them, but I can't accept removing other people names ( or mine ) without explicit permission from the authors.

Quite a few people tried to spin this and somehow accuse the authors of too much "ego" or "brick buying" or make it look like signing on your work is against the community spirit. After all, the code is written by a community and any suggestion that there are actually individual people who wrote the code would discourage others to help. This is usually backed with examples from the Avalon or Cocoon community - probably considered a perfect model to be followed by others.

There is also the silly argument that spammers ( or just people needing help ) will collect the names from the code, look them up on google to find emails, and then flood the author with spam or questions. Like our postings on the mailing lists and the public archives and the news gateways don't exist. It is each person choice to not respond when people ask for help, or to not include their name in the code if they think that would make others to ask their opinion wrt code they wrote - but there is no reason to assume everyone has the same attitude and remove their name as well.

If you are willing to accept that code is a form of communication or speach - then removing the author name from the code is as bad as removing it from a book and claiming the book was written by the "community". If programming is a form of science - then code and authors should be given the same respect as in a scientific paper, where typically authors names are not removed from their research.

The vague legal arguments used by the ASF board ( you can find the exact text on few public mailing lists, I don't know if it was officially released so I won't include the quote ) are a problem by themself. It is too often that ASF board is using legal arguments ( which are hard to fight by the programmer community ) to push their view. I can accept that ASF has better lawyers then FSF or Mozilla or all other open source projects that credit the authors in their code and don't push the development process to an extreme of control. But if all other open source lawyers are incompetent and are not actually protecting their contributors - I think we deserve some more explanations to justify this point of view, and maybe we should also warn them of the dangers. It's not only @author tags - there is also the claim that committers can't have any right on the code or community unless they are in the PMC ( project management commitee ), or that they are not legally protected unless something.

It seems a lot of people believe that Apache code is so good because of some "process" that is followed and it's actually the community, management and great environment that deserve all credit. Somehow it's the "community builders" that are responsible for the creation and success of those communities, by teaching the regular coders the virtues of ego-less open source. I've allways believed in community - and accepted a lot of shit and unfairness for it's sake, but I think it's the individuals that form the community who make the difference and deserve credit too.

If companies can be destroyed by arogance, the same holds true for open source projects.

Other opinions:Ovidiu points to other projects who preserve the author names. It is interesting to look at linux or FSF sources and see the long list of people who have contributed, and the fact that those projects don't consider it a problem.

Posted by costin at March 23, 2004 11:05 AM