depends on what you are looking to use it for i guess.

the way we've been using @author is to tag people who design/change behavior
in a class - not by how important the change was.

so that if someone has a question about a certain class you can look at the
list of authors and go bug those people on the hows and whys.

svn revision history is too noisy because you see all the changes - spelling
fixes, the < to > changes, etc. you can ask the initial committer as that
person is always the author, but then you leave out all the other people who
have significantly changed/added functionality.

-Igor


On 10/3/06, Ross Gardler <[EMAIL PROTECTED]> wrote:

Matej Knopp wrote:
> Johan Compagner wrote:
>>>
>>> >  - shall we remove the @author tags (not forbidden but frowned upon
>>> @apache)
>>>
>>> Really, frowned upon? If I recall correctly, any Apache project I have
>>> seen the sources from have them... Now that most people are more or
>>> less maintaining them, I kind of like them.

The author tag is considered to be too difficult to manage within a
community project. At least within the ASF projets I work on, however, I
believe there are projects that still use it.

The general feeling, as I understand it, is that it opens the question
of "how much does one have to change before they deserve entry into the
author tag?". In a community driven project no contribution is more or
less important than any other. For example, a single character change
that fixes an elusive bug can be just as important as the contribution
of a new class, consider, for example, a '<' changed to a '>' in a
rarely executed conditional statement.


>> aarghh
>> no author tags? Then all my:
>>
>> @author Igor Vaynberg (ivaynberg)
>>
>> are removed?
>> How do we blame igor then?
> Exactly. I think also the @author tag is too useful to be just removed.
> So -1 on removing that.

SVN commit logs provide all the necessary information for tracking
individual contributions. Many projects also manage a status document
that describes significant changes in a code base.

Ross


Reply via email to