Sorry for my delay. First: Sorry Ceki for not knowing who you are! I am thankful for all the work you put in logging!
Topic: Gavin speaks out of my heart. I prepared an e-mail before I got Gavins and now discarded mine. However, I am afraid before that bureaucracy in sourcecode. It never came to my mind that one could feel that it is "unfair" that his lines of code are exactly noted. Furthermore, if one contributes 3 lines of code an a comment makes clear who did that, what is if one has to make a fix in the 3 lines? Isn't it unfair too that his fix probably would not get the same attention? Or if the fix gets the same attention as the original? I would prefer not to think about fairness in sourcecode comments. I would prefer to be fair in a team.xml or NOTICE or Changes.xml file as already suggested. I also think that @author-tags are bad for team-building. Reading "who is responsible for a class" makes me afraid if I want to propose changes. I think it's a bit different at apache (cause people are cool), but I never would use author tags in a commercial projects for this. However, after this dicussion I decided to search the mailinglists. there was already atempt to remove the author tags: http://www.mailinglistarchive.com/[email protected]/msg00068.html I think this lead to the issue. I will try to find out the e-mail adresses of the original authors and will ask them if they are fine with moving their names to changes.xml or wherever. If all agree, no problem, if none agrees we have to discuss more. Maybe there are others jumping into discussion Cheers! Christian On Mon, Apr 27, 2009 at 2:19 PM, Gavin <[email protected]> wrote: > > >> -----Original Message----- >> From: Ceki Gulcu [mailto:[email protected]] >> Sent: Monday, 27 April 2009 10:04 PM >> To: Log4PHP Dev >> Subject: Re: Getting rid of author tags: >> >> Hello Gavin, >> >> It looks like you have considered the issue and made up your mind. > > Nope not yet :) It is the way that it is being done on the other project I'm > involved in. As said, some work one way some work the other. I'm open to all > suggestions. And you have made some valid arguments. > > Lets see if we get other opinions on the matter first. > > Thanks > > Gav... > >> It is >> not >> hard to establish a authorship tag policy for contributors. Any >> contribution of >> over 10 lines (just an indication) merits mention as an author with an >> author >> tag. It is a very easy policy to follow. >> >> Moreover, author tags are not exclusive. Nothing prevents the creation of >> a >> central team-members page. Author tags provide an easy and *fair* way to >> reward >> for contribution, regardless of the number of contributors. >> >> > Apart from having their name in lights as it were in those few files it >> > points out only to other coders that look in those few files that they >> > contributed. It also points out that they did not contribute to the >> other >> > 200 or so files and so actually it makes them less worthy than those >> initial >> > contributors that are named in all files. This does not give incentive. >> >> If a contributor makes a contribution to N files, then they are mentioned >> as an >> author in N files, as should be the case in a meritocracy. The fact that >> they >> did not contribute to other files is a mere given of this hypothetical >> case. >> If making less of contributions makes those contributors feel less, then >> they >> should contribute more. Treating all contributors equal, while ingrained >> in the >> collective Apache psyche, is not an incentive, on the contrary, it is an >> incentive killer. >> >> > Having multiple @authors and contributors named in these files also >> takes >> > away the limelight from the original @author(s). >> >> I hope that you see the logical flaw in this reasoning. In order to avoid >> taking >> away the limelight away from the original authors, you are ready to >> penalize all >> authors, new and old. >> >> There exists a fair alternative. Add authorship tags as the project moves >> along. >> New author tags can be added in either alphabetical or chronological >> order. >> >> > And what those other than only contribute in terms of documentation >> and/or >> > applying of patches put into Jira from passers by that contribute a >> line or >> > two to fix a bug? It can also get messy maintenance-wise into the >> future. >> >> Documents can mention their author or authors. The same procedure applies >> to >> documentation as to source code. As for passers by who contribute a line >> or two, >> they can be mentioned as contributors in a line of comment in the source >> code. >> This line of comment does not carry an @author tag. >> >> Anyway, we can argue about @author tags all day. If you feel strongly >> about >> removing them, then go ahead and remove them. >> >> Gavin wrote: >> >> > Having @author tags is great when a program only has that initial author >> or >> > when one or two more join in his/her efforts. At that stage it is also >> > unlikely that the author uses much in the way of attributing their works >> in >> > cvs/svn commit messages, issue tracker contributions etc etc. So having >> the >> > @author tag is a good way of identifying the origins of the code. >> > >> > When it becomes a project that comes with all the maintenance overhead >> of >> > website, issue tracker, wiki, official releases, documentation as well >> as >> > the code itself then moving forward I see things differently. >> > >> > Supposing over the next 2 years another 15 committers join the ranks >> > contributing to log4php (well there always hope :) ). We need to look >> ahead >> > at what will work best. If 6 or so of those new committers contribute >> code >> > to maybe 10 or so existing files and another 4 or 5 of their own then >> you >> > suggest they add their names to the @author tags in those files. >> > >> > Apart from having their name in lights as it were in those few files it >> > points out only to other coders that look in those few files that they >> > contributed. It also points out that they did not contribute to the >> other >> > 200 or so files and so actually it makes them less worthy than those >> initial >> > contributors that are named in all files. This does not give incentive. >> > Having multiple @authors and contributors named in these files also >> takes >> > away the limelight from the original @author(s). >> > >> > And what those other than only contribute in terms of documentation >> and/or >> > applying of patches put into Jira from passers by that contribute a line >> or >> > two to fix a bug? It can also get messy maintenance-wise into the >> future. >> > >> > Having said all the above, I do not want to take away from the original >> > author(s) and what it is that they achieved before that CLA was signed >> > handing over all code as donation to the ASF. >> > >> > I (and Christian no doubt) would be extremely happy to see Marco >> > Vassura' name on a page on the main website of log4php somewhere, on the >> > changes-report pages listed with him as having donated the original code >> > along with any other names from the initial import. Other contributors >> who >> > were not part of the original code import would also be listed on the >> > changes-report page. Perhaps they could also be listed on the main >> logging >> > http://logging.apache.org/team-list.html or log4php could create its own >> > team-list with all @authors listed in the Contributors section. >> > >> > For the future and moving on, all contributors/committers etc would be >> > listed in the changes-report for any significant code changes that >> warrant >> > listing on there. Any contributors providing patches will do so via Jira >> so >> > we will know where that bug fix/feature came from and a committer will >> > mention there name and link to the Jira issue in their commit of that >> code. >> > >> > Having contributors/committers/PMC Members listed in one place on the >> > website and in the changes-report and in svn commits is I think a better >> way >> > to show who is/was part of a project. It also gives everyone involved in >> > that project equal limelight, in other words no one contributor should >> be >> > considered better than another. I think this far better exposure than >> hiding >> > away in an @author tag that only coding peers are likely to look at -- >> > remember that much software is created for users, users that are >> comfortable >> > with seeing a how to use manual rather than having to look under the >> bonnet. >> > >> > As log4php is a project destined to become a sub-project of logging >> > services, it is in our best interests to interact and align ourselves >> well >> > for the future. It is also in this projects interests to be independent >> in >> > its own right as far as possible. >> > >> > I'd welcome your thoughts further on what I've said above before we >> decide >> > and move on towards a first release. >> > >> > Thanks >> > >> > Gav... >> > >> > >> > >> > >> >> Christian Grobmeier wrote: >> >>> Hi, >> >>> >> >>> I have a commit ready for getting rid of the author tags and would >> >>> like to apply. Since a user complained about this and I am unsure >> >>> about all this legal stuff, I would like to have your "go"s for this >> >>> commit. (Please don't check the patch applied to the bug, I have >> >>> re-created it) >> >>> >> >>> In my opinion it's enough to give credits in the NOTICE file of the >> >>> project. We do this at Commons to. I also think that it is enough to >> >>> credit real names, not names like VXA or something. >> >>> >> >>> See also: https://issues.apache.org/jira/browse/LOG4PHP-1 >> >>> >> >>> If no other opinions occur, I will go ahead and commit it. >> >>> >> >>> Cheers, >> >>> Christian >> >> -- >> >> Ceki Gülcü >> >> Logback: The reliable, generic, fast and flexible logging framework for >> >> Java. >> >> http://logback.qos.ch >> >> >> >> >> >> -- >> >> No virus found in this incoming message. >> >> Checked by AVG. >> >> Version: 7.5.557 / Virus Database: 270.12.4/2081 - Release Date: >> 4/26/2009 >> >> 9:44 AM >> > >> >> -- >> Ceki Gülcü >> Logback: The reliable, generic, fast and flexible logging framework for >> Java. >> http://logback.qos.ch >> >> >> -- >> No virus found in this incoming message. >> Checked by AVG. >> Version: 7.5.557 / Virus Database: 270.12.4/2081 - Release Date: 4/26/2009 >> 9:44 AM > >
