Recent Features
| |||||||||||||
All rights reservedAll rights reservedPosted Jan 17, 2013 15:50 UTC (Thu) by mjw (subscriber, #16740)In reply to: All rights reserved by rcweir Parent article: A discordant symphony
> You're missing the forest for the trees if you are fixated on file headers.
But even other Apache hackers don't want or can edit those files currently because IBM seems to refuse to correct the headers:
And it seems ASF policy is (reasonably) only the copyright holder may change the license headers on files:
And others have commented that the current situation is so murky that it is really hard to see who has rights to hack on and redistribute which files. So effectively nobody but IBM employees can work on this code even though technically it has been "donated" to Apache. Would it really be that hard to get IBM to fix that? Just like Oracle eventually fixed the headers for the original openoffice.org source files. It would mean a lot more people can actually hack on it and get more of the code in the hands of end users sooner. Thanks.
(Log in to post comments)
All rights reserved Posted Jan 17, 2013 16:14 UTC (Thu) by rcweir (subscriber, #48888) [Link]
>But even other Apache hackers don't want or can edit those files currently >because IBM seems to refuse to correct the headers:
Confusion is not limited to Lwn.net, clearly. But I hope my comments have clarified the state of these files.
If anyone really wants to help integrate these files, then post a note to dev@openoffice.apache.org and introduce yourself. I'll be more than happy to give you some hints and hopefully remove any doubts you might have.
But I'll be honest, so far all I've heard so far is complaints from non-lawyers and non-coders who are not interested in helping with the project. They tend not to get a lot of my time.
-Rob
All rights reserved Posted Jan 21, 2013 8:14 UTC (Mon) by ceplm (subscriber, #41334) [Link]
I would say that releasing source code under conflicting licenses and not fixing the headers when reminded about it seems to me like a classic case of spreading Fear, Uncertainty, and Doubt about the legal status among potential users (e.g., LO developers). Don’t they have sed(1) in IBM?
Matěj Cepl
All rights reserved Posted Jan 21, 2013 14:15 UTC (Mon) by rcweir (subscriber, #48888) [Link]
Actually, we are "fixing the headers" as we merge the code into OpenOffice. That is how we will be publishing these features. We won't publish a release containing Symphony code until that is done. The code we release is extensively reviewed and the licenses and notices carefully documented. That is part of the value-add that Apache releases are known for.
Your complaint seems to be that we're merging the Symphony code into OpenOffice rather than releasing it as-is. As stated before, the community considered both options but decided to do the merge approach. This is not FUD. This was done out of careful consideration of the benefits and liabilities of either approach. The discussion was held in public, on our mailing lists, which you are free to examine at your leisure.
Regards,
-Rob
All rights reserved Posted Jan 21, 2013 14:38 UTC (Mon) by ceplm (subscriber, #41334) [Link]
> Your complaint seems to be that we're merging the Symphony code into OpenOffice rather than releasing it as-is. As stated before, the community considered both options but decided to do the merge approach.
No, it isn't my complaint ... what I would prefer would be to release Symphony (or parts which you want to release) as it is, unmerged, unrebased, raw, uncompiliable (if it happens), with bugs, smell and everything (but with copyright statements corrected ... see my comments about Perl scripting of that). You would have a nice precedent with open sourcing of Java by Sun (which was not compilable in the beginning, if you recall the history).
We are adults here and we can understand what does releasing of parts of the code means. Programmers are usually pretty good in fixing, patching, rebasing etc. and community of them can do it better than a closed group under adult supervision ... that’s kind of the point of the open source development, isn’t it?
Thank you for replying
All rights reserved Posted Jan 21, 2013 15:28 UTC (Mon) by rcweir (subscriber, #48888) [Link]
What makes you think the existing copyright statements are incorrect? I don't see any error. Remember, an SGA does not reassign copyright. It provides a license.
-Rob
All rights reserved Posted Jan 21, 2013 16:01 UTC (Mon) by mjw (subscriber, #16740) [Link]
You are talking past each other.
The question was for *IBM* to update the license headers of the files since they are the copyright holder and claim to have granted a license. The answer was that the *Apache* project decided not to change the license headers for now even though they believe they have a grant from IBM to do so.
It would probably help to be explicit whether a question is asked of Rob/IBM and/or if an answer is given by Apache/Rob.
The original article (and a lot of the comments here) contain confusion about whether or not IBM actually donated the code or not. Part of that confusion comes from whether or not Apache accepted it or not. The current situation of confusing/wrong license headers on the "contributed" files doesn't help make that situation very clear. The answer would be immediately clear if IBM just clarifies the issue by cleaning up the headers. Or hopefully Matthews poking will help make clear whether the files as is can be seen as already under the Apache license because of the SGA and being distributed by the ASF and who can update/clean up the headers to give everybody the warm fuzzy feeling that a contribution was actually made and available as Free Software to all users.
All rights reserved Posted Jan 18, 2013 13:02 UTC (Fri) by man_ls (subscriber, #15091) [Link] And others have commented that the current situation is so murky that it is really hard to see who has rights to hack on and redistribute which files.That is a pity. I think (IANALly) that the files have clearly been donated by IBM under the ASL, and that the contents of individual headers is not relevant since there is an overarching permission. A real world analogy: imagine that I sign a contract to sell you a pile of my books, but then you refuse to take them because some books have "property of man_ls" written in the first page, until I erase that line from all of them. Hey, the books are now legally yours; feel free to modify them as you see fit.
All rights reserved Posted Jan 18, 2013 14:52 UTC (Fri) by pboddie (subscriber, #50784) [Link]
Nice analogy! One thing that has to be considered is that statements of ownership are easily disregarded if everyone involved knows about agreements to the contrary, but the real problem is where people who are not aware of (or party to) such agreements see such statements and are then thrown into confusion about the real status of the work in question.
It would be a bit like someone pulling a book of the shelf in the aforementioned purchaser's house, having been told by the purchaser to "take any of my books if you like them", and then reading the "property of man_ls" statement. One would be obliged to check once again with the purchaser or potential owner to make sure that the book is not one that has been borrowed and then mixed up with the rest of the purchaser's collection.
Such manual checking that everything is still legal or acceptable is precisely the kind of thing that conflicts with Free Software licences because one of the aims of such licences is to eliminate such bureaucratic obstacles, especially when they can be misused to effectively limit distribution even if grand promises of generosity have been made.
All rights reserved Posted Jan 18, 2013 14:55 UTC (Fri) by man_ls (subscriber, #15091) [Link] Your comment is completely on target. Just wanted to say.
|
|||||||||||||