directory-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Emmanuel Lecharny <>
Subject Re: svn commit: r782968 [1/7] - in /directory/sandbox/slp: ./ src/main/java/org/apache/directory/slp/ src/main/java/org/apache/directory/slp/codec/ src/main/java/org/apache/directory/slp/extensions/ src/main/java/org/apache/directory/slp/impl/ src/ma
Date Wed, 10 Jun 2009 21:03:23 GMT
Rellermeyer Jan Simon wrote:
> Hi Emmanuel, 
>> You just committed a bunch of code produced by two person (namely you
>> and Lorenz), one of them not being a committer (Lorenz). This is
>> obviously something we can't accept, as Lorenz is not a committer, for
>> legal reasons : as a committer, you transfer the ownership of your
>> code to The ASF. If you are not a committer, then the code is assumed
>> to be owned by the one who wrote the code. Sadly, we don't accept
>> proxies, so you can't endorse the code he wrote.
> Okay, sorry for this. I did not know that the sandboxes are under the same legal constraints.
In fact, I just put the focus on ownership, but there is also another 
aspect we have to take care of: IP issues. As a committer, you are 
supposed to understand the way The ASF works, and you are supposed to 
have donw the IP clearance before committing some code. The sandbax is 
not really different, as soon as it's created by a committer, becasue it 
can be moved to the trunk whenever the committer consider necessary to 
do so.

Otherwise, we have an incubator, where the IP issues and ownership 
issues can be cleared.

Ok, enough legal stuff, because this is a bit different in your case.
> Lorenz is a (now former) student of mine and since he wrote the code as part of a master
thesis, he had to transfer the rights for the code to ETH which in turn gave me the permission
to open source it. I have the forms (somewhere) on my desk. So from our legal point of view,
I can actually transfer the ownership of the entire code to the ASF. If this does not work
for Apache, we can surely find a better way.

he told me about that when I asked him if becoming a committer was an 
option. Namely, he told me that until he get an authorization from the 
ETH, the code he produced is not his own, so he cannot contribute to the 
project as a committer.

I don't think there is a problem for this transferm, but we probably 
have to ask Lorenz if he agrees.
>> - second option, we can vote Lorenz as a committer,
> Yes, I think this is a good idea since he will continue to work on the code. I already
asked him to file some patches.
It was a pleasure to work with him and having him as a committer will be 
great !
>> Assuming that the previous point has been fixed, it would be we better
>> to know which part has been written by Jan and which part has been
>> written by Lorenz (not that it's important, but at least for a correct
>> attribution).
> Every class has an @author tag attributing the original author. Do you need this information
in a different format?
We don't put names in @author tags, we refer to the project itself. 
However, as everything is stored on svn with the committer name, it's 
easy to trace. There are many rason we don't want to see direct 
attribution in the code :
- it gives the impression that the code is 'owned' but someone else than 
- we don't know if we should allow anyone who modified a portion of a 
code to add his name in the file
- and it leads some people to contact the one whose name is in the file, 
something we want to avoid.

Thanks Jan, I suggest that for the moment we keep the code in the 
sandbox, and try to get all those legals stuff out of the way ! As soon 
as everyone is aware of what's going on, there is no harm.

cordialement, regards,
Emmanuel L├ęcharny

View raw message