I agree that using a bug tracking system is a useful way to track (and not 
lose) contributions - however, I can't find anything mentioning 'scope' 
requiring a CLA, in ASF or LS docs.  

This particular commit is very limited in scope, and has nothing to do with the 
other logging projects - only that it enables them to write a reverse-connect 
appender, similar to our SocketHubAppender.  There are three source files.  As 
Elias mentioned, it's essentially a duplicate of code from SocketHubReceiver - 
hardly controversial.

The CLA is a requirement for those of us making commits.  For contributors to 
mailing lists, it seems that isn't the case.  At a minimum, different projects 
may have differing policies, and I don't think LS has a policy above and beyond 
what's described as required on the ASF license page.

Accepting contributions via mailing list is a common practice.  I assume that's 
one of the reasons mailing lists were explicitly described in ASL 2.0 as one 
way for a 'contribution' to be 'submitted'.  The fact that he applied the ASL 
2.0 license to his source files and submitted then to a mailing list maintained 
by the ASL seems to qualify as a valid contribution.  At that point.  There are 
steps folks can take to try and make the odds that a patch will be committed, 
to be sure, but it's not a requirement (I could just have easily created a 
bugzilla issue for him).

My primary interest in responding is to make sure we do everything we can to 
-encourage- contributions from the community.  Anything we can do to remove 
barriers to contribution is positive (assuming, of course, that all ASF 
policies are adhered to).

Scott Deboy
COMOTIV SYSTEMS
111 SW Columbia Street Ste. 950
Portland, OR  97201

Telephone:      503.224.7496
Cell:           503.997.1367
Fax:            503.222.0185

[EMAIL PROTECTED]

www.comotivsystems.com



-----Original Message-----
From: Curt Arnold [mailto:[EMAIL PROTECTED]
Sent: Mon 11/20/2006 8:26 PM
To: Log4J Developers List
Subject: Re: XMLSocketHubReceiver available - answer to Scott Deboy  e-mail
 
> But you're not giving him "commit rights".  He's just contributing  
> code and you are going to commit it for him, no?  To me, that  
> sentence only applies to people like Elias Ross, who we are voting  
> on to give commit rights (and he's already signed the CLA).
>
> Jake

It is more a question of the scope of the contribution, not the  
frequency of contributions.  There is a judgement call as to the  
point an signed CLA is required to accept a contribution.  I had not  
reviewed the patch but that it required an archive and not just a  
simple patch file and its applicability spanned multiple projects  
(log4net, log4j and log4cxx) suggested that it might end up over the  
line into the area that a signed CLA would be justified.  Elias had a  
signed CLA on file since his concurrent appender contribution was of  
sufficient scale that it justified a signed CLA to have his explicit  
statement that the code was his original creation, etc.  In any case,  
a contributor should at least be familiar with the CLA and have no  
objections to it even for the most trivial of contributions even if  
we don't have a CLA on file.

Having the patch attached to a bug report by the original contributor  
is an essential bit of process so that we can effectively track the  
chain of custody of any code in case there are any dispute.  Mailing  
list archives often do not archive attachments and using the bug  
tracking system as the single point for significant code contributions.




---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


<<winmail.dat>>

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to