Hi All,

Getting caught up after being out last week…

 

Succinctly:  The use of any shared account to contribute code is strongly 
frowned upon regardless of whether the code under development is closed source 
or open source.  In fact at most companies doing so is grounds for termination.

 

I completely understand the concerns raised by Davide, but the current 
“workaround” is highly problematic for several reasons. From my analysis it 
tries to address a perceived problem rather than a real one.  Vodafone is most 
definitely NOT the only carrier with contractors performing work using an 
account within the carrier’s domain space.  I can point to dozens of such 
examples.  The likelihood of any such individual ever attempting run for a 
reserved seat in the context of representing the carrier is virtually 
non-existent. Doing so would not only put them jeopardy of losing their job by 
angering the carrier, the contract agency they work for would be at risk of 
losing the carrier’s business altogether. 

 

Given the current criteria and “named-nine” reserved carrier seats, if Active 
Community Member [email protected] <mailto:[email protected]>  is a 
contract developer:

*       Is it likely she would intentionally run as CarrierX’s candidate?  No.
*       Should she be able to vote in the election? Yes.
*       Should she be able to run in the election for a non-reserved seat as an 
employee of her staffing firm? Yes.
*       If for some reason she is the only person from the carrierx.com domain 
that qualifies as an Active Community Member, then the TSC rep for CarrierX 
simply needs to communicate that BobbySue is really a contractor developer from 
XYZ Staffing Firm and not an actual CarrierX employee.

Also to clarify another possible misconception, Davide you may not have been 
aware of it but Vodafone voted in favor of the TSC criteria and definitions 
that are currently in place today, so to say that “We have always opposed that 
rule”, is a bit inaccurate.  Objections were not raised by Vodafone until after 
the vote was passed and the provisions took effect.

https://wiki.onap.org/display/DW/TSC+2018-07-12

https://lists.onap.org/g/onap-tsc/message/3364

 

As a final note to everyone, the criteria and definitions for the TSC elections 
will be coming up for review in the near future. 

If there things you would like to see adjusted, please work directly with your 
TSC rep.

 

Hope sharing my perspective helps everyone to move forwardly constructively.

Let’s get those contactors doing their own commits 😊

 

-kenny

 

 

From: Williams, Marcus <[email protected]> 
Sent: Wednesday, April 10, 2019 8:05 AM
To: Cherubini, Davide, Vodafone Group <[email protected]>; Seshu m 
<[email protected]>; [email protected]; Kenny Paul 
<[email protected]>; Phil Robb <[email protected]>
Cc: Abu Aisheh, Razanne, Vodafone Group <[email protected]>
Subject: RE: Commit Code on Common Account? - SO-1421 code review

 

Davide, Thanks for explaining your reasoning. 

 

*       TSC, what is your position on code authorship?

 

Vodafone is submitting code under a group account due to the TSC election 
eligibility rules (see Davide’s email below). They are putting co-authors in 
the patches (who are the author), but the author and committer appears as  
<https://gerrit.onap.org/r/#/q/owner:onap%2540vodafone.com+status:open> 
Vodafone Group. This seems like bad open source practice and precedent to me 
and at first I -2 their patch.

 

I’m a committer in SO, APPC and former committer in SDNC, CCSDK, OpenDaylight 
and OPNFV. I’ve contributed to those projects and OpenStack. Based on Open 
Source precedent with very few exceptions we should be pushing, authoring, 
committing and merging code as individuals. The reason other open source 
projects don’t generally allow this is that it opens a host of code 
administration and management issues for large projects and it generally 
increases the amount of time the already time strapped committers/PTLs need to 
take when approving/managing code base.

*       Is there a middle ground on TSC elections eligibility we can find?
*       Or is this practice fine in the TSC’s mind?

 

Thanks,

 

Marcus Williams

IRC, Twitter, etc. @ mgkwill

Intel Corp.

 

From: Cherubini, Davide, Vodafone Group [mailto:[email protected]] 
Sent: Wednesday, April 10, 2019 7:42 AM
To: Williams, Marcus <[email protected] 
<mailto:[email protected]> >; Seshu m <[email protected] 
<mailto:[email protected]> >
Cc: Abu Aisheh, Razanne, Vodafone Group <[email protected] 
<mailto:[email protected]> >
Subject: Re: SO-1421 code review

 

Hi Marcus,

 

I agree with you the best practice is to commit the code with real names. 

 

The reason behind the decision to commit with a common email (and have the 
engineer who wrote the code added as co-author) is based on the fact that stats 
are registered by “author” (which is , btw, correct). 

 

However, those stats are then used to “identify” the person that can be elected 
to the TSC (that’s a rule decided some time ago). We have always opposed that 
rule because Vodafone is a very big company and we use external (read 
non-Vodafone) developers to write part of the code. In the past (although 
briefly) it has come to my attention that one of those external consultants 
resulted to be the “TSC-eligible” person for Vodafone which, of course, it does 
not make any sense.

 

So the options I see are 3:

1.      We review the TSC rule, e.g., by measuring the stats per Company and 
not per Individual Contributor. In that way the external developers (who have a 
Vodafone email account) can commit the code using their names and avoid the 
risk to become the TSC candidate (the TSC candidate for Vodafone should be 
decided BY Vodafone)
2.      The external developer writes the code and then she/he submits it using 
one of the Vodafone architects accounts. I don’t like this option because, 
e.g., I should share my LFN account details with external employees
3.      The external developer writes the code and then she/he submits it using 
the common Vodafone account (the name of the developer is always added as 
“co-author”). This is our current option

I hope this clarifies our decision.

 

Many thanks

 


 <http://vodafone.com/> 

Davide Cherubini 
Lead Software, Open Source and Labs 
Strategy, Planning and Operations

Cloud & Automation CoE


Babbage House, The Connection, Newbury, Berkshire RG14 2FN 
+44 7770 700172 


 <mailto:[email protected]> [email protected] 
 <http://vodafone.com/> vodafone.com


The future is exciting. 
Ready?

 

 

 

From: "Williams, Marcus" <[email protected] 
<mailto:[email protected]> >
Date: Wednesday, 10 April 2019 at 15:20
To: Razanne Abu Aisheh <[email protected] 
<mailto:[email protected]> >, Seshu m <[email protected] 
<mailto:[email protected]> >
Cc: Davide Cherubini <[email protected] 
<mailto:[email protected]> >
Subject: RE: SO-1421 code review

 

HI All,

 

I’ve removed my -2 based on our chat in the SO Weekly Call. I still think this 
is bad precedent to set in an Open Source project. What is Vodafone’s issue 
with having engineers submit their code with their name and Vodafone email, 
just like everyone else? I think we should have a chat in TSC about code 
authorship.

 

The reason other open source projects don’t generally allow this is that it 
opens a host of code administration and management issues for large projects 
and it generally increases the amount of time the already time strapped 
committers/PTLs need to take when approving/managing code.

 

Just my 2 cents.

 

Thanks,

 

Marcus Williams

IRC, Twitter, etc. @ mgkwill

Intel Corp.

 

From: Abu Aisheh, Razanne, Vodafone Group 
[mailto:[email protected]] 
Sent: Wednesday, April 10, 2019 3:08 AM
To: Seshu m <[email protected] <mailto:[email protected]> >; 
Williams, Marcus <[email protected] <mailto:[email protected]> >
Cc: Cherubini, Davide, Vodafone Group <[email protected] 
<mailto:[email protected]> >
Subject: SO-1421 code review

 

Hi Seshu/Marcus,

 

We’ve got a comment from Marcus “Please, rebase this change and submit it with 
a person as the author and committer” for below given two commits:

1.       SO:  <https://gerrit.onap.org/r/#/c/82798/> 
https://gerrit.onap.org/r/#/c/82798/

2.       SO/chef-repo:  <https://gerrit.onap.org/r/#/c/82816/> 
https://gerrit.onap.org/r/#/c/82816/

However, the commit for OOM ( <https://gerrit.onap.org/r/#/c/84475/> 
https://gerrit.onap.org/r/#/c/84475/) seems to have been accepted and merged 
with the same Vodafone id. We have been using Vodafone ONAP competency ID to 
commit the code into ONAP Gerrit. The mail box for this email ID is actively 
monitored by the members of the ONAP competency group and for additional 
reference, we mentioned the ID of the developer in the Co-Authored field.

 

PFA – The code that is committed by the same id and merged with master branch.

 

Regards,

Razanne


-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.

View/Reply Online (#4860): https://lists.onap.org/g/onap-tsc/message/4860
Mute This Topic: https://lists.onap.org/mt/31019923/21656
Group Owner: [email protected]
Unsubscribe: https://lists.onap.org/g/onap-tsc/leave/2743226/1412191262/xyzzy  
[[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

Reply via email to