incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ted Dunning <>
Subject Re: [Result][VOTE] Graduate Apache Kylin from the Apache Incubator
Date Tue, 03 Nov 2015 01:25:30 GMT
On Mon, Nov 2, 2015 at 5:17 PM, John D. Ament <> wrote:

> Actually, 5/5 releases were done without the SGA.  I'm not too worried
> about that for a number of reasons.

The missing SGA was a defect in form, but not in substance.  All of the
original code came from eBay who had a strong commitment to open sourcing
the code as evidenced by their placing the code on github under ALS.

> 1. We imported the full history from the original repository, the podling
> has built upon that instead of flattening the commit.


How is this a defect? It increases transparency.

> 2. The code coming in was already AL v2 compliant, so its really a question
> as to whether releases were proper.

Correct modulo the kinds of nits that tend to be surfaced by incubator

> 3.  From continuing to look into this, I also noticed we're missing an IP
> Clearance for the donation.  In addition, there are 31 contributors per
> github, but the project proposal includes 8, the current roster shows 13 +
> 5 mentors.  So there is a potential gap related to IP.

All of the original contributors to the code were employees of eBay and
thus, there was no IP problem from then. Later contributors may have been
from outside of the committer circles, but all commits were pulled into
Apache by a committer. The Apache push logs record exactly who brought the
commits into Apache while commits themselves record who actually authored
the original code. This is relatively standard for git operations at Apache.

> 4. We've had prior cases where podlings failed to get the SGA done properly
> before starting incubation.  It happens, and as long as we can find an SGA
> indicating the initial import we should be fine

We have that.

> Sorry, I feel like every time I dig into this I find something new. :/

I am sorry you get that impression.

Other than the SGA (now filed, acknowledged by the project to be late in
obsequious subsequent prostrations) and the font license (covered by the
acks in the containing package and the overall project acknowledgements and
difficult to mark otherwise) what new actual issues have you uncovered?

My own opinion is that each of the issue you just mentioned are actually
virtues rather than vices.

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message