cloudstack-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Sam Ruby <>
Subject Re: External fork of Cloudstack (was Re: [GitHub] cloudstack pull request: Is the project attempting a fork on Githu...)
Date Sat, 19 Mar 2016 14:18:24 GMT
On Sat, Mar 19, 2016 at 9:50 AM, Daan Hoogland <> wrote:
> On Fri, Mar 18, 2016 at 8:24 PM, Jim Jagielski <> wrote:
>> That sounds like a cop-out to me related to what's really going
>> on.
> Jim, I am not a native english speaker and this remark has no meaning to
> me. It sounds somewhat hostile, can you explain what you mean?

I think there was a misunderstanding and history involved, and I'm
working to clean that up.

The history involved is not unique to CloudStack.  It often comes up
when large projects or large companies are involved.  I work for a
large company (IBM), and often when IBM contemplates donating a
project to the ASF, the people involved are referred to me.  The most
extreme example I recall was when a high level executive told me he
wanted to take a project to Apache, but wanted a different license, to
be able to control who got commit access, and to run the project on
different hardware.  My response was simply: "then you don't want to
come to the ASF".

Here we had the CloudStack PMC make a reasonable request to ASF
infrastructure team (i.e., for more granular permissions), and were
not only told no, but that their request was placed on the back
burner.  I'm not proud of that response.  A technical solution to the
problem was developed (kudos!) and a proof of concept was deployed
(cool!).  Unfortunately, the proof of concept was poorly communicated,
and many (not just Jim!) saw this as an unfriendly act.  And to be
very clear, the optics were very bad: within 5 days of opening a JIRA
that was rejected, the CloudStack team looked like they were
unilaterally moving off of the ASF provided GitHub repository.

I don't think that there are any easy answers.  In particular, I don't
think that projects should ever have to simply take no for an answer.
And the fact that the board didn't provide a response to the top issue
listed in the December board report, and didn't reply to the attempt
to provide an out-of-cycle report last month didn't help.

Despite this clear failure of the board, I would suggest that the
CloudStack team alert the board before taking such an action again.

- Sam Ruby

View raw message