kyua and Google Contributor License Agreement
rodrigc at FreeBSD.org
Wed Sep 30 20:36:57 UTC 2015
This link is busted: ttps://github.com/jmmv/kyua/wiki/Contributing
but this seems to have the latest version with instructions for signing the
On Wed, Sep 30, 2015 at 1:20 PM, Craig Rodrigues <rodrigc at freebsd.org>
> The kyua test framework which is used to run all the tests in FreeBSD under
> /usr/tests/ is available on github: https://github.com/jmmv/kyua/
> However, if someone submits a bugfix or enhancement,
> Julio requires that they sign a Google Contributor License Agreement with
> I understand that this is necessary, but it is quite unfortunate.
> I know a number of folks at EMC/Isilon who have walked away from
> fixes to kyua because of the Google CLA.
> I have signed the Google CLA for kyua through an online submission form in
> September 2014,
> but sometime this year, Google changed their system and lost a bunch of
> submitted CLA signatures (!!) including mine. Honestly, I was kind of
> pissed, and wasn't
> motivated to sign the form again.
> This link: https://github.com/jmmv/kyua/wiki/Contributing doesn't seem to
> go anywhere,
> so I can't even figure out the steps to sign the CLA.
> So we are in a state now where good folks who want to contribute
> (EMC/Isilon) lose interest
> because of needing to sign paperwork. We also have instance where random
> on the Internet try to contribute fixes, but then the fixes are in limbo,
> the contributors never bother to sign the Google CLA:
> De facto, what has resulted is that Julio is the sole person who can work
> on kyua and enhance or fix it.
> This is really sad, and a major buzzkill for an open source project. kyua
> is good stuff too, and very helpful for FreeBSD, which makes it even sadder.
> Is there anything we can do to get out of this dead-end?
> -> get Google to remove the need for people to sign the Google CLA?
> -> fork the kyua code, and relicense it so it doesn't need the Google
> -> other?
> Julio has done great work, but it is never good to have a single point of
> failure for
> something as important as this.
More information about the freebsd-testing