On 09/11/2015 12:59 PM, Adam Miller wrote:
On Fri, Sep 11, 2015 at 10:59 AM, Matthew Miller
<mattdm(a)fedoraproject.org> wrote:
> On Thu, Sep 10, 2015 at 01:02:10PM -0400, Matthew Miller wrote:
>> * Vagrant boxes:
>> - same tunir-based test suite in VM environemnt
>
> Followup! Kushal points out that we are testing the KVM vagrant images
> in this way, but not testing VirtualBox. (Because we don't have
> VirtualBox in Fedora or EPEL, because out-of-tree kernel modules.).
> Like the qcow2->ec2 thing, these are the same bits as something that
> _is_ autotested, but run in a different environment. Unlike qcow2->ec2,
> we aren't even doing a boot test.
>
>
> Things which could go wrong which I see are:
>
> * some VirtualBox-specific thing with booting an updated kernel or
> grub2 (for example, updated kernel missing some drivers or something
> that VirtualBox needs)
>
> * some corruption or something in the image conversion
>
> These seem mostly unlikely, but far from impossible.
>
>
> Since VirtualBox is the format the vast majority of Vagrant users will
> want, that's... kind of a big deal. *sigh* Options I can see here are:
>
> A) Scramble to find some way to do the VirtualBox testing.
>
> B) Don't publish the VirtualBox images.
>
> C) Publish the VirtualBox images, but put them in a Penalty Box with
> extra warnings
>
> Any other ideas? Preferences? B seems the most responsible, yet also
> the most sad. A would be highly unusual for our infrastructure. C could
> expose us to looking bad if support breaks and no one notices.
I'm pretty neutral on B or C. I don't really care and also don't think it
should even remotely be a concern of ours. Not only do we not have
testing for it but we don't even have the building blocks in place to
work towards testing it. VirtualBox is bad and those who use it should
feel bad.[0]
Damn, man. That's harsh and probably not a great way to bring people
into the fold.
("should feel bad" I mean. I don't disagree on the merits of VB, it's
not good.)
This is probably not a popular opinion and I'm fine with that,
but we
would have to install something that we very publicly speak out
against in order to test this. I'm not yet ready to throw out Fedora's
values for the sake of some OS X user's convenience but that's just
me.
Which of the values would we be compromising here to test something with
GPL'ed software? I realize that out of band modules aren't awesome, but
I wasn't aware this was a project-level value [1].
The "some OS X user" that we're trying to reach with VirtualBox-friendly
images today is a potential Fedora desktop user tomorrow. I would
totally agree we shouldn't compromise by using proprietary software, but
using fully GPL'ed software to test something... that doesn't seem like
a violation of Fedora values to me. (Note I'm differentiating between
"values" and best practices/packaging guidelines/etc. here.)
[1]
https://fedoraproject.org/wiki/Overview#Our_Core_Values
--
Joe Brockmeier | Community Team, OSAS
jzb(a)redhat.com |
http://community.redhat.com/
Twitter: @jzb |
http://dissociatedpress.net/