[Fedora-packaging] Proposal to reduce anti-bundling requirements

[Josh Boyer]

On Mon, Sep 14, 2015 at 1:10 PM, Stephen John Smoogen <smooge at gmail.com> wrote:
> On 14 September 2015 at 10:43, Josh Boyer <jwboyer at fedoraproject.org> wrote:
>> On Mon, Sep 14, 2015 at 12:12 PM, Adam Williamson
>> <adamwill at fedoraproject.org> wrote:
>>> On Mon, 2015-09-14 at 10:13 -0400, Stephen Gallagher wrote:
>>>
>>>> We also haven't established any kind of migration plan between the
>>>> two
>>>> repositories. In the darktable example, one of the reasons we ended
>>>> up
>>>> going back to the FPC and re-requesting an exception was that we
>>>> don't
>>>> have any mechanisms for moving an application out of the Fedora
>>>> repositories and into COPR automatically. So upgrades would be
>>>> broken.
>>>
>>> The big thing for me was that we can't use stuff from outside the
>>> Fedora repos to build Fedora products - so we couldn't include
>>> darktable in the design spin if it was in COPR.
>>
>> So you aren't incorrect, but I think that actually probably won't
>> really hold over the long run.  Workstation can already include a set
>> of curated COPR repo files that are disabled by default.  That allows
>> Software to present the contents of the COPR in searches and then
>> allow a user to install from there following a proper warning.  I do
>> not see why the Design Spin could not follow the same steps.
>>
>
> The issue is that most of the Spins want to be able to work as Live
> only images. Mizmo and others take the Design spin to various classes
> as a cdrom or a USB key and then reboot the Windows box to using the
> cdrom. I don't know if darktable is extremely important in these
> environments, but if it is.. then this would make the spin less
> useful. [In the case of people teaching these classes in South America
> or Southern Asia.. the entire ability to have access to Coprs is
> probably not happening.]

Fair point, though I would counter that if one of the goals of these
classes (even if a minor one) is to gain additional Fedora users and
eventually contributors then education spent on use the OS and its
tooling isn't wasted.  To put another way, these are classes using
Fedora to do design work.  Certainly the main focus is going to be on
the tools designers use, but some focus is surely going to be on why
Fedora is the OS to use here instead of any other OS that has said
tools, right?  "Very up-to-date software available in COPRs" might be
one compelling point.

(Again, perhaps not for darktable but some future design tool.)

josh