EPEL (was Re: RFC: Proposal for a more agile "Fedora.next" (draft of my Flock talk))

[Toshio Kuratomi]

On Mon, Jul 22, 2013 at 12:33 PM, "Jóhann B. Guðmundsson"
<johannbg at gmail.com> wrote:
> On 07/22/2013 05:53 PM, Toshio Kuratomi wrote:
>>
>> On Mon, Jul 22, 2013 at 04:47:12PM +0000, "Jóhann B. Guðmundsson" wrote:
>>>
>>> On 07/22/2013 04:34 PM, Toshio Kuratomi wrote:
>>>>
>>>> This was actually not the rationale.  The rationale was that it wasn't
>>>> harmful to Fedora and so if individual maintainers felt that it was
>>>> something that they wanted to ship they could.
>>>
>>> Did the FPC even bother to test what they had approved in practices?
>>> ( Do they in general? )
>>>
>>> Have you tried installing an legacy sysv initscript file and using
>>> after it has been replaced with a native systemd unit.
>>>
>> The sysv initscripts subpackages are not for use with systemd.  They are
>> for
>> use at places that use systemv init scripts.
>
>
> So is it safe for me to assume what you are getting at here to use with
> other init system than systemd right?
>
> If that is the case look at those what <30 components ship legacy sysv
> initscript along side their native systemd unit counterpart and tell me if
> you can deliver bootable Fedora using either sysv or upstart from those
> components.
>
> Needless to say that reasoning/arguments does not hold water..
>
If by does not hold water, you mean fails to convince you, then that's
fine.  If by fails to hold water you mean, does not have any validity
then I'm afraid you're wrong.  As I said, we saw no reason to ban the
scripts as they are not causing harm.  If maintainers want to ship
them because they perceive them to be useful to people who are not
running systemd then we did not want to stand in their way.  If you
want to argue about whether they are useful or not, take that portion
up with the maintainers of the individual packages.  If you have a
case of harm done by having the subpackages available, then that could
go to the FPC.

>
>>   When the guidelines were
>> initially written, having both installed was tested and they didn't cause
>> problems for systemd (although having both installed and still using
>> systemd
>> was seen as not the use case for having the subpackages).  Has something
>> changed in the systemd code since then that changes that equation?
>>
>
> No and this has nothing to do with systemd.
>
> Again try to use what you approved and are so reluctant to remove and or
> change.
>
> Install a sub package containing a legacy sysv initscript and try to replace
> that legacy sysv initscript with the unit file it has been replaced with and
> remember perform update/upgrades on the components at the same time, then
> tell me if you still think we should be shipping legacy sysv initscripts
> once they have been migrated instead of dropping them to avoid confusion and
> crappy user experience for everybody that try to think they can still (
> easily ) use it.
>
> And there is no need for you to reply to this response on this thread until
> you have done the above because you will not understand what I'm getting at
> until you try this for yourself.
>
Well, I did what you asked and I don't know what you are getting at.
So I suppose either your instructions were unclear or you just wanted
me to see that the FPC subpackage guidelines work as designed.

So, thanks!
-Toshio