Re: Bug#737493: RFS: iceowl-l10n/2.6.4-1 [NMU]
On Mon, Feb 3, 2014 at 4:35 AM, Wookey <wookey@wookware.org> wrote:
> +++ Vincent Cheng [2014-02-02 21:27 -0800]:
>> On Sun, Feb 2, 2014 at 9:20 PM, Jerome Charaoui <jerome@riseup.net> wrote:
>> > -----BEGIN PGP SIGNED MESSAGE-----
>> > Hash: SHA256
>> >
>> > Le 2014-02-03 00:08, Vincent Cheng a écrit :
>> >> Have you tried contacting the current maintainer prior to sending
>> >> out this RFS? If they haven't responded in a timely manner, please
>> >> ping the MIA team and go through the MIA process; if they did reply
>> >> and simply don't have time to update their package, please get them
>> >> to say so on a public list / bug report and include a link to it in
>> >> your RFS bug. Otherwise, this would be considered a hostile NMU.
>> >
>> > As the maintainer is listed in LowThresholdNmu, I thought it would be
>> > okay to upload without delay. And I did leave a note in bug #693150
>> > announcing my intention and asking for feedback. Could the upload still
>> > be considered as such in light of this?
>>
>> No, low threshold NMUs doesn't give an unconditional license to upload
>> new upstream releases and/or make 0-day uploads (it's also never been
>> formalized in Policy, so there aren't exactly any clear-cut rules as
>> to what low threshold NMUs do allow...).
>
> Reading the bug, this upload seems reasonable to me. The original reason
> for not doing it was that versions of various ice* things needed to be
> in sync. Other things are now updated so that reason for delay is gone
> and this package is now uninstallable without an update. The maintainer
> said that help was appreciated. This looks like help to me. Combined
> with LowthresholdNMU, no complaint and a grave bug I really don't think
> we should be telling Jerome that he's doing it wrong.
Ack, in retrospect I may have been too cautious/strict with applying
regular NMU rules.
That being said, my understanding (interpretation?) of LowThresholdNMU
is that it doesn't give NMUers the ability to liberally upload new
upstream releases, and LowThresholdNMU or not, it's _always_ a good
idea to at least attempt to contact the maintainer beforehand (that,
and my willingness to sponsor NMUs in general drops sharply if the
uploader has never tried to get in touch with the maintainer...and I'm
rather wary of sponsoring NMUs to begin with, to be honest).
It'd help if LowThresholdNMU could be formalized in Policy, or at
least mentioned in devref.
Regards,
Vincent
Reply to: