[arch-dev-public] Web based signoffs

[Eric Belanger]

On Sat, 16 Aug 2008, Dan McGee wrote:

> On Sat, Aug 16, 2008 at 1:13 PM, Eric Belanger
> <belanger at astro.umontreal.ca> wrote:
>> On Sat, 16 Aug 2008, Dusty Phillips wrote:
>>
>>> Esteemed ArchLinux Developers,
>>>
>>> Partially because Aaron asked for it and thinks it is cool and
>>> partially because all those [signoff] e-mails in my inbox are kind of
>>> boring to me (as a non-packager), I have implemented a web based
>>> signoff mechanism into archweb_dev.
>>>
>>> https://dev.archlinux.org/packages/signoffs/
>>>
>>> This is a thoroughly un-tested beta version that will probably break.
>>> I'm far too lazy/important/busy (depending how you perceive my levels
>>> of ambition/arrogance/available time...) to do my own testing, so
>>> please break it and report bugs.
>>>
>>> It works pretty simple. All packages in testing are listed in the
>>> signoff page. If you test a package and deem it usable you can click
>>> the "signoff" link to sign off for it. If two or more people sign off
>>> for a package, the package 'approved' flag is set to true.
>>
>> The  'approved' flag for a package (both i686 and x86_64 versions) should be
>> set to true when the package has received at least one signoff per arch.
>>
>> Apart from that bug, it looks nice and it would be more convenient than the
>> ML expecially to keep track of the signoffs.
>
> Yes and no. I can't think of an easy way to record the packager as
> having signed off, but it would be nice to know who exactly signed off
> rather than just doing the assumed signoff. For instance, I have built
> x86_64 packages before that I have no possible way of signing off on
> (GUI application or something), and in that case I do want multiple
> signoffs and my implicit one doesn't exist.
>
> However, I can't think of an elegant way to do this. Anyone else?
>
> -Dan

We could just stop assuming that the packager has automatically signed off 
and toggle the 'approved' flags after each arch has received 2 signoffs. 
Then, the packager will need to go to the web page and signoff his own 
packages (if he could test them, of course).

Eric

-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.