Re: Should maintainers receive copies of their own BTS mails?
* Don Armstrong (22-07-2007):
> On Sun, 22 Jul 2007, Oleg Verych wrote:
>> * Don Armstrong
>> * Date: Sun, 22 Jul 2007 07:29:25 -0700
>> > If you're interested in the discussion around this bug, you should
>> > subscribe to it.
>>
>> 1. There's a new feature -- subscribing.
>
> It's actually not that new; Joachim and Pascal did the l.d.o side and
> I did the b.d.o side over two years ago now.
Well, i meant -- you are pointing on this feature, that's why i'm
concluding, that somebody (still) don't know about it.
>> pros: while done, it's easy to follow bug's messages
>> cons:
>> - handled by another party: l.d.o; unneeded complexity
>> - subscribing have 2 stages with total 2/2 messages sent from
>> both sides
>> - unsubscribing -- the same
>
> This is a requirement for any subscription system, even if a message
> to a bug automatically causes you to be subscribed, as From is trivial
> to forge.
But what if one isn't bound to the "subscription system"?
> I am interested in perhaps allowing people to confirm a subscription
> once, and then be able to subscribe to any bug using that same address
> without further confirmation (or perhaps with a GPG signature), but to
> date the code has not been written.
This isn't flexible (storage/dynamic management of GPG, etc).
>> 2. Thus, here are two sides
>> - BTS maintainer
>> - package maintainer
>> and they have different experience and goals, while trying to
>> understand each other.
>
> In this particular instance, I'm sitting in both positions, because
> this bug is against debbugs, and I am both a BTS maintainer as well as
> a maintainer of debbugs.
I assume, that something has higher priority here. Your way of
replying suggests the first choice.
>> > I'm open to adding a header to messages so that you can easily
>> > indicate your desire to be subscribed to a bug, and I probably
>> > will do that as soon as it's possible to get the subcription
>> > information off of l.d.o.
>>
>> I think BTS's `reply-to' rules must not be implied, but based on
>> some policy. That policy can be delivered from either `Cc',
>> `Mail-Followup-To' or both.
>
> The reply-to: rules are actually kind of silly, and probably will be
> junked once its easier to subscribe to bugs without confirmation.
> Plus, it's always appropriate for people to cull the Cc: list.
Please, tell me what do you think about scheme, i've described in message
to Josseline. I just need a wise opinion, because i will implement it
anyway.
Thanks.
____
Reply to: