Proposal - "Slow updates" repo

[Casey Dahlin]

Jeff Spaleta wrote:
> On Tue, Nov 18, 2008 at 10:02 AM, Casey Dahlin <cdahlin at redhat.com> wrote:
>   
>> This is the one drawback to this strategy. Here a human judgement call would
>> have to be made: Take the package and all its deps or run the risk. The
>> right answer would depend on the case, and on what the stable group decides
>> their use case is best served by.
>>     
>
> Security updates break the model for testing to. This is a known problem space.
>
> If you take your idea...but also make it so client side users can
> specifically know what is tagged as security and what is
> not...regardless of repo "speed" then you probably have something
> robust enough for everyone.
>
> Give client side the option to pull ALL security updates...from any
> repo "speed." That way the "slow" update repo doesn't have to be
> burdened with dealing with security updates as part of their mission
> as a special case.
>
> -jef
>
>   
Not a bad idea.

--CJD