[CentOS-devel] Re: CentOS Errata

Kevan Benson kbenson at a-1networks.com
Mon Sep 10 20:02:34 UTC 2007

On Monday 10 September 2007 12:22, R P Herrold wrote:
> On Mon, 10 Sep 2007, Kevan Benson wrote:
> > So, should I take the silence on this topic as no, there
> > isn't any desire for a web-based errata list beyond the
> > searchable archives of the announce list, or just that
> > everyone's too busy to comment?
> Kevan ...
> Produce a mock up, and accessible underlying generating code,
> database schema, and the rest, if you wish it to be reviewed
> for comment. It is easy to propose work for others; harder to
> do the work.

The whole point of this (tangent to this) thread is because I AM offering to 
do it, I just didn't want to devote hours of my time to something that wasn't 
going to be accepted because of some unknown (to me) criteria.

> By way of Requirements it will need to cover all NERVA
> population,


> 3 active releases, base, and all adjunct archives, 
> track CVE's with live links, and emit an RSS, and yet be
> searchible,

All can be handled through a very simple querying of the DB it would be stored 
in, although I'm not sure what you are referring to by adjunct archives.

> capable of subscription self-management, 

?  We're talking about RSS here, right?  I take it you mean differing feeds so 
you don't have to see all errata (E.g. Redhat 5 errata feed, etc.).  That's 
easily handled through the utilization of a DB back-end as well.

> and not  
> add to workflow.

Well, that's a tall order.  Everything adds to workflow unless it replaces 
something else or utilizes existing data automatically.  Karanbir already 
noted that the errata posted to the list are automatically generated, 
otherwise I would suggest generating them out of this as well, which would 
replace some existing workload (possibly) while adding new functionality.

The real question is, what level of added workflow are you prepared to accept?  
At a bare minimum (and easiest  for me), errata would have a name, short 
description, associated set of affected releases, and content (most likely 
being the same content as the mail sent to the announce list).  That's a copy 
and paste on the CentOS dev's end at most, and an automated insertion 
mechanism at least.

I'm serious about providing this, if you guys are serious about using it.

- Kevan Benson
- A-1 Networks

More information about the CentOS-devel mailing list