[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Wed, 6 Apr 2011 22:16:17 -0400
From: Michael Gilbert <michael.s.gilbert@...il.com>
To: oss-security@...ts.openwall.com
Subject: Re: Closed list
Solar Designer wrote:
> On Tue, Apr 05, 2011 at 09:52:10AM +0100, Benji wrote:
> > Fixing issues secretly is definitely a no-go in my book.
>
> I think you're mixing up distinct things:
>
> 1. Fixing security bugs secretly, then releasing the fixed software
> without notifying others of the fixes.
>
> 2. Fixing security bugs secretly, then releasing the fixed software
> along with information on the fixes on the coordinated release date.
>
> I think #1 has worse drawbacks than #2. I think that with the current
> state of the community/industry/technology, we should avoid #1, but we
> can do #2.
>
> Is your opinion on #2 different, and why?
>
> > It will and clearly
> > has, created hostility between different developer groups and those that are
> > allowed in and those that aren't.
>
> Unfortunately, yes, both #1 and #2 may create hostility.
>
> > >>However, my proposal, which I am going to try to enforce, is to only
> > >>discuss medium-severity issues on this new list. I think that an
> > >>embargo period of 1-2 days does not make sense for those; if that's all
> > >>we can afford, we can as well make them public right away.
> >
> > So.... if this list isnt for high-severity issues what is the point of it?
> > Why not use OSS-Sec.
>
> For low-severity issues, I propose that we use oss-security right away.
>
> I propose that we use the new closed list(s) for medium-severity issues,
> where immediate disclosure on oss-security could do some harm.
>
> In this context, I propose to use overall severity defined as the
> product of risk probability and risk impact. Of course, we'll use
> guesstimates.
>
> > I thought the only way this el8 mailing list was even
> > justified was the fact that the vulnerabilities were mission-critical and
> > the POCs for these vulnerabilities would potentially lead to throwing us
> > back into the ice-ages.
>
> That's not my justification. In those special cases, I'd try to see who
> is affected before sending out the detail. However, the list may in
> fact be useful to probe for affected vendors/distros - post a heads
> up, with no detail on the issue, and ask to contact the reporter for
> detail. Also, propose a much shorter embargo period (than is usual for
> the list). vendor-sec was used like that on some occasions, and I think
> it was an improvement over mailing the same heads up to an arbitrary
> subset of distros, which happens in the absence of such a list.
>
> > >>That said, I agree that a closed list should be a last resort, to be
> > >>used whenever other options are determined to be less appropriate for a
> > >>particular security issue. Unfortunately, this determination is usually
> > >>made by just one person (whoever brings the issue to the list), so it is
> > >>likely to sometimes be "wrong".
> >
> > So why are you using a last resort for 'medium-severity issues'?
>
> The key words above were: "whenever other options are determined to be
> less appropriate". "Less appropriate" does not mean that it would be
> the end of the world if the issue were disclosed publicly right away.
> Things would just be worse, in the reporter's opinion. So we provide a
> convenient way for one distro to share info (or just a heads up) with
> other likely-affected distros. In the absence of such a list, the
> reporter would likely end up notifying an arbitrary subset of the distros.
>
> > Currently, from what you've said, it seems like you're trying to, as some
> > people apparently correctly feared, an elite mailing list where you can all
> > boost your egos and, excuse the term for lack of a better one, 'circlejerk'.
>
> I fail to see what in this discussion thread makes you arrive at that
> conclusion, other than presumably you readily having this opinion of any
> closed discussion groups. If that's not the case, then can you name a
> closed discussion group that you would not categorize that way, and
> explain why not? This might help me and others understand you better.
>
> > Question; now that vendor-sec has been compromised, I suppose we can expect
> > a full public archive of all the emails?
>
> Maybe, or maybe not. This may happen if someone just goes ahead and
> posts it publicly. Other than that, making it public in an ethical
> fashion feels unrealistic (we'd need to ask everyone who has ever posted
> to the list).
This is something to consider with the new list as well. Better to get
approval from all participants now. No sense in taking on people who
are just going to "gum up the works" (sorry for the lame cliche, but I
couldn't think of a more succinct way to put it).
Also, completely unrelated to this particular thread, will you be adding
participants/keys to your wiki so people know who they can contact
securely if they do want to responsibly submit an issue? A use case
may be that a researcher doesn't want to disclose an issue to the
entire list, but instead to a limited group that they feel is more
trusted. Such a feature would also alleviate the inadvertent disclosure
concern so you could add more participants to the overall list, since
it seems to be exploding.
Best wishes,
Mike
Powered by blists - more mailing lists
Please check out the
Open Source Software Security Wiki, which is counterpart to this
mailing list.
Confused about mailing lists and their use?
Read about mailing lists on Wikipedia
and check out these
guidelines on proper formatting of your messages.
