[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Tue, 5 Apr 2011 10:56:15 -0600
From: Vincent Danen <vdanen@...hat.com>
To: oss-security@...ts.openwall.com
Subject: Re: Closed list
* [2011-04-05 18:02:54 +0200] Andrea Barisani wrote:
>On Tue, Apr 05, 2011 at 09:40:13AM -0600, Vincent Danen wrote:
>> * [2011-04-05 08:43:29 -0400] Josh Bressers wrote:
>>
>> >
>> >----- Original Message -----
>> >>On Tue, Apr 05, 2011 at 07:19:08AM -0400, Josh Bressers wrote:
>> >>> Not adding Apple to any coordination list would be plain silly. They
>> >>> were far more active than most of the distributions.
>> >>
>> >>Yes. But why do they need to be aware, say, of glibc vulnerabilities
>> >>(ones that are in fact believed to be glibc-specific)?
>> >
>> >This is an excellent point. It's a hard problem to solve honestly. I guess
>> >the question really comes down to this. Do the disadvantages of one list
>> >outweigh the benefits? I'm not sure what the answer is. There probably
>> >isn't an "answer" though, just lots of opinions.
>>
>> Just throwing this out there (I've read the thread but haven't
>> contributed at all yet).
>>
>> A lot of userland stuff is shared between BSD and Linux, and probably
>> some other operating systems. About the only things that differ between
>> a lot of these are the Linux kernel, and the *libc. There is a lot of
>> cross-over with other stuff, which means there will likely be a lot of
>> cc'ing going on (which I imagine might be complicated due to encryption
>> requirements).
>>
>> Where does the line get drawn? If vendor A ships with exim, and another
>> with postfix, which one belongs on the "Linux list"? Obviously
>> discussions of exim don't matter to the postfix shipper, and vice versa.
>> Pick any other software that has a competing open source alternative.
>> Does Debian not get on the list because they don't technically ship
>> firefox?
>>
>> I think if the disqualifier to Apple is that they don't ship a Linux
>> kernel and glibc, then we're doing them (and ourselves) a disservice.
>> Apple contributed a lot to vendor-sec (and I'm not going all pro-Apple
>> here, just stating a fact).
>>
>> I think it would be reasonable to s/Linux list/open source vendor list/,
>> like vendor-sec used to be. Yes, Apple will see some glibc stuff and
>> some Linux kernel stuff. So what? They may also see some exim stuff
>> that doesn't apply to them, and (hopefully!) the Linux vendors may seem
>> some stuff that isn't applicable to them, but it is useful for the *BSD
>> vendors who would (hopefully!) be on the list and feel welcome enough to
>> use it. Who knows, it might even be beneficial to have a glibc issue
>> and someone from Apple or FreeBSD or whatever pipes up and indicates
>> that the BSD libc once had a similar problem and tells us how they fixed
>> it.
>>
>> I think getting hung up on "Linux vendors only" and "BSD vendors can
>> have their own list" and we end up cross-posting 90% of the issues is
>> going to be an exercise in frustration.
>>
>> Either that, or we start to work more closely with a *CERT and deal with
>> their process for passing along information to other vendors for
>> userland things that are shared; no offence to oCERT or anyone else, but
>> that seems like more of a headache than just letting
>> Apple/FreeBSD/OpenBSD/etc. have a seat at our table.
>>
>> Just my $0.02.
>>
>
>None taken :)
>
>Some random reasons about why we value coordination over a "catch all" list
>(which was considered at oCERT beginning):
>
>- some vendors/projects got annoyed by reports not relevant to them, in the
> long run it tends to lower the "attention level" when some matters are
> really meaningful to them, that's why having a trusted purposed channel
> often worked
>
>- it is not feasible to have every single OSS project on such a list and at
> some point there is the need to address individual maintainers in a timely
> fashion along with the affected parties, using a list + cc for that often
> doesn't work as the communication level greatly differs most of the times
> (unfortunately).
>
> We found out the hard way that the usual level of technicality that was
> happening with most vendors or lists like vendor-sec was perceived as
> "threatening" or absolutely obscure to some developers/maintainers.
>
>I am not suggesting that coordination ala oCERT is the only true way of
>course, and I am not trying to pitch our project here, just wanted to give
>some elements for the discussion.
Thanks for that Andrea. I get where you're coming from, but also keep
in mind that the *CERTs also deal with upstream and issues that cover a
multitude of potential upstreams (think protocol-level issues). For
that, absolutely the *CERTs have their place.
I don't think anyone is suggesting this list is a replacement for the
work that the *CERTs can do in coordinating with various upstreams and
vendors. I think rather they can often be complimentary.
>I personally think that we need a blend of both approaches in the long term,
>it is just a matter of using the right tool for the job.
Absolutely agree with this.
--
Vincent Danen / Red Hat Security Response Team
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.
