Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Sun, 15 Jun 2014 16:55:31 +0100
From: Richard Moore <>
Subject: Re: CVE Request for KIO/kmail

On 15 June 2014 16:17, <> wrote:

> Hash: SHA1
> [ actual security content, for people who don't want to read the
> meta-discussion:
>   Is this related to an
>   vulnerability?
> ]
That is unrelated.

> > From: Henri Salo <>
> > More details are needed for public CVE request.
> It might be worthwhile to clarify what problems we're solving, in the
> specific case of a request by a product's upstream vendor, by
> requiring that more vulnerability details be sent here at the time of
> the CVE request.
> MITRE is almost certainly willing to accept this as a valid reason, if
> there's general agreement:
>   - people here don't want to see announcements of vulnerabilities
>     that are both vague AND unfixed, and would just unsubscribe if there
>     were many of these
>     (vague vulnerability details for fixed issues typically haven't
>      had objections in the past, e.g., the
> post)
> (In 2013 and earlier, vulnerability details here had a de-duplication
> purpose for MITRE, but they often don't now. If we were to assign a
> CVE ID for this, we'd already know whether any kio CVE requests had
> been sent directly to us. The CVE-OpenSource-Request-HOWTO.html page
> and examples aren't directly applicable to us, and we're generally
> more concerned with knowing whether something is a unique issue that
> everyone will want to fix, than with knowing every possible detail.
> But we realize the list charter may imply a minimum level of details.)
> Going back to the security content, KDE Bug 330795 mentioned above says:
>   Product  kio
>   KDE's KIO completely ignores the first part, "must not automatically
>   redirect without prompting the user"
> and possibly this is (or was) a violation of RFC 7231 section 6.4,
> with security implications:
>   "Automatic redirection needs to done with care for methods not
>    known to be safe, as defined in Section 4.2.1, since the user
>    might not wish to redirect an unsafe request."
> Does anyone want a CVE ID for that?
> Finally, for the original "a vulnerability in KIO that causes a security
> issue in kmail" request: to get a CVE ID for this more quickly, please send
> to mentioning whether it is or isn't Bug 330795.
> We don't want to send a CVE ID here immediately because MITRE doesn't
> determine the list charter and it might be an off-topic request.
In the past when I've tried to use the cve-assign address it has basically
been a black hole. Since then I've either asked redhat or one of the other
OSS vendors for a CVE. I've used the now as a

I'd also note as part of the meta discussion that I'm not going to release
details of vulnerabilities to a public list  before the fix, and just
because someone asks for more details doesn't mean I will provide them. - I
have no way to know who the person asking is. If getting a CVE assigned
becomes a problem then the solution from my side will be to release the
advisory then let someone else sort out assigning CVEs afterwards.



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.