Networking

NAVIGATION
CATEGORIES
REFERRENCE
LINKS
  • rfc3920 document: use of TLS

    5 answers - 365 bytes - related search similar search Add To My Delicious Add To My Stumble Upon Add To My Google Mark Add To My Facebook Add To My Digg Add To My Reddit

    Hi,
    In the rfc3920 document (XMPP: Core), section 5.2 Narrative, step 5
    states that: "The receiving entity MUST reply with either a <proceed/>
    element or a <failure/element"
    My question is upon what conditions the receiving entity replies with
    a <proceed/and upon what conditions it replies with a <failure/>?
    Thanks,
  • No.1 | | 927 bytes | |

    Chen, Hao wrote:

    In the rfc3920 document (XMPP: Core), section 5.2 Narrative, step 5
    states that: "The receiving entity MUST reply with either a <proceed/>
    element or a <failure/element"

    My question is upon what conditions the receiving entity replies with
    a <proceed/and upon what conditions it replies with a <failure/>?

    Hmm, that could be better specified, eh? We'll have to fix that in
    rfc3920bis.

    I can see two reasons for returning a <failure/>:

    1. The server is temporarily not prepared to offer TLS negotiation (some
    internal server problem).

    2. The STARTTLS command is malformed (i.e., something other than
    <starttls xmlns=''/because the
    namespace is wrong, there is XML character data contained in the
    <starttls/element, or whatever).

    the server would return <proceed/>, I think.

    Peter
  • No.2 | | 1209 bytes | |

    how about the server deployment not being enabled with TLS? Would that
    also result in failure (if the client ignored the server capabilities
    because it is configured to require TLS)?
    -David Waite

    7/22/05, Peter Saint-Andre <stpeter (AT) jabber (DOT) orgwrote:
    Chen, Hao wrote:

    In the rfc3920 document (XMPP: Core), section 5.2 Narrative, step 5
    states that: "The receiving entity MUST reply with either a <proceed/>
    element or a <failure/element"

    My question is upon what conditions the receiving entity replies with
    a <proceed/and upon what conditions it replies with a <failure/>?

    Hmm, that could be better specified, eh? We'll have to fix that in
    rfc3920bis.

    I can see two reasons for returning a <failure/>:

    1. The server is temporarily not prepared to offer TLS negotiation (some
    internal server problem).

    2. The STARTTLS command is malformed (i.e., something other than
    <starttls xmlns=''/because the
    namespace is wrong, there is XML character data contained in the
    <starttls/element, or whatever).

    the server would return <proceed/>, I think.

    Peter
  • No.3 | | 2591 bytes | |

    If the server does not enable TLS, then it should not even broadcast
    the <starttlsfeature namespace within the <stream:features
    element. Thus, in your client, you should double check first if the
    feature exists before proceeding with the TLS negotiation.

    However, if you D proceed with TLS and the server does not support
    it, then I would think it's reasonable for the server to return a
    <failureand disconnect the connection.

    Thanks,
    Chris

    Jul 22, 2005, at 12:02 PM, David Waite wrote:

    how about the server deployment not being enabled with TLS? Would that
    also result in failure (if the client ignored the server capabilities
    because it is configured to require TLS)?

    -David Waite

    7/22/05, Peter Saint-Andre <stpeter (AT) jabber (DOT) orgwrote:
    >
    >Chen, Hao wrote:
    >>
    >>

    In the rfc3920 document (XMPP: Core), section 5.2 Narrative, step 5
    states that: "The receiving entity MUST reply with either a
    <proceed/>
    element or a <failure/element"

    My question is upon what conditions the receiving entity replies
    with
    a <proceed/and upon what conditions it replies with a <failure/>?

    >>

    >Hmm, that could be better specified, eh? We'll have to fix that in
    >rfc3920bis.
    >>

    >I can see two reasons for returning a <failure/>:
    >>

    >1. The server is temporarily not prepared to offer TLS negotiation
    >(some
    >internal server problem).
    >>

    >2. The STARTTLS command is malformed (i.e., something other than
    ><starttls xmlns=''/because the
    >namespace is wrong, there is XML character data contained in the
    ><starttls/element, or whatever).
    >>

    >the server would return <proceed/>, I think.
    >>

    >Peter
    >>

    >--
    >Peter Saint-Andre
    >Jabber Software Foundation
    >
    >>
    >>

    >
    >jdev mailing list
    >jdev (AT) jabber (DOT) org
    >
    >>
    >>
    >>
    >>


    jdev mailing list
    jdev (AT) jabber (DOT) org

    jdev mailing list
    jdev (AT) jabber (DOT) org
  • No.4 | | 769 bytes | |

    Fri, Jul 22, 2005 at 09:58:11AM -0600, Peter Saint-Andre wrote:
    Hmm, that could be better specified, eh? We'll have to fix that in
    rfc3920bis.

    I can see two reasons for returning a <failure/>:

    []

    2. The STARTTLS command is malformed (i.e., something other than
    <starttls xmlns=''/because the
    namespace is wrong,

    If the namespace is wrong, then that is not a StartTLS request and
    server should not treat it this way. It should not even be treated as
    a malformed StartTLS request, as if the namespace is not know, then
    there is no way to say what the element means (element name doesn't
    matter).

    Greets,
    Jacek

    jdev mailing list
    jdev (AT) jabber (DOT) org
  • No.5 | | 836 bytes | |

    Jacek Konieczny wrote:
    Fri, Jul 22, 2005 at 09:58:11AM -0600, Peter Saint-Andre wrote:

    >>2. The STARTTLS command is malformed (i.e., something other than
    >><starttls xmlns=''/because the
    >>namespace is wrong,


    If the namespace is wrong, then that is not a StartTLS request and
    server should not treat it this way. It should not even be treated as
    a malformed StartTLS request, as if the namespace is not know, then
    there is no way to say what the element means (element name doesn't
    matter).

    Sure, that was a bad example. But it *is* possible for the STARTTLS
    command to be malformed -- e.g., the initiating entity sends XML
    character data or child elements in the <starttls/element.

    Peter

Re: rfc3920 document: use of TLS


max 4000 letters.
Your nickname that display:
In order to stop the spam: 8 + 7 =
QUESTION ON "Networking"

EMSDN.COM