IETF-SSH archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: clarifications for SFTP extensions draft



denis bider wrote:
> On Tue, 6 Dec 2005 21:28:41 +0100, denis bider wrote:
>> For example, it is not explicitly stated what kind of an
>> extension "vendor-id" is. Is this extension included in the
>> server's VERSION packet? The specification doesn't say so;
>> one might even be lead to think that it's a response to an
>> SSH_FXP_EXTEND packet that must be sent by the client.

It is both.  The server sends it in it's version packet,
the client sends it as a SSH_FXP_EXTENDED packet.

> Also, the spec says that the client MAY send the "vendor-id"
> as an SSH_FXP_EXTEND request to the server. The specification
> should specify the exact encoding of the request in this case,
> and should also determine how the server needs to respond to
> such a request. Does it send SSH_FXP_STATUS with SSH_FX_OK?
> Should it respond with SSH_FXP_EXTENDED_REPLY, perhaps
> re-sending its own vendor ID? Or does it simply ignore it?
> 
> These questions aren't answered by the current draft.

When it is sent as a SSH_FXP_EXTENDED packet,
the contents vendor-structure becomes the extension
specific data:

       byte   SSH_FXP_EXTENDED
       uint32 request-id
       string "vendor-id"
       string vendor-name     [UTF-8]
       string product-name    [UTF-8]
       string product-version [UTF-8]
       uint64 product-build-number

The server responds with a status packet, which
should always be SSH_FX_OK.

I will clarify this.

Thanks,

Joseph



Home | Main Index | Thread Index | Old Index