IETF-SSH archive

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

Re: kdbinteract-03



On 15 Feb, Frank Cusack wrote:
>   - [3.1] Language tag deprecated in SSH_MSG_USERAUTH_REQUEST.
> 
>     When originally written, ssh had no provision for language choice
>     (and didn't need it -- all protocol exchanges were byte codes).
>     However, the better place for this is in the transport, which
>     is where it is now.  So I want to deprecate use of it in kbdint.
>     The doc says it may be removed in the future, but I don't think
>     this is easily done if compatibility with existing implementations
>     is desired.  Can someone comment on that?

I do not think it is worth changing this from the current draft since
this is already widely deployed.

>   - [3.2] Server no longer should limit name/prompt fields.
> 
>     Instead, the server should just consider that the client may
>     truncate these fields, and choose its prompts appropriately, if
>     possible.

This is  a difference in degree only, but I can live with it.

>   - [3.3] Client now MUST prompt the user "as follows".
> 
>     The end of 3.2 said the client MUST display the name and
>     instruction as indicated in 3.3, but in 3.3 it only said SHOULD. 
>     This is a disconnect which I didn't like.  So the 3.3 section is
>     now MUST, but the description is a little more flexible.

I would prefer to have SHOULD at both places as to not limit the client
more than absolutely needed. Somebody might want to build a client which
recognizes some prompts from the server and which knows how to answer,
the MUST forbids that.

	/MaF




Home | Main Index | Thread Index | Old Index