Re: Can an SSH_MSG_EXT_INFO have zero entries?

Mouse <[email protected]> Tue, 4 Apr 2023 11:35:18 -0400 (EDT)
Newsgroups gmane.ietf.secsh
Message-ID <[email protected]>
>> For what it's worth, empty lists are valid (and common) in some
>> other places, e.g., the languages lists in SSH_MSG_KEXINIT.

> Right, but that's just some pointless baggage tacked onto a message
> whose actual job it is to communicate crypto parameters.  In the case
> of the extensions, the entire message contains nothing.

(a) Even if true, so what?  Why does that make it wrong?  Why would a
pointless message be any worse than a pointless appendage to a pointful
message?

(b) Not quite; the message contains a definite assertion that none of
the extensions which might have been listed are supported on that
connection (or not supported _yet_, in the case of a server sending at
its first opportunity).  That is, it's evidence of absence, whereas not
sending SSH_MSG_EXT_INFO at all would be absence of evidence.

/~\ The ASCII				  Mouse
\ / Ribbon Campaign
 X  Against HTML		[email protected]
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B