questions about "Requesting a Pseudo-Terminal"
- From: yawnmoth <terra1024@xxxxxxxxx>
- Date: Sun, 12 Apr 2009 20:28:06 -0700 (PDT)
There are two sections of the SSH specs I have questions about. The
first is Section 6.2: Request a Pseudo-Terminal:
http://tools.ietf.org/html/rfc4254#section-6.2
The above link uses vt100 as an example TERM environment variable
value for SSH_MSG_CHANNEL_REQUEST. PuTTY uses xterm. What advantages
does xterm offer over vt100?
Also, according to the above link, "the dimension parameters are *only
informational" (emphasis mine). In that case, why have them?
I did "cat test.txt" on a file which had the letter "a" repeated 100
times. The terminal was an 80x24 one but instead of adding a "\n"
after the 79th "a" nothing happened. This suggests that the above
link is correct when it says "the dimension parameters are *only*
informational".
Finally, I have another question about Section 6.5: Starting a Shell
or a Command:
http://tools.ietf.org/html/rfc4254#section-6.5
Per that link, "only one of these requests can succeed per channel".
I can understand why "shell" should only have one
SSH_MSG_CHANNEL_REQUEST per channel but what about "exec"? Why not
allow additional "exec" channel requests after SSH_MSG_CHANNEL_EOF is
received? As is, there's really no point in doing something like
"cd .." via "exec" when it's just going to be undone when the channel
is closed.
.
- Prev by Date: Re: putty 0.58
- Next by Date: ssh private and public keys known_host etc...
- Previous by thread: ssh hangs when executing a command
- Next by thread: ssh private and public keys known_host etc...
- Index(es):
Relevant Pages
|
