summaryrefslogtreecommitdiff
path: root/networking/telnet.c
diff options
context:
space:
mode:
Diffstat (limited to 'networking/telnet.c')
-rw-r--r--networking/telnet.c22
1 files changed, 5 insertions, 17 deletions
diff --git a/networking/telnet.c b/networking/telnet.c
index e8e51dc..7081a6a 100644
--- a/networking/telnet.c
+++ b/networking/telnet.c
@@ -186,22 +186,6 @@ static void con_escape(void)
static void handle_net_output(int len)
{
- /* here we could do smart tricks how to handle 0xFF:s in output
- * stream like writing twice every sequence of FF:s (thus doing
- * many write()s. But I think interactive telnet application does
- * not need to be 100% 8-bit clean, so changing every 0xff:s to
- * 0x7f:s
- *
- * 2002-mar-21, Przemyslaw Czerpak (druzus@polbox.com)
- * I don't agree.
- * first - I cannot use programs like sz/rz
- * second - the 0x0D is sent as one character and if the next
- * char is 0x0A then it's eaten by a server side.
- * third - why do you have to make 'many write()s'?
- * I don't understand.
- * So I implemented it. It's really useful for me. I hope that
- * other people will find it interesting too.
- */
byte outbuf[2 * DATABUFSIZE];
byte *p = (byte*)G.buf;
int j = 0;
@@ -216,7 +200,11 @@ static void handle_net_output(int len)
if (c == IAC)
outbuf[j++] = c; /* IAC -> IAC IAC */
else if (c == '\r')
- outbuf[j++] = '\0'; /* CR -> CR NUL */
+ /* See RFC 1123 3.3.1 Telnet End-of-Line Convention.
+ * Using CR LF instead of other allowed possibilities
+ * like CR NUL - easier to talk to HTTP/SMTP servers.
+ */
+ outbuf[j++] = '\n'; /* CR -> CR LF */
}
if (j > 0)
full_write(netfd, outbuf, j);