Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.debian.user > #264104 > unrolled thread

Telnet

Back to article view | Back to linux.debian.user

Contents

  Telnet William Torrez Corea <willitc9888@gmail.com> - 2023-12-02 17:00 +0100
    Re: Telnet Marco Moock <mm@dorfdsl.de> - 2023-12-02 17:10 +0100
      Re: Telnet Greg Wooledge <greg@wooledge.org> - 2023-12-02 17:20 +0100
      Re: Telnet William Torrez Corea <willitc9888@gmail.com> - 2023-12-02 17:40 +0100
    Re: Telnet Andy Smith <andy@strugglers.net> - 2023-12-02 23:00 +0100
      Re: Telnet Michel Verdier <mv524@free.fr> - 2023-12-03 12:10 +0100
        Re: Telnet Marco Moock <mm@dorfdsl.de> - 2023-12-03 17:10 +0100
          Re: Telnet Charles Curley <charlescurley@charlescurley.com> - 2023-12-03 20:00 +0100
            Re: Telnet Greg Wooledge <greg@wooledge.org> - 2023-12-03 20:10 +0100
              Re: Telnet Charles Curley <charlescurley@charlescurley.com> - 2023-12-03 20:30 +0100
                Re: Telnet debian-user@howorth.org.uk - 2023-12-03 21:30 +0100
                  Re: Telnet Anssi Saari <anssi.saari@debian-user.mail.kapsi.fi> - 2023-12-04 11:30 +0100
                    Re: Telnet gene heskett <gheskett@shentel.net> - 2023-12-04 12:40 +0100
                      Re: Telnet debian-user@howorth.org.uk - 2023-12-04 17:20 +0100
                        Re: Telnet Anssi Saari <anssi.saari@debian-user.mail.kapsi.fi> - 2023-12-05 11:20 +0100
              Re: Telnet Tim Woodall <debianuser@woodall.me.uk> - 2023-12-04 11:40 +0100
                Re: Firefox, was: Telnet Michael Kjörling <2695bd53d63c@ewoof.net> - 2023-12-04 13:10 +0100
                  Re: Firefox, was: Telnet Tim Woodall <debianuser@woodall.me.uk> - 2023-12-04 15:30 +0100
            Re: Telnet Nicolas George <george@nsup.org> - 2023-12-04 09:30 +0100
              Re: Telnet Marco Moock <mm@dorfdsl.de> - 2023-12-04 17:30 +0100
                Re: Telnet Nicolas George <george@nsup.org> - 2023-12-04 17:40 +0100
            Re: Telnet Marco Moock <mm@dorfdsl.de> - 2023-12-04 09:30 +0100
              Re: Telnet Curt <curty@free.fr> - 2023-12-04 17:30 +0100
                Re: Telnet Nicolas George <george@nsup.org> - 2023-12-04 17:40 +0100
                  Re: Telnet Curt <curty@free.fr> - 2023-12-04 17:50 +0100
                    Re: Telnet Nicolas George <george@nsup.org> - 2023-12-04 18:00 +0100
                  Re: Telnet <tomas@tuxteam.de> - 2023-12-04 18:40 +0100
                    Re: Telnet Nicolas George <george@nsup.org> - 2023-12-04 20:10 +0100
                      Re: Telnet Nicolas George <george@nsup.org> - 2023-12-04 20:30 +0100
                        Re: Telnet <tomas@tuxteam.de> - 2023-12-05 06:10 +0100
                          Re: Telnet Nicolas George <george@nsup.org> - 2023-12-05 08:50 +0100
                      Re: Telnet tomas@tuxteam.de - 2023-12-04 20:30 +0100
            Re: Telnet Nicolas George <george@nsup.org> - 2023-12-04 09:30 +0100
              Re: Telnet Nicholas Geovanis <nickgeovanis@gmail.com> - 2023-12-04 17:20 +0100
              Look, ma no telnet [was: Telnet] <tomas@tuxteam.de> - 2023-12-04 20:00 +0100
                Re: Look, ma no telnet [was: Telnet] <tomas@tuxteam.de> - 2023-12-04 20:10 +0100
                Re: Look, ma no telnet [was: Telnet] Nicolas George <george@nsup.org> - 2023-12-04 20:10 +0100

Page 2 of 2 — ← Prev page 1 [2]

#264226

Marco Moock (12023-12-04):
> Is that really the case?

Yes.

> Other applications like telnet or vi don't care about it, so I
> assume(d), it is up to the application to handle it.

Applications can decide to change the mode of the tty or catch SIGINT.

Regards,

-- 
  Nicolas George

[toc] | [prev] | [next] | [standalone]

#264193

Am 04.12.2023 um 09:23:16 Uhr schrieb Nicolas George:

> If you want to test a network protocol, you should use a really
> transparent client. Traditionally people use netcat (nc), but it
> handles EOF approximatively.

ncat also uses ^C to kill the process.

[toc] | [prev] | [next] | [standalone]

#264223

On 2023-12-04, Marco Moock <mm@dorfdsl.de> wrote:
> Am 04.12.2023 um 09:23:16 Uhr schrieb Nicolas George:
>
>> If you want to test a network protocol, you should use a really
>> transparent client. Traditionally people use netcat (nc), but it
>> handles EOF approximatively.
>
> ncat also uses ^C to kill the process.
>
>

Telnet doesn't alter the actual data being transmitted, which is the
denotation of transparent in this context, so I think his point is
ludicrously small, if inexistant.

[toc] | [prev] | [next] | [standalone]

#264229

Curt (12023-12-04):
> Telnet doesn't alter the actual data being transmitted

Yes it does, read the doc before posting wrong information here.

-- 
  Nicolas George

[toc] | [prev] | [next] | [standalone]

#264230

On 2023-12-04, Nicolas George <george@nsup.org> wrote:
> Curt (12023-12-04):
>> Telnet doesn't alter the actual data being transmitted
>
> Yes it does, read the doc before posting wrong information here.
>

I think you're buggering yet another fly here.

[toc] | [prev] | [next] | [standalone]

#264233

Curt (12023-12-04):
> I think you're buggering yet another fly here.

I think you should read the docs and shut up. I know what I am saying.

-- 
  Nicolas George

[toc] | [prev] | [next] | [standalone]

#264235

[Multipart message — attachments visible in raw view] — view raw

On Mon, Dec 04, 2023 at 05:32:20PM +0100, Nicolas George wrote:
> Curt (12023-12-04):
> > Telnet doesn't alter the actual data being transmitted
> 
> Yes it does, read the doc before posting wrong information here.

Which, in the case of interaction with HTTP (and most others) actually
comes in handy. Those explicit \r\n get old pretty fast...

;-)

Cheers
-- 
t

[toc] | [prev] | [next] | [standalone]

#264241

[Multipart message — attachments visible in raw view] — view raw

tomas@tuxteam.de (12023-12-04):
> Which, in the case of interaction with HTTP (and most others) actually
> comes in handy. Those explicit \r\n get old pretty fast...

Just hope you will not need to emit a LATIN SMALL LETTER Y WITH
DIAERESIS in ISO-8859-1.

Anyway, the treatment done by telnet is not bad per se, provided we know
(1) that they happen, (2) if we need them in our use case and (3) how to
turn them off. I guess most people who use telnet as a general network
client do not know either (1) nor (2) nor (3).

Regards,

-- 
  Nicolas George

[toc] | [prev] | [next] | [standalone]

#264244

[Multipart message — attachments visible in raw view] — view raw

tomas@tuxteam.de (12023-12-04):
> Back then (TM) (must have been 1990ies or so) I knew. And I sometimes still
> miss the "easy interactivity". I haven't investigated whether there is an
> equivalent socat mode (say line-mode with readline editing or something). That
> would be a market niche, wouldn't it?

I am not sure what you are saying: telnet does not have a line mode with
readline editing; anything of that kind you observe is on the server
side. socat, OTOH, has.

Regards,

-- 
  Nicolas George

[toc] | [prev] | [next] | [standalone]

#264268

[Multipart message — attachments visible in raw view] — view raw

On Mon, Dec 04, 2023 at 08:29:04PM +0100, Nicolas George wrote:
> tomas@tuxteam.de (12023-12-04):
> > Back then (TM) (must have been 1990ies or so) I knew. And I sometimes still
> > miss the "easy interactivity". I haven't investigated whether there is an
> > equivalent socat mode (say line-mode with readline editing or something). That
> > would be a market niche, wouldn't it?
> 
> I am not sure what you are saying: telnet does not have a line mode with
> readline editing;

It does have a line mode with local echo (meaning you can type without anything
being sent until you hit ENTER, with limited line editing capabilities (backspace
and things). No readline's full power, though: that was more like a Christmas's
wish :-)

>              anything of that kind you observe is on the server
> side.

I know Telnet too well, I'm /that/ old (and I just tried it against the local
Web server: yes, it has a line mode and it even is the default). I think we
don't remember how bad network latency could be back then. Line mode was essential.

> socat, OTOH, has.

Now that's good news: Christmas can come -- thanks for the hint :)

Cheers
-- 
t

[toc] | [prev] | [next] | [standalone]

#264274

tomas@tuxteam.de (12023-12-05):
> It does have a line mode with local echo (meaning you can type without anything
> being sent until you hit ENTER, with limited line editing capabilities (backspace
> and things). No readline's full power, though: that was more like a Christmas's
> wish :-)

Oh, you mean that. That is not telnet, that comes from the kernel. Even
sleep has this.

Regards,

-- 
  Nicolas George

[toc] | [prev] | [next] | [standalone]

#264245

[Multipart message — attachments visible in raw view] — view raw

On Mon, Dec 04, 2023 at 08:04:40PM +0100, Nicolas George wrote:
> tomas@tuxteam.de (12023-12-04):
> > Which, in the case of interaction with HTTP (and most others) actually
> > comes in handy. Those explicit \r\n get old pretty fast...
> 
> Just hope you will not need to emit a LATIN SMALL LETTER Y WITH
> DIAERESIS in ISO-8859-1.

Most definitely :-)

> Anyway, the treatment done by telnet is not bad per se, provided we know
> (1) that they happen, (2) if we need them in our use case and (3) how to
> turn them off. I guess most people who use telnet as a general network
> client do not know either (1) nor (2) nor (3).

Back then (TM) (must have been 1990ies or so) I knew. And I sometimes still
miss the "easy interactivity". I haven't investigated whether there is an
equivalent socat mode (say line-mode with readline editing or something). That
would be a market niche, wouldn't it?

Cheers
-- 
t

[toc] | [prev] | [next] | [standalone]

#264194

Charles Curley (12023-12-03):
> True. None the less, there is at least one perfectly good use for
> telnet: testing connections to servers.

Wrong. The telnet client is not entirely transparent, as the telnet
protocol defines an escape octet to introduce commands.

If you want to test a network protocol, you should use a really
transparent client. Traditionally people use netcat (nc), but it handles
EOF approximatively.

For that use, I strongly recommend socat.

Regards,

-- 
  Nicolas George

[toc] | [prev] | [next] | [standalone]

#264221

[Multipart message — attachments visible in raw view] — view raw

On Mon, Dec 4, 2023, 2:23 AM Nicolas George <george@nsup.org> wrote:

> Charles Curley (12023-12-03):
> > True. None the less, there is at least one perfectly good use for
> > telnet: testing connections to servers.
>
> Wrong. The telnet client is not entirely transparent, as the telnet
> protocol defines an escape octet to introduce commands.
>

The show-stopper early-internet issue was that the protocol did not hash,
encrypt or secure the exchange of credentials at session start time.

If you want to test a network protocol, you should use a really
> transparent client. Traditionally people use netcat (nc), but it handles
> EOF approximatively.
>

There's a little chunk of perl in the Camel book that does what you want on
any port you want. In the 2nd edition it's pg. 349-351.


For that use, I strongly recommend socat.
>
> Regards,
>
> --
>   Nicolas George
>
>

[toc] | [prev] | [next] | [standalone]

#264237 — Look, ma no telnet [was: Telnet]

[Multipart message — attachments visible in raw view] — view raw

On Mon, Dec 04, 2023 at 09:23:16AM +0100, Nicolas George wrote:
> Charles Curley (12023-12-03):
> > True. None the less, there is at least one perfectly good use for
> > telnet: testing connections to servers.
> 
> Wrong. The telnet client is not entirely transparent, as the telnet
> protocol defines an escape octet to introduce commands.
> 
> If you want to test a network protocol, you should use a really
> transparent client. Traditionally people use netcat (nc), but it handles
> EOF approximatively.
> 
> For that use, I strongly recommend socat.

These days, even bash can do it:

tomas@trotzki:~$ exec 3<>/dev/tcp/localhost/80
tomas@trotzki:~$ echo -ne "GET / HTTP/1.0\r\n\r\n" >&3
tomas@trotzki:~$ cat <&3
HTTP/1.0 200 OK
Content-Type: text/html
ETag: "3243577940"
Last-Modified: Mon, 30 Oct 2017 17:09:28 GMT
Content-Length: 3378
Connection: close
Date: Mon, 04 Dec 2023 18:55:37 GMT
Server: lighttpd/1.4.59

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head>


These days, even bash can do it:

  tomas@trotzki:~$ exec 3<>/dev/tcp/localhost/80
  tomas@trotzki:~$ echo -ne "GET / HTTP/1.0\r\n\r\n" >&3
  tomas@trotzki:~$ cat <&3
  HTTP/1.0 200 OK
  Content-Type: text/html
  ETag: "3243577940"
  Last-Modified: Mon, 30 Oct 2017 17:09:28 GMT
  Content-Length: 3378
  Connection: close
  Date: Mon, 04 Dec 2023 18:55:37 GMT
  Server: lighttpd/1.4.59

  <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
  <html xmlns="http://www.w3.org/1999/xhtml">
  <head>
  [... rest elided ...]

;-)

(alas, it can't do TCP server, which is a pity)

Cheers
-- 
t

[toc] | [prev] | [next] | [standalone]

#264238 — Re: Look, ma no telnet [was: Telnet]

[Multipart message — attachments visible in raw view] — view raw

On Mon, Dec 04, 2023 at 08:01:10PM +0100, Nicolas George wrote:
> tomas@tuxteam.de (12023-12-04):
> > These days, even bash can do it:
> 
> Zsh could years before.
> 
> > (alas, it can't do TCP server, which is a pity)
> 
> socat can do TCP server. And UDP client, and UDP server, and TLS/SSL
> client and server, and Unix sockets, and SOCKS, and tun, and…

Don't tell /me/: Back then, captive behind a corporate firewall I used
to wrap my outgoing ssh in a tls tunnel with a 443 endpoint "out there",
courtesy of socat (I didn't want to take chances and let the TLS really
look like TLS :-)

> (And zsh seems to be able to do TCP server.)

That one sounds really nice :-)

Cheers
-- 
t

[toc] | [prev] | [next] | [standalone]

#264239 — Re: Look, ma no telnet [was: Telnet]

[Multipart message — attachments visible in raw view] — view raw

tomas@tuxteam.de (12023-12-04):
> These days, even bash can do it:

Zsh could years before.

> (alas, it can't do TCP server, which is a pity)

socat can do TCP server. And UDP client, and UDP server, and TLS/SSL
client and server, and Unix sockets, and SOCKS, and tun, and…

(And zsh seems to be able to do TCP server.)

Regards,

-- 
  Nicolas George

[toc] | [prev] | [standalone]

Page 2 of 2 — ← Prev page 1 [2]

Back to top | Article view | linux.debian.user

csiph-web