:: Re: [DNG] Claywand dosplay bananage…
Góra strony
Delete this message
Reply to this message
Autor: Rainer Weikusat
Data:  
Dla: dng\@lists.dyne.org
Nowe tematy: Re: [DNG] X forwarding over SSH over ADSL (Was: Claywand dosplay bananager)
Temat: Re: [DNG] Claywand dosplay bananager
Simon Hobson <linux@???> writes:

[...]

> Rainer Weikusat <rainerweikusat@???> wrote:
>
>>> Is that what I get with ssh -X? I've noticed it's sometimes quite klunky.
>>
>> ssh -X is basically 'straight X' but with the protocol traffic
>> transparently forwarded over the SSH connection and some convenience
>> features like "setting up a suitable DISPLAY" and
>> "handling MIT magic cookie authentication".
>>
>> For this to work well (for applications where there's any hope that it
>> could work well), the remote system needs to have good upstream
>> bandwidth to "the internet" which will usually not be the case if ADSL
>> is being used.
>
> I disagree. I've used remote X forwarding many times, and found it ran
> "quite nicely" with 400kbps upstream from my home ADSL. Obviously it
> depends what you are doing, and "graphics intensive" stuff slows
> enormously, but for anything "text and widgets" based it's like being
> connected locally


Disagreeing with facts is a little pointless. X-over-TCP worked nicely
for me in a LAN. While the computer I was using remotely was connected
via 2MBit leased-line, 'ssh -X' worked ok, the compression delivering a
notable improvement. Going over a 'BT business' DSL-connection required
more aggressive/ targetted compression (using dxpc) for me to be able to
use (non-GTK) Emacs running on the remote machine fluently.

This will obviously vary depending on the bandwidth that's actually
available but the rule-of-thumb is that X will work well over a
network with 1MBit or more up and down.

NB: 'Work well' is supposed to mean that one can't usually tell
which applications are running locally and which remotely (except by
knowing, of course).