[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-devel
Subject:    Re: [kde-promo] (fwd) NX Project Announcement
From:       "Kurt Pfeifle" <kpfeifle () danka ! de>
Date:       2003-03-31 10:52:50
[Download RAW message or body]

Martin Konold wrote:

Unusual times are dawning -- this must be about the longest email by you,
Martin, that I have ever read"   ;-)

> Am Samstag, 29. M?rz 2003 01:51 schrieb Kurt Pfeifle:
> 
> Hi Dirk,
> Hi Kurt,
> 
> 
>>Dirk Mueller wrote on kde-promo:
> 
> 
>>>might be interesting.
>>
>>Definitely it *IS* interesting
> 
> 
> I very much agree with Kurt here. After the NX stuff was announced on the 
> xfree forum ml I gave it a try. These guys really make it right! They 
> definetly know what they are doing. Especially I checked their low latency 
> stuff and the persistent file based cache.
> 
> 
>>I am testdriving NX since a few days and this is by *far* the fastest
>>"remote destop-ing" system I have ever seen. The performance across
>>an internet link is really astonishing. It is better than what I have
>>even on my LAN (or via a crosslink cable between desktop and laptop)
>>using krdc/krfb!
> 
> 
> Kurt: You are comparing apples and oranges.

Both are fruit.

> Instead of comparing NX with vnc 
> you should compare it to pure X11.

I've done that too (but forgot to mention) and NX is far faster than pure X.

> X11 already knows about all the remote 
> desktop features of NX. NX basically modifies X11 in a way LBX always tried 
> but never succeeded.
> 
> Still NX is an extreme improvement to XFree86.
> 
> 
>>[Be
>>warned: you will be encountering a semi-broken RedHat desktop there -- but
>>I think Gian Filippos offer to help develop an Open Source version of NX is
>>worth while exploring.]
>>
>>A KDE version of NX could be a reall killer application for its usage
>>on the desktop in large offices and organizations.
> 
> 
> The NX Client already works nicely with KDE.

Yes -- but NX is planned to be a *sold*, commercial product. (Dunno about
the pure *client*, but at least the part called "NX server" is.

OTOH, NoMachine have publically offered to support the development of a free
(as in beer, and as in freedom) KDE (and GNOME) version of a GUI frontend.
The core NX libraries are already GPL (and may be used on the commandline
to do what NX frontends do in a more intuitive way.

> Actually when playing an mp3 
> remotely the client already proxies the audio stream to my local artsd.

Ah -- you even got that to work for you. I didn't succeed.

> Even 
> klipper works as expected. Of course native KDE config dialogs etc. would be 
> nice. Maybe even some D&D could lead to interesting usage patterns.
> 
> 
>>NX seems to be able provide (clustered) application servers, which can
>>serve and satisfy hundreds or even thousands of client machines. It will
>>open the door to a new kind of "network computing" for many users.
> 
> 
> X11 was able to do this for many years! 

No. Not really. Only in theory. Since performance sucked, no-one really did
it set up as the standard way to do enterprise-wide computing. With the NX
technology,  I think it could become a real option.

> So in order not to repeat myself to 
> often:
> 
> - The main thing NX does is on one hand a much more efficient version of the 
> X11 protocol and on the other hand providing a lean and free (as in beer) 
> client for many operating systems incl. Unix/Linux, Windows, Opie, 
> Symbian....
> 
> IMHO the main feature with regards to kde-promo and NX is that NX finally 
> makes thin client / fat server computing a REAL alternativ. Sofar ICA was 20 
> times better than the bloated X11 protocoll.
> 
> Regards,
> -- martin
> P.S.: My NX testdrive was my first real usage of the RH KDE setup. Oh my god 
> it is really _extremly_ broken :-((

I warned you...

But it shouldn't stop everyone to testdrive NX. "The proof for the pudding
lies in the eating":

   http://www.nomachine.com/testdrive.php

> Dipl.-Phys. Martin Konold
> 

Cheers,
Kurt

 
>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<
[prev in list] [next in list] [prev in thread] [next in thread]