NX Compression Technology To Go Closed Source

An anonymous reader writes "NoMachine has sneakily revealed it is closing its source of the NX compression technology with NX 4.0: 'This release marks an important milestone in the history of the company. Version 4.0 of the software, in fact, will be only available under a closed source license.'"

Comments

Sneakily revealed?

[RAMMS EIN]

"NoMachine has sneakily revealed..."

That's quite remarkable.

Re:Sneakily revealed?

[Monkeedude1212]

It would appear they've achieved the impossible.

Congrats NoMachine! No matter how often I try to sneakily reveal something, I'm either too sneaky and nothing is revealed, or its revealed and not very sneaky at all!

Re:Sneakily revealed?

[Jah-Wren Ryel]

It would appear they've achieved the impossible.

It's part of their new NX 4.0 implementation - they have other forms of encryption likes AES and 3DES but Sneaky Reveal is their own proprietary encryption algorithm.

Re:Sneakily revealed?

[newcastlejon]

Indeed! I thought nothing could be seen in a NoRoo...

Oh, wait. Nevermind.

Re:Sneakily revealed?

[jelle]

Hmm... "proprietary encryption"?

http://www.interhack.net/people/cmcurtin/snake-oil-faq.html

Re:Sneakily revealed?

[siddesu]

"Sneaky Reveal" is also apparently a part of their new lingerie line. The part for playful adults.

Re:Sneakily revealed?

[fucket]

Hmm... 'Hmm... "proprietary encryption"?'?

http://en.wikipedia.org/wiki/Joke

Re:Sneakily revealed?

[mysidia]

Sneaky Reveal is their own proprietary encryption algorithm.

Which you should not trust, because it has not been through cryptanalytic rigor and attack attempts by the community at large.

They can publish their "Sneaky Reveal" algorithm details, and then, maybe we can take a look at it and see if the algorithm is cryptographically strong, or if it's just smoke and mirrors.

Re:Sneakily revealed?

[cOldhandle]

People following Jo Moore's infamous advice might just achieve that elusive "sneaky reveal" - http://news.bbc.co.uk/2/hi/uk_news/politics/1823120.stm

The more reason to use something else.

[sethstorm]

VNC or reverse engineer an open alternative.

Re:The more reason to use something else.

[KiloByte]

There's no need to reverse engineer anything. Version 3 is there, and no one is going to give a damn about 4.

Anyone who feels the open version lacks something is free to extend it on their own. If not... well... it's not going to stop working outright, and at least security bugs will be fixed in a timely manner.

Re:The more reason to use something else.

[madprof]

Except VNC lets you control a Windows/Unix/Mac machine from a Windows/Unix/Mac machine just fine. Unlike NX.

Re:The more reason to use something else.

[harrkev]

Version 3 is in fact very good, but not perfect. It seems to have problems if the client is on a system with multiple monitors. Also, I have seen crashes when I full-screen SOC-Encounter. An update/bug fix would be very welcome.

This product is simply the BEST remote software for *NIX systems, period. VNC (all flavors) runs like an absolute dog compared to NX and, depending on the program, it as times completely unusable, while NX is generally very smooth.

Re:The more reason to use something else.

[Svartalf]

Why bother? The NX compression tech prior to this release is GPLed.

You don't NEED to reverse engineer anything. Just re-name it and take it down the way.

Re:The more reason to use something else.

[characterZer0]

Version 3 has one fatal flaw. It requires password authentication, and cannot do public key authn. If version 4 addresses this, I'd pay for it.

Re:The more reason to use something else.

[dpilot]

I've never used NX because it always seemed almost insurmountable to do it "free". There was free NX available, but from everything I read, it was much more difficult to set up, and I never had the time to go through it all. Add to that the fact that it seemed necessary to pick up "freely available" parts from NoMachine. I work for a company that's very picky about licenses, and knowing that an attorney's reflexive answer is "NO!" I just stick to known-good licenses like GPL, BSD, and other things I know are already approved.

So I figure in the cumulative time it would take to navigate the maze called NX, setup, licensing, etc, I can just use vnc occasionally.

Re:The more reason to use something else.

[ehrichweiss]

VNC only offers a low form of control compared to what I get out of NX(IIRC, I never could get VNC to recognize more than 3 buttons on my mouse but I need the use of all 5), or rdesktop for that matter, and you don't need any special ports open, just SSH. And to top it off, NX offers better compression which is why they're likely closing the source. I have tried to convince NoMachine to make a Windows version but apparently they don't care about it despite my offers of wealth.

Re:The more reason to use something else.

[ehrichweiss]

You can use PK authentication for the SSH traffic(for example if you want to thwart bruteforce attempts by allowing ONLY PK authentication) but the NX login requires a password regardless, I assume for security measures; I personally am cool with that.

Re:The more reason to use something else.

[Belial6]

Make that Windows/Unix/Mac/Wii/Atari/Android/iPhone/Amiga/Anything that runs java/Palm/WinCE/C64/Etc./Etc./Etc...

I would love to see a better format for screen sharing, but if it doesn't work with everything, it will have a hard time replacing VNC.

Re:The more reason to use something else.

[Galactic Dominator]

Caveman!

I couldn't possibly get any work done without my 18 button mouse. Standard tasks are ridiculously inefficient on anything less.

Re:The more reason to use something else.

[ehrichweiss]

Well, yeah, I have several friends who need at least 9 buttons...I'm just a bit of a simpleton in some regards.

Re:The more reason to use something else.

[breagerey]

*shrug
I use the free version all the time and it's super easy to setup (less than 5 minutes ... seriously ... go look at nomachine.org).

It's *way faster than VNC and runs my session silently while another user is logged into the remote box.
(maybe I can do that with VNC as well and I just don't know how to ... I just never bothered trying since NX seems so much faster)

Re:The more reason to use something else.

[RzUpAnmsCwrds]

Right, I've used NX over a WAN link (albeit a very fast one) and it's able to handle non-intensive tasks like running Eclipse or general GNOME apps on a 2560x1600 display with reasonable performance. You can forget trying to do much of anything at that resolution with VNC over a WAN.

That said, NX totally dies if you try to do anything with animation or video. Protocols like PCoIP or HP RGS do a lot better here since they compress more.

Re:The more reason to use something else.

[sl3xd]

I've never seen NX perform as well as VNC - espescially Turbo/Tiger VNC with VirtualGL.

If you're doing remote 3D/scientific visualization, Turbo/Tiger VNC and VirtualGL is far and away better than NX. I've gotten 20 fps for 1280x1024 3D graphics -- over a 2 Mbit connection. VirtualGL/TurboVNC can also handle clusters of GPU's (ie. many nodes, each rendering part of an extremely complex image). The final screen buffer is then tunneled over SSH. It's amazing, frankly, to have some-odd 16 or so GPU's rendering a hellaciously complex 3D scene, and then it gets sent over a tiny pipe using TurboVNC & VirtualGL, and is then displayed on a netbook with very usable frame rates.

I didn't believe it was possible until I tried it myself. Recall you can stream a movie at close to DVD quality at 2 MBit; so I guess it's not that unbelieveable after all.

NX, in comparison, couldn't even start glxgears.

you don't need any special ports open, just SSH

SSH is all you need for anything. It's port forwarding and tunneling capabilities can be used for anything that uses TCP/IP - NFS, VNC, Samba, HTTP, video games - anything.

NX does the same thing - it uses SSH port forwarding & tunneling; it just handles it transparently for the user.

This is hardly unique to NX, as many VNC implementations also uses SSH transparently for the user.

And it's been pointed out - there isn't a major platform that VNC doesn't support.

Re:The more reason to use something else.

[MemoryDragon]

The problem simply is X itself, the protocol is in a serious need for an overhaul, maybe now that things are moving towards wayland and push the X protocol on top of the rendering stack than being the base of everything as huge big X server blob will get things moving. Making Cairo remote seems like a sane choice and has been done as websocket demonstration already.

Re:The more reason to use something else.

[phtpht]

the NX login requires a password regardless, I assume for security measures

Ouch. Getting your password transferred to & used at arbitrary locations is hardly more secure than the challenge/response PK schema in which your secrets never leave your computer.

IMO the reasons behind this are merely laziness/incompetence on top of bad design.

I personally am cool with that

I am not. First of all it is un-secure to enter your password somewhere without really having control of where it goes.

Second, it is a pain in the ass. Even if you use PK for all your ssh access you will have to maintain a password just for nx sake. Come on in the year 2010 (or 2011 for that matter) it is totally idiotic to use ssh with passwords for normal daily jobs.

That's why NX made its way out of my door very shortly after trying it out.

If you want to know the whole story then read on.

What really happens with NX is that your client first opens a ssh session to nx@server - this session uses PK auth. Unfortunately the private key is well known, as it ships with a default one and almost noone bothers with replacing it (in fact it is not advised - lol). That means anyone in the world can open that connection to your server and get authenticated and proceed to step 2.

Coming to step 2 after logging to your server as user nx the nx server program is launched. This program actually manages translating the X window protocol to the nx protocol - all the "compression" and stuff. It will set up a DISPLAY variable to point to itself and then launches ANOTHER ssh to you@localhost (localhost being the server). That's where your password comes in.

After that X applications (usually the whole desktop) can be launched under your account, they are going to talk to the sshd spawned by ssh @localhost, where the X protocol is encrypted and compressed, shortly afterwards decompressed and decrypted by the nx server running the ssh command, then translated to nx protocol, encrypted and compressed again through the ssh session for user nx all the way to your client pc where finally decrypted, decompressed, translated to X and displayed on your screen.

As you can see the weird part here is the user nx running the nx server. It is technically absolutely possible to avoid at all this step. The nx server could be easily launched on behalf of the user with the applications spawned from there. IIRC the introduction of the nx account is something as ill as licensing.

Even with the nx account present it is totally irresponsible to leave it protected by well known PK, for the sake of making the installation & deployment 2 clicks easier. The impact is that anyone can launch a nx server on your server. They claim it is secure - hah. At the very least you can start brute forcing passwords for local users, at the other extreme you can find a neat hole to hijack the process and make your way in to the system. Combine with the knowledge of some kernel loophole and you can start scanning for nx enabled systems all over the internet with satisfaction guaranteed.

And even with the ssh@localhost weirdness it is feasible to use PK auth, either by use of ssh agent or some sort of channel forwarding. But no, that's not supported, because...

... enter the customized ssh client shipped with nx which one of the steps involves. Being customized it is never kept up to date so say goodbye to features and say welcome another lot of security holes.

And finally when I read the forums about the open source client wondering if all this mess was fixed at last, I found out that it wasn't/won't - for COMPATIBILITY reasons.

Sad story.

Re:The more reason to use something else.

[hbr]

Above post is informative.

When I installed nx the reason for the nx user is the wierdest thing about the installation - I still don't think I understand the rationale here. It doesn't seem like good design to me.

Added to that you have the necessity to do a full xdm login session - that also seems overkill as usually I just want a single xclient running remotely. I think you maybe can do a single client at a time, but I don't understand how to do it.

Otherwise, though, I have to say, though, that NX is fantastic - it does the job it is designed for amazingly well. It seems much better than VNC to me because it deals with the X protocol itself, rather than just redrawing patches of screen. I qualify this with the fact I am only interested in *nix.

I wonder what constraints there would be on developing the open-source version of nx - e.g. to throw out the nx user and make it simple to launch xclients a client at a time? I don't know how much better version 4 is than version 3, but version 3 is still 2-orders of magnitude better than doing x "directly" over ssh for slow connections.

Re:The more reason to use something else.

[PybusJ]

I've gotten 20 fps for 1280x1024 3D graphics -- over a 2 Mbit connection.

On the other hand for usable browsing and general desktop sessions NX doesn't need close to 2Mbit, it works well for me at 56kbit. So it's horses for courses, I guess.

It's not the VNC protocol that gives you your good 3D performance, it's the architecture of VirtualGL. I don't think there's a good reason VirtualGL couldn't be made to work with NX as well as VNC.

The future probably belongs to SPICE, which redhat (a company who do know how to develop open source code) are creating for remote access to virtualised systems.

Re:The more reason to use something else.

[ultranova]

I couldn't possibly get any work done without my 18 button mouse. Standard tasks are ridiculously inefficient on anything less.

Just put the mouse electronics on the bottom of a keyboard and enjoy your new 102-button two-handed mouse.

Re:The more reason to use something else.

[GameboyRMH]

What's the CPU usage like? TurboVNC sounds like it could make remote desktops less painful over slow connections, my only two concerns are client compatibility and CPU usage. I use a lot of different client apps, and it has to work on my N900 which has a stock clock speed of 600Mhz.

Re:The more reason to use something else.

[GameboyRMH]

It's *way faster than VNC and runs my session silently while another user is logged into the remote box.
(maybe I can do that with VNC as well and I just don't know how to ... I just never bothered trying since NX seems so much faster)

Uh isn't that how VNC normally works? :-\

I know there's a way to have it work Windows-style, like a true remote control, but it doesn't seem useful to me and it's a PITA to set up that way so I never bothered.

Re:The more reason to use something else.

[Anonymous Coward]

If I was you, I'd give a try to the 4 and see if your statements still hold true.

The 4 was a complete rewrite aimed at making NX the best remote desktop software ever in Windows and Mac environments (while obviously maintaining the leadership on Linux/X11). It's four years of work, spanning from making it 100% multithreaded, accelerated on the client and native on all platforms (read no more Cygwin). In fact the 4 will debut the server for Windows and Mac. This means that the target was not the remote Unix worker interested in accessing the corporate desktop (we had a good product solving this problem already), but the consumer interested in watching DVDs from the laptop computer on a NX enabled device in the living room. We are trying to go beyond the definition of "remote desktop". There must be nothing that you can do with your physical computer that you can't do with NX.

Re:The more reason to use something else.

[ehrichweiss]

"Second, it is a pain in the ass. Even if you use PK for all your ssh access you will have to maintain a password just for nx sake. Come on in the year 2010 (or 2011 for that matter) it is totally idiotic to use ssh with passwords for normal daily jobs."

Ok, I've gotta call you out on something... You have a good and informative post, I will agree, but you also called the devs at NX lazy, complained about your perception of their security measures and yet you're here complaining about having to use a password....

I personally like the password because that means that someone can't get unchecked graphical access to my remote machines by simply catching me on a bathroom break with my screen unlocked. In your scenario it would only take a single click. Yes, they would have access to my laptop but I'd rather have at least a password between them and access to the remote systems I admin.

I don't ever remember being encouraged to keep the same public key for NX but I can safely say that my setup uses my own PK.

Also, I don't think that you have the NX login sequence entirely correct because while the first part seems accurate, I can safely say that NX can NOT be spawning a second ssh session using my password because passwords are forbidden in ssh by policy on my system, so it'd have to be using my PK, or doing some other form of login. I could be wrong but that's the impression I got while learning to set it up with RSA keys a while back. And just to let anyone know who might be interested in setting up RSA PK authentication with NX, you have to use NoMachines node/server/client from their site, the FreeNX/OpenNX only do DSA.

Also, if I have my own PK pair setup for ssh, I'm pretty certain that even if I leave the NX keys as default, there isn't much chance of someone using the NX account to bruteforce passwords as you suggest because even if my system allowed passwords for ssh, they'd still need my PK to even get the NX session spawned. Again, I could be wrong on any of these points but I checked this pretty thoroughly and that's what I've seen thus far.

Regardless, I see your complaints boiling down to the out-of-the-box experience of NX versus having to actually work to configure and secure the system properly. I don't really have issues with having to do extra work to do that.

Re:The more reason to use something else.

[ehrichweiss]

Windows and Mac server?!?!? Finally, no more rdesktop!

Re:The more reason to use something else.

[Guspaz]

VirtualGL is strictly for 3D-based stuff, though. Useless in the typical desktop use case (unless you can remote an OpenGL compositor with it?). TurboVNC is still ultimately just sending JPEG frames over the internet, which is both CPU intensive, and bandwidth intensive.

The big advantage of NX is that you're sending the underlying XDMCP data, executing the draw instructions directly, rather than trying to detect changes on the screen and then sending chunks of the screen over the internet as images. NX's primary speedup comes by caching (not compression), as it eliminates most of the round-trips that XDMCP normally requires.

In my personal experience, I've found NX to be enormously faster than VNC, more and more dramatically so on slower and slower connections.

A more interesting solution would be VirtualGL/NX, although it doesn't sound like VirtualGL is terribly useful over the internet.

Re:The more reason to use something else.

[phtpht]

I personally like the password because that means that someone can't get unchecked graphical access to my remote machines by simply catching me on a bathroom break with my screen unlocked. In your scenario it would only take a single click. Yes, they would have access to my laptop but I'd rather have at least a password between them and access to the remote systems I admin.

I don't get the point here. You leave your screen unlocked then I gain instant access to where ever you're currently logged on. Plus I can stick a neat backdoor to your user account to automagically hijack all your further sessions (if time allows, depending on the magnitude of bathroom service you're utilizing...)

So please leave your screen locked. (BTW even if you lock it I can stick a key sniffer to your keyboard and pwn you anyway.)

Anyway what you probably originally meant is that PK auth is passwordless. Well it can be used that way but you can also put a password on your private key. Then you have to type that password in order to auth. The key difference here is that neither the password nor the private key leaves your computer and the login is unrepeatable without possession of the private key. That's in contrast to classic password auth where the pw is sent to the server where it can be sniffed and re-used. So what I meant in my original post is that in 2010 there is really no reason to use oldschool insecure auth methods.

Also, I don't think that you have the NX login sequence entirely correct because while the first part seems accurate, I can safely say that NX can NOT be spawning a second ssh session using my password because passwords are forbidden in ssh by policy on my system, so it'd have to be using my PK, or doing some other form of login.

Okay my experience with NX is about 18 months old and on behalf of this discussion I re-checked their docs today. It seems there is another way, that is using an internal database of users. The relevant source is http://www.nomachine.com/documents/admin-guide.php section 4. So either you're using that option or you should really re-check your SSH logs and config ;-)

I don't ever remember being encouraged to keep the same public key for NX but I can safely say that my setup uses my own PK.

Now I can't find the reference but I swear I saw a paragraph saying something like if you replace the keys then out of box installs will not connect blah blah. Yeah I hear you say "big deal" but ffs this is about serious loss of security so there should never be a default key thus negating the necessity for mentioning "replacing the default key" somewhere in the manual. The keys should be simply generated at install time, with a message popping out "give this [public key content] to any Joe Schmoe you want to access that server of yours".

And just to let anyone know who might be interested in setting up RSA PK authentication with NX, you have to use NoMachines node/server/client from their site, the FreeNX/OpenNX only do DSA.

Sounds weird to me, why the heck would that be so? Both use the same ssh, which supports both.

Re:The more reason to use something else.

[ehrichweiss]

"I don't get the point here. You leave your screen unlocked then I gain instant access to where ever you're currently logged on."

You don't get the point because you don't have my setup, nor did you pay attention to a key word: "graphical". Non-GUI access to my systems isn't nearly as damaging as GUI access to them. For example, if I accidentally leave a document on which I'm working on my desktop, it's not readily apparent via CLI, but you can see it plain as day if you logon graphically. That's a simple example, but I have far more reasons than just that. BTW, my screensaver crashes for some unknown reason from time to time so while I do typically lock it, I don't like to take the risk, so a password is a good extra step to provide that security.

"So either you're using that option or you should really re-check your SSH logs and config ;-)"

Nope, neither of those is true. My sshd_config has PasswordAuthentication = no. Nothing can login via ssh(not even 127.0.0.1) without the proper PK.

I do agree they should generate a new key upon install.....and they offer that as well. If you read a bit more there's a step in the setup that asks if you want to use the default key pair. If you answer "no", it'll generate one for you. Are you sure you used NX for any *real* period of time? I ask because it seems you didn't give it much of a fair chance because this was a feature(at least in Free/OpenNX, and I'm sure NoMachine's had it as well but I haven't used it in this regard in over a year so I can't be certain) since I first started using it some 4+ years ago.

"Sounds weird to me, why the heck would that be so? Both use the same ssh, which supports both."

Apparently because the Open/FreeNX had issues with RSA. I don't remember what that problem was but I do remember spending several months hammering out the problem after moving to PK-only authentication only to discover that it was solved by using NoMachine's setup which took me 30 seconds to configure. I actually have to do that again because of a recent system crash so maybe I'll play with it a bit and see if I can answer that question for you in depth.

Re:The more reason to use something else.

[Abcd1234]

I've never seen NX perform as well as VNC

You're nuts.

No, seriously, you're fucking crazy.

NX over a low-bandwidth or high-latency link blows VNC clear out of the water. It's not even comparable. Combine that with the ability of NX to run rootless, and it's a *far* better solution than VNC will ever be.

The twats who modded you up have clearly never used NX before. And you, it seems, have only tried it over a large pipe, where NX will, of course, lose out to other technologies better suited for that usage model.

Re:The more reason to use something else.

[Kagetsuki]

You are not using TightVNC or something equally capable. I just checked and it recognized all 5 buttons and the scroll wheel on my mouse, not to mention I use a Japanese keyboard and there is no problem there either. Furthermore ssh tunneling is supported painlessly with the -via argument in tightvnc. I have no ports open other than ssh, which is on a non-standard port to begin with.

Re:The more reason to use something else.

[Kagetsuki]

And VNC doesn't work for you on 56kb? Have you tried enabling compression?

Re:The more reason to use something else.

[phtpht]

You don't get the point because you don't have my setup, nor did you pay attention to a key word: "graphical". Non-GUI access to my systems isn't nearly as damaging as GUI access to them. For example, if I accidentally leave a document on which I'm working on my desktop, it's not readily apparent via CLI, but you can see it plain as day if you logon graphically. I don't like to take the risk, so a password is a good extra step to provide that security.

What I said is universally applicable. Graphics does not matter. "A password" can be used on private key too. And "a password" will not help you against someone determined to do the damage if they already have access to your account. Unless this is some sort of protection from 9 year olds that is.

my screensaver crashes for some unknown reason from time to time so while

Crashed screen saver usually renders the X session unusable because it would make precautions so that if it crashed then it would not expose working desktop to anyone while you're taking your break. That is at least the old school ones do this, maybe the new fancy ones don't :[

My sshd_config has PasswordAuthentication = no. Nothing can login via ssh(not even 127.0.0.1) without the proper PK.

Then the documentation lies...

I do agree they should generate a new key upon install.....and they offer that as well. If you read a bit more there's a step in the setup that asks if you want to use the default key pair. If you answer "no", it'll generate one for you. Are you sure you used NX for any *real* period of time?

Yes I devoted some fair time to make it working. (I failed on the PK auth step.) I installed the open/free version from packages so there were no questions asked. Hey maybe I am wrong and all crap like this has already been fixed - good for you/them/whoever then. But back then I assure you it wasn't just any one thing that pissed me off, it was the whole concert of them.

Fast remote X connection...

[oldhack]

Such a niche tech. Who are their customer base?

Re:Fast remote X connection...

[kriston]

Try as I might, I simply could not get this "whatever-it-is" software to work, and I know something about getting "whatever-it-is" software to work.

Nothing of value is lost here.

Re:Fast remote X connection...

[MichaelKristopeit319]

idiots...

i know some people that demand to use remote X to open a local terminal window rather than SSH'ing to the machine.

Re:Fast remote X connection...

[rduke15]

I am a real CLI addict, but some things still require a GUI.

For example comparing a server's /etc tree with another one, and applying changes.I found Meld to be great for that. But to be able to effectively run Meld on an otherwise headless remote server, connected through a slow ADSL link, I need NX.

Plain X forwarding is fine on a LAN, but it's not really usable over ADSL.

Re:Fast remote X connection...

[gatkinso]

Your kidding I hope.

On my CentOS box it entailed the following gruelling sequence of steps:

$ sudo yum install nx freenx

Then, install a client, connect to your server, pretend you are working on your server.

This is useless for the runlevel 3 crowd (however my box is at 3 and it runs quite nicely).

Re:Fast remote X connection...

[gatkinso]

Ha!

In practice, NX doesn't provide more than PuTTy and XMing.

But is is impressive software.

Re:Fast remote X connection...

[vgerclover]

For the runlevel 3 crowd ssh will always be there.

Re:Fast remote X connection...

[micheas]

Maybe ssh X11 forwarding with compression would help?

What I have done to get X over ssh to work acceptably over dsl.

Set a simple theme for gtk and qt apps. Having images as the background for menu items seems to slow things down.

In addition to enabling compression, make sure that ssh is configured so that CompressionLevel is 9 instead of the default of 6.

Finally, if I am running a windowmanager remotely blackbox (and its offspring) are really light on network resources.

I have not had to do anything else to get things working acceptably.

Re:Fast remote X connection...

[seifried]

For example comparing a server's /etc tree with another one, and applying changes.

Ever consider "diff" and "patch"? Seriously....

Re:Fast remote X connection...

[X0563511]

Your kidding I hope.

On my CentOS box it entailed the following gruelling sequence of steps:

$ sudo yum install nx freenx

Then, install a client, connect to your server, pretend you are working on your server.

This is useless for the runlevel 3 crowd (however my box is at 3 and it runs quite nicely).

Yea, installing the software isn't very hard. Configuring it and getting the client etc to play nice is where it gets not so fun.

Even so - you can blame RedHat (or whoever maintains that package) for making your package so nice and easy. Not everyone has that indirect expert assistance handy.

Re:Fast remote X connection...

[Gaygirlie]

Ssh with compression simply doesn't even nearly compare to NX and its speed. NX does in fact run over ssh and uses compression, but the NX server also does lots of other tricks on its end, like caching of tiles, selectively updating only certain regions of the screen and so on. It all adds up and once you've tried NX you'll notice the difference. I've been using NX for several years now and I have not found anything remotely as capable or as fast as it.

Oh, and then there's the fact that NX client is useable on Windows whereas X11 forwarding to a Windows client would require quite a lot of hacking around.

Re:Fast remote X connection...

[Hobart]

Grandparent is a whippersnapper.

twm + xterm work fine over X11-over-compressed-SSH on dialup PPP links on a 28.8 modem.

You'd think these people had never had to install their OS on a 9600bps serial link on an 80x25 screen with no cursor control.

*shakes fist*

Re:Fast remote X connection...

[sl3xd]

I've given NX too many chances as it is. I kept giving it "one more try", but I decided to call it quits before this announcement.

I've always found VNC to be a better option - espescially after working with TurboVNC and its successor, TigerVNC in concert with VirtualGL.

Being able to get astonishingly high frame rates (15-20 fps) at high resolution with 3D rendered graphics over a 1.5 MB connection is an impressive feat - one which I can't duplicate using NX.

VNC lets you control the connection security - meaning you can do something sane like use a passphrase and an authentication agent. In secure environments, you can go without any ssh tunneling at all.

So after trying NX, and pretty much every other thing out there, I've always found that VNC works - and works well - when other options fail.

Add to that the fact you can get VNC for every platform out there, and NX just doesn't stack up well.

Re:Fast remote X connection...

[jimicus]

There is one crucial difference - it's a hell of a lot faster than X over SSH.

Re:Fast remote X connection...

[kasperd]

i know some people that demand to use remote X to open a local terminal window rather than SSH'ing to the machine.

Sometimes running xterm over X forwarding over ssh is the best way to work. It's a bit easier to get that working correctly than just plain ssh in those cases where one end of the connection uses ISO 8859-1 and the other end uses UTF-8.

Re:Fast remote X connection...

[Stephen Chadfield]

We use this at work. We have a collection of fast servers which run complex semiconductor simulation jobs that can take days to complete. The engineers start these in an NX session then detach. They can subsequently re-attach to check the progress of this work from any site or by connecting into the company network from home via VPN. The performance of NX is acceptable for the graphical applications which display and manipulate the data.

Re:Fast remote X connection...

[dovgr]

Use sshfs to mount your remote machine's disk and then run meld locally. No remote access protocol will be faster than that.

Re:Fast remote X connection...

[TheRaven64]

Usually that makes things worse. X11 doesn't actually use much bandwidth, but it is very sensitive to latency. Adding compression on top of encryption adds more latency. A lot of this is the design of xlib, which tries to provide a synchronous API on top of an asynchronous protocol - XCB is a lot better if used correctly, as it lets you easily overlap requests and responses.

The compression that NX provides is okay - it's tailored to the X11 protocol, so it is a bit better than using a generic stream compressor, but the real benefit comes from the caching. The NX proxy will reply immediately to a lot of requests from the client, reducing latency.

Re:Fast remote X connection...

[GameboyRMH]

Or rsync, if he's distributing changes from one server to many others...

Re:Fast remote X connection...

[Anonymous Coward]

Fist? Oh, you young people. We used to shake our fins.

Re:Fast remote X connection...

[buchner.johannes]

Use sshfs to mount your remote machine's disk and then run meld locally. No remote access protocol will be faster than that.

You are ignoring the fact that all the data has to be transferred to localhost then, which will be more costly than running a GUI remotely.

Re:Fast remote X connection...

[phantomcircuit]

X Forwarding works just find over ADSL if you turn on ssh compression. ssh -X is practically useless ssh -XC is fast.

Re:Fast remote X connection...

[rduke15]

Ever consider "diff" and "patch"?

Meld actually uses diff. But it then displays the results in a practical and clear interface for reviewing which change to keep and which not. diff/patch and Meld serve very different purposes, so your remark doesn't make much sense.

Re:Fast remote X connection...

[dkf]

Plain X forwarding is fine on a LAN, but it's not really usable over ADSL.

Only because you insist on having masses of eye candy. Forego that, and you can run just fine with ADSL. Heck, I've done real work with remote X running over a 14k4 modem (ick!) and that worked (by "real work" I mean Framemaker and xfig; everything else that I needed at the time was just fine locally or in a plain terminal). It's only for eye candy (or photo editing, though I've never needed that professionally) that you really want to have a high-bandwidth low-latency connection between application and graphics display.

Re:Fast remote X connection...

[Kagetsuki]

If you set up X11 forwarding through ssh properly (which for some reason is actually harder than it should be) things like Meld will run fine. I actually use Meld a lot when I merge branches in a git repository... on my server in another city. Once you get X11 forwarding to stop locking up your application every 5 seconds on slow links it works quite well.

Re:Fast remote X connection...

[Kagetsuki]

You can X11 forward TO a Windows client with Cygwin, just install X11 and SSH. FROM Windows on the other hand would obviously not work.

Re:Fast remote X connection...

[Kagetsuki]

I'm sorry, but Meld is extremely nice. I use it constantly when dealing with git repositories and having to use diff and patch in place of it would be a nightmare.

Re:Fast remote X connection...

[Siffy]

Or NFS, sshfs, or even Samba/CIFS if he must have a GUI for diff. In /etc it makes sense.

Re:Fast remote X connection...

[kriston]

Yea, installing the software isn't very hard. Configuring it and getting the client etc to play nice is where it gets not so fun.

Even so - you can blame RedHat (or whoever maintains that package) for making your package so nice and easy. Not everyone has that indirect expert assistance handy.

I always need custom configuration. This software is really difficult.

Like X0563511 said, installation is not the problem. Using it is the problem.

Nothing of value is lost here.

On the bright side, I use Logmein.com and jump using realvnc to the unix host, but then again, I haven't had much need for graphical unix hosts lately, so I just use ZOC or PuTTY.

New?

[RAMMS EIN]

Perhaps I don't remember it right, but in my recollection, NoMachine has always been a bit possessive with their (definitely impressive) technology. To the point that lesser alternatives have continued to be used and even developed.

Re:New?

[Trepidity]

Well, they previously opened under an "open source it but make the open-source version a pain-in-the-ass to use" business model. You got a random code dump that wasn't even buildable. However, there was code there, and it was possible to fix it up, which is what the FreeNX and OpenNX projects did (along with adding a few things). With no GPL release of the core libraries, it's no longer possible to even use it as a base for a cleaned-up open-source release--- either projects will have to independently develop a fork of the previous version, or come up with something else.

Re:New?

[Hatta]

Indeed. It's like they *want* people to not use it.

Re:New?

[countSudoku()]

Yes, they want people to use VNC, like any normal person would do. What they really want is people to depend on their product, then lock them in with a new version that is paid, but that ain't gonna happen. VNC is just fine, and X11 is free, plus every company I've ever worked for wanted to buy my Hummingbird, so why would I even bother with NoMachine? The only admin who ever recommended it to me was of the Windows side of the house, and I'm not buying. So long, and thanks for all the fish!

Re:New?

[mattdm]

Presumably, the success of FreeNX and the advent of OpenNX is what pushed them over the edge -- the "code dump that no one can use hahaha" model wasn't working for them.

Re:New?

[Junta]

I'm hoping this will be the impetus for the forks to abandon all pretense of interoperability with NoMachine's crap session management and do it right.

Re:New?

[keeboo]

NoMachine has always been a bit possessive with their (definitely impressive) technology.

To be fair, it's not like NX appeared from nowhere: it's a fork from DXPC.

Re:New?

[ne0n]

you might even classify some alternatives, eg x2go as superior. freenx and neatx are ok sometimes but if you have the option...

Re:New?

[segedunum]

I wouldn't use VNC over anything that doesn't provide LAN speeds, which kind of defeats the object of a remote desktop in a way. Sorry. Normal people know VNC is slow. People might like to pretend that the open source world has something like RDP, but they don't.

Re:New?

[PhilipJLewis]

xrdp on Linux works for me - not quite as fast as NX but certainly better then VNC variants I have tried.

Re:New?

[GameboyRMH]

VNC can definitely get painful at low speeds, but LAN speeds? It works okay over a 3Mbps connection.

violating software patents?

[MichaelKristopeit319]

why else would they close their source when the compression industry is already saturated with near optimal free products?

Re:violating software patents?

[Rockoon]

"Optimal" is subjective.

The patents would most likely be related to speed, rather than compression ratio. For instance, all the good (fast) arithmetic encoders are still under patent. These trade a small bit of compression efficiency to avoid the expensive division per symbol that is otherwise required.

Re:violating software patents?

[julesh]

there is a fixed and easily definable limit to the amount of lossless compression a dataset can withstand.

Um. Okay. How many bits are there in the optimal coding for this post?

Turns out it isn't *that* easy to define.

Re:violating software patents?

[calmofthestorm]

Or performance on the input type. Sure lossless compression cares less what it operates on, but you can still have algorithms which make more sense for different types of input. As a trivial example, if the input to your compressor is one of five GB-long sequences known in advance, you can compress that down to 3 bits.

Re:violating software patents?

[dpilot]

One of the antecedents of NX was a thing called "dxpc", which I used to use. It did compression, but more performance came from simply being a proxy and short-circuiting many of the round-trip communications X did. X is a very chatty protocol, and much of that chatter can be intelligently done away with. That's one of the things that dxpc, and now NX did.

A year or two back I found dxpc source, dusted it off, and it actually built cleanly. But it didn't work worth spit. I guess X has moved on.

Re:violating software patents?

[Rockoon]

uh... optimal is NOT subjective at all.

Wrong, douche bag.

There are at least 4 metrics in the case of data compression that can be optimized for, and further that any metric can have a weighted level of importance:

1) Compression ratio.
2) Speed of compression.
3) Memory Overhead (the good compressors use a LOT of memory.. one compression competition limits memory use to "only" 1 gigabyte)
4) Recovery rate from corruption/transmission errors.

You are obviously one of those compression noobs that talks the big talk but doesnt have any applied experience in the matter.

Re:violating software patents?

[noidentity]

It's not a general-purpose compression product; it's specialized for the X windows protocol.

Re:violating software patents?

[thePowerOfGrayskull]

Pun intended? ah, given who's posting, probably not...

Re:violating software patents?

[Fast Thick Pants]

Good list -- also needs items 2.5 and 3.5 for the speed and memory requirements of the decompression side.

Re:violating software patents?

[X0563511]

Who shit in your cereal this morning? Nice job on the personal attacks.

Re:violating software patents?

[Rockoon]

also considering that those very weighted levels DEFINE AN OPTIMAL SOLUTION... I'm an idiot."

Fixed that for the douche bag.

Re:violating software patents?

[segedunum]

You'll have to tell us what those 'optimal' products are, because as we should all know by now remote X is totally unusable over anything but a LAN and ditto for VNC.

If no one does anything with NX or the state of remote graphical protocols in the open source world then there is nothing that is comparable to RDP. I wouldn't reply really, but you see trolls repeating this stuff when they have not a clue.

Re:violating software patents?

[RAMMS EIN]

Indeed, usually, the slowness in X is caused by network latency (the exception being if you are rendering a lot of pixels, e.g. for movies). Moreover, the slowness is often not inherent in the X protocol, but rather caused by how an application uses it. Some X clients are amazingly fast, even over moderate latency links.

This is also why you often get a more responsive UI by using something that just pushes pixel data, like VNC, instead of X. They work faster, even though they are less efficient in terms of amount of network traffic. It's not the throughput, it's the latency.

You could actually do the same thing with X: just render your whole app to an XImage, then render that to the server. This will be faster than synchronously performing all your drawing operations over the network, if you do lots of drawing operations. On the other hand, if you have lots of images that you tend to reuse, store them on the server as XPixmaps, and then you can render them faster than you could by pushing the pixels each time. X offers you this choice, and when used well, can actually be _faster_ than other technologies over any kind of network. The only thing I haven't found is a way to compress pixel data, but perhaps that is just because I haven't looked hard enough.

Re:violating software patents?

[kasperd]

How many bits are there in the optimal coding for this post?

He said it was easily definable, not that it was easily computable. The length of the optimal coding would be the kolmogorov complexity. It is easy to define, but it is undecidable. Computing the length is equivalent to solving the halting problem.

Re:violating software patents?

[Rockoon]

there is only ONE optimal solution for fixed parameters and fixed input.

Which means that optimal is based on subjective choices of parameters, which you claimed wasn't true.

You: modded both troll and flamebait
Me: modded informative

You: don't understand the subject
Me: been dealing with this subject for several decades

Re:violating software patents?

[eriqk]

A year or two back I found dxpc source, dusted it off, and it actually built cleanly. But it didn't work worth spit. I guess X has moved on.

This thread had me looking for dxpc in the Ubuntu repos and whaddayaknow, it's there! I'm tempted to give it a whirl.

Re:violating software patents?

[Rockoon]

Let me quote you:

First you declare "optimal is NOT subjective at all. there is a fixed and easily definable limit to the amount of lossless compression a dataset can withstand." as a response to "optimal is subjective"

Now you claim "defining the average content stream and system requirements were NECESSARY BEFORE defining THE optimal solution."

Cart before the horse much? You originally had no idea that other factors besides "amount of lossless compression" were commonly important in data compression..

And the funniest part is that you declared that there was a "fixed and easily definable limit" to lossless compression. No sir, you are wrong. It is not easy to define at all. The estimate for English is somewhere between 0.6 and 1.3 bits per character, with the best compressors still performing on average worse than 1.0 bits per character. Nobody knows if 0.6 is possible or not, hence there cannot be an "easily definable limit"

Re:violating software patents?

[Rockoon]

i said there was a fixed and easily definable limit TO A GIVEN DATASET.

No there isn't.

Re:violating software patents?

[Rockoon]

There is no way to measure the total entropy of a sequence, but instead only the entropy given a specific model of that sequence. There is also no way to show that a given model describes the optimal entropy.

So you have used a circular reference, by stating that optimal data compression is rigidly defined by moving the undefined parts to a word that you dont understand.

Re:violating software patents?

[MichaelKristopeit328]

ur mum's face have used a circular reference you don't understand.

give me a fixed bitstream. i will give you an optimal compression method.

you're an ignorant hypocrite.

cower some more, feeb.

you're completely pathetic.

Re:violating software patents?

[Rockoon]

OK. The fixed bitstream I submit is the enwik8 text file used in the Hutter Prize compression competition.

I do not want your results. I want you to claim the bulk of the Hutter Prize money. I look forward to seeing Michael Kristopeit as the top entry.

Re:violating software patents?

[Rockoon]

SO claim the prize, son. Claim the prize.

Re:violating software patents?

[MichaelKristopeit335]

only my father would correctly call me son.

you're an idiot.

cower some more, feeb.

you're completely pathetic.

So what

[gd23ka]

Someone will come along and make a better opensource alternative to it.

Re:So what

[man_of_mr_e]

And nobody had created anything as good as BitKeeper either until there was suddenly a need for someone to do so when the free license was pulled.

Of course that person that came along was Linus Torvolds, not your typical hack.

Re:So what

[MemoryDragon]

Actually not only Thorvalds dont forget about Mercurial which is as good as GIT (although I prefer GIT over Merc despite its rotten command set)

Re:So what

[barnacle]

slashdot is dead when there are 2 misspellings of Linus' last name in the same thread and nobody corrects it

Re:So what

[bcmm]

In this case, however, the previous version's source is available as a starting point. Remember XFree86?

Re:So what

[HBI]

Everyone left except the left.

Re:so what

[pak9rabid]

Correction: x2go actually uses the NX protocol...but they've cleaned up the installation process quite a bit...I'd imagine they'll just continue using the current NX version to keep it free.

Re:So what

[juan2074]

Torr-volts?

Buried in tl;dr

[tepples]

It would appear they've achieved the impossible.

Not impossible. To me, "sneakily reveal" sounds like "bury in a tl;dr changelog".

Re:Buried in tl;dr

[Monkeedude1212]

The problem with trying to hide something in a tl;dr changelog is that Someone - somewhere - WILL read the change log, and likely make mention of anything out of the ordinary on their blog.

In fact - I went to renew my Xbox live Gold membership a month ago or so... And they said that the terms and conditions of the service had changed since I last was on Gold. So I decided I would read through them. However - to continue this anecdote and help explain why it might be remotely funny - is that I had previously set my regional settings of my Xbox to Spain, and language to Spanish - so that when I got achievements in Halo 3 they came up as a different language. It's true, you can go and change it to like, Korean, then get some achievements, and no matter what language you go from there on out - they will come up in whatever language you achieved them in. I did this for a while, finding it to be of great amusement when someone new came over to my house and just happened to look at which achievements I had.

Anyways, so this ended up backfiring on me because the EULA and TOC of Live was now in Spanish, but I thought perhaps there was English at the bottom. However, there is no fast scroll when looking at the TOC - its very slow and you have to hold the analog stick down. To my dismay, there wasn't English. Afraid that going next would Mean I accepted to terms I didn't actually agree too - I quickly pulled up my computer and typed it in verbatim the entire thing into Google Translator.

If you've ever tried reading legalese in a language you do not understand - I highly recommend you NEVER EVER try it. Even after going through an internet translator you still will have no idea what the heck they are saying.

In hindsight I probably just should have not accepted the terms, gone and changed my language settings, and then gone through it - but I guess that seems obvious now, it didn't back then.

Re:Buried in tl;dr

[Midnight Thunder]

I was about to say something snarky and suggest you should have read it online, except that it seems pretty difficult to find. I haven't found it yet. Maybe you can't read it unless you are on the XBox, which would be rather bothersome?

Re:Buried in tl;dr

[kiwimate]

Well, I thought exactly the same thing as the grandparent - "sneakily revealed" is a fairly remarkable statement.

The link is "http://www.nomachine.com/news-read.php?idnews=331", so it would appear that it's on a regular news portion of their site - hardly sneaky. Furthermore, it's mentioned in the very first paragraph.

Rome, Italy, December 21, 2010 - NoMachine, a global leader in cross platform remote access and application delivery solutions, announced a software preview of its upcoming new products and technologies which offer a completely redesigned client GUI and restructure its flagship suite of NX Server. The new products will not only extend the current functionalities of NX application delivery and remote access products, there will also be new naming conventions adopted. This release marks an important milestone in the history of the company. Version 4 of the software, in fact, will be only available under a closed source license.

Beats me how they're being sneaky about doing it, at any rate. Shall we agree that the summary is (shock, stunner, surprise!) badly written, or at least very biased, and go on to debating the impact of the move?

Re:Buried in tl;dr

[TheLink]

Anyways, so this ended up backfiring on me because the EULA and TOC of Live was now in Spanish,

Nobody expects the Spanish interpretation!

Re:Buried in tl;dr

[timeOday]

Microsoft intentionally makes dealing with XBox Live extremely difficult and time-consuming. You can't unregister to stop their billing charges online - you have to call & wait. Cancel the credit card? They lock the account (stop providing the service) but keep accumulating charges, then hit you with back charges if you ever put in a valid card number. They're jerks.

Re:Buried in tl;dr

[Score Whore]

Really? I stopped my service online. Maybe you just have to log into their service with a web browser? Did you try that?

Re:Buried in tl;dr

[FatdogHaiku]

So, last sentence of the first paragraph of a press release that starts with "Rome, Italy, December 21, 2010" is not sneaky? They even hid a link to it in plain sight on the front page of the site... how much more evidence do we need?
Unloose the villagers with their digital pitchforks and torches!!!

Re:Buried in tl;dr

[timeOday]

I did it online once too, a few months ago. But last night I spent about 20 minutes searching the menus and searching for help, and I think they removed the function.

Re:Buried in tl;dr

[Kilrah_il]

Our chief weapon is surprise. Surprise and fear... Our two weapons are fear and surprise... and horrible machine-language interpretation. Amongst our weapons.... Amongst our weaponry...are such elements as fear, surprise.... I'll come in again.

Re:Buried in tl;dr

[WWWWolf]

You think that's weird?

The latest XBL EULA update was shown to me in Swedish. I've specified that I'm in Finland, the email notices from XBL come in Finnish, I'm using the dashboard in English (Finnish isn't available), and the xbox.com website shows up in Finnish for me. And boom, EULA in Swedish. What the hell. Granted, it's one of the official languages, but it's, um, a bit rusty for me. I definitely can't comprehend legalese.

Re:Buried in tl;dr

[jimshatt]

IANAL, but wouldn't you have effectively sidestepped any legal consequence of the EULA changes? You can easily claim you just pressed a button and didn't realize you were accepting the new conditions at all (and thereby not accepting them). I guess the same argument can be made for the English version, i.e. you don't have to understand or even read EULAs, much less agree with them, and still click the 'Next' button. This is especially valid for EULA dialog's that don't have a checkbox you have to check first before continuing ("By clicking 'Next' you agree to the following...").

Re:Buried in tl;dr

[GameboyRMH]

They're jerks.

Who would ever have imagined Microsoft would treat their customers like dirt! This must have been quite a shock to them!

Re:Buried in tl;dr

[timeOday]

So, because you're too stupid to navigate their web site, it's now impossible?

Also this:

Cancel your Xbox LIVE Gold Membership

Learn how to cancel your Xbox LIVE Gold Membership.

* To cancel your Xbox LIVE Gold Membership or to turn off auto-renewal, you need to contact Xbox Support.

And if you follow the link on "contact XBox Support," the only way listed is to call.

I'm surprised anybody would bother to start making insults without bothering to simply supply a link. (Then again we are talking XBox Live here. There would be more incentive for normal people to stick around if there were an easy way to segregate all the morons such as yourself into your own little playpen).

Re:Buried in tl;dr

[buchner.johannes]

Every 10th story on slashdot tells what Apple and Microsoft have "quietly announced", so sneakily revealing should not be too much of a oxymoronic surprise.

Re:Buried in tl;dr

[DaVince21]

Still better than the Wii, which doesn't let you set any language other than the one of the country you're in. Set anything else, and half of the built-in Wii channels suddenly stop working. I want my games in English, not in broken, badly translated Dutch (my native language)!

NX is a bandaid

[girlintraining]

Rather than bitch about how they're making it closed source, or dismissing the gesture entirely, maybe this should be taken as a sign that the problem NX solves needs a different solution. Like, oh I don't know... maybe revising the X windows protocol so it doesn't suck so hard it has its own event horizon?

Re:NX is a bandaid

[Grishnakh]

Unless I'm missing something, Wayland isn't network transparent like X is. Wayland's advantage is that it's simpler than X, so it has higher performance. For those needing network transparency, X can be run on top of Wayland, but then you're still stuck with X's obsolete protocol, rendering it not very useful on slower network connections.

Re:NX is a bandaid

[spitzak]

It should be possible to make Wayland "network transparent". This would be the job of the "compositor". The Compositor takes the window images and assembles them into the screen display, and also takes user interface events and sends them to the programs. There is no reason it cannot assemble a window onto a remote display by talking to the Compositor running there, and return input events from that remote display. It can take hints about what areas of the windows were updated and doing comparisons and data compression so the images are sent quickly.

This is how Microsoft does remove windowing and it works reasonably well. Also how VNC works though they don't have access to the low levels similar to the Compositor.

Re:NX is a bandaid

[walshy007]

X can be run on top of Wayland, but then you're still stuck with X's obsolete protocol,

Once again, wayland does not do a third of what X does, I repeat, wayland is NOT a suitable replacement for x, it does not handle window creation or mouse events or anything (which is what x11 is mostly used for these days)

Wayland is still useful as a screen multiplexer etc. But it does not do what most people think it does, hell read it's website with it's goals.

It is far simpler because it doesn't have anywhere near the scope of X, by itself it is useless. This is why x is still needed.

Re:NX is a bandaid

[IGnatius T Foobar]

One might even think about writing a Wayland compositor that speaks RDP, making use of all of the existing clients on every platform.

Re:NX is a bandaid

[Billly Gates]

Do people use old school terminals anymore? Remote work is done with SSH today anyway.

X is redundant and even the creator admitted it is obsolete and is holding Linux behind.

If you like the idea of using remote access then using something through a browser.

Re:NX is a bandaid

[dbIII]

Bullshit. Where I am everyone is busy using interactive software that involves a lot of mouse clicks and instant response on displays generated by a big cluster. Try doing that in a web browser instead of in X.
As we move back to distributed environments (cloud is the new buzzword for that), things like X will become more important.

Re:NX is a bandaid

[Gaygirlie]

maybe revising the X windows protocol so it doesn't suck so hard it has its own event horizon?

That would be a start. The X protocol is really chatty and there's lots of unneeded roundtrips. NX does filter away what isn't needed and compresses the rest into larger packets that then get sent all at once instead of small bits all the time. So, perhaps X.org should start doing the same.

But even that simply isn't enough. X11 forwarding would still be immensely slow compared to what NX does as NX for example keeps a cache of pixmaps on both server and client side, thus rendering the need for those pixmaps to be transmitted over the network all the time obsolete. Not to mention all the other small tricks it does that all add up.

I don't know, but I feel that simply revising X11 forwarding will not cut it. Besides, NX client works on other platforms too, not just Linux/Unix. I am atleast not aware of any good way of doing X11 forwarding to a Windows client.

Re:NX is a bandaid

[ogdenk]

I don't know, but I feel that simply revising X11 forwarding will not cut it. Besides, NX client works on other platforms too, not just Linux/Unix. I am atleast not aware of any good way of doing X11 forwarding to a Windows client.

It's called downloading PuTTY (the SSH client) and downloading a Winblows X11 server.....

X11 forwarding isn't all that bad unless you are a moron with your desktop environment. There's also the good ol' LBX extensions and such if they're still around.

Re:NX is a bandaid

[ChunderDownunder]

I'm not across the technical details but this would seem a great niche for wayland to develop a remoting protocol minus the chattiness of X you mention, using lessons learnt.
Backwards compatibility is the catchcry but are we still going to be debating X remoting 5 years from now if and when all the major distros have adopted 'something better'?

Re:NX is a bandaid

[RAMMS EIN]

``maybe this should be taken as a sign that the problem NX solves needs a different solution. Like, oh I don't know... maybe revising the X windows protocol so it doesn't suck so hard it has its own event horizon?''

Yeah, maybe. Of course, this has been tried many times before, and not with much success, I believe.

The truth to the matter is that the X protocol is actually pretty good. It's the software that implements it that could be better. That at least used to be true of both servers and Xlib (range checks, anyone?), as well as clients. If you think X is slow, it is almost certainly because you are using a client that causes a lot of network round trips and waits for them. There is absolutely nothing in the X protocol that forces you to do so. But if you do so, then, yes, your app will be very slow over a moderate-latency link.

Re:NX is a bandaid

[vlm]

That would be a start. The X protocol is really chatty and there's lots of unneeded roundtrips. NX does filter away what isn't needed and compresses the rest into larger packets that then get sent all at once instead of small bits all the time. So, perhaps X.org should start doing the same.

For the past 17 years I've been hearing about this, but the problem is it becomes less useful every year. So back in the 14.4K era I heard with some effort I could compress my X data stream when I connect remotely, but frankly the compatibility and deployment hassles have not been worth it. Now 17 years later maybe they've got the perfect solution to run on my multi-megabit cablemodem. Excellent, thats just like someone finally perfecting their floppy disk automated cataloger/indexer (remember those?) in 2010. Or finally developing the idea gopher server in 2010. Thanks guys, but if I wanted to pay for something obsolete I'd get a classic car instead.

Re:NX is a bandaid

[quintesse]

It's true that Wayland doesn't do everything that X does, but it does do some of the hardest parts, which mean that somebody might think about writing a replacement and not be called a nutter.

In fact it's what they're currently aming for: making GUIs for resource constrained environments where X is just overkill (like tablets and netbooks).

Combine that with years of experience with X and things like RDP and VNC and there are people who are saying we could build network transparancy on top of Wayland without too much trouble and probably out-perform X by a large factor.

Anyway, all these developments are quite interesting, at least things are moving, let's just see where they will take us.

Re:NX is a bandaid

[Gaygirlie]

You quite obviously don't run remote desktops over *DSL link; if you did you'd know it's really laggy if you just try to go with standard X11 port forwarding over SSH.

Besides, you're overlooking a LOT things, like for example mobile networks are quite popular nowadays. If you're on a mobile network any speed increases you can get are a plus. Oh, and what about mobile devices themselves? There's plenty of reasonable things one might want to/need to do on their real desktop computer while on the road, and then a remote desktop is the only solution.

Re:NX is a bandaid

[Abcd1234]

Like, oh I don't know... maybe revising the X windows protocol so it doesn't suck so hard it has its own event horizon?

Or we could just keep using the existing, already open NX codebase, fork, and move on with a perfectly usable technology, rather than starting from scratch under the naive belief that new == better?

Missing explanatory link in summary

[Anonymous Coward]

Here is an explanation of what NX compression technology is.

Re:Missing explanatory link in summary

[noidentity]

In summary, it speeds X windows sessions over a remote link, by compressing and caching data going across the link.

Who's had sucess?

[jvillain]

I am trying to think of any company that has closed their source and been more successful and I can't think of any. It sure didn't help SSH. NX used to be really great and a number of people I know used it. But I don't know any one that still does.

Re:Who's had sucess?

[Junta]

SSH lives on as Tectia and still has quite the revenue stream selling to companies that presume it's more secure because they have to pay for it. Commercial SSH has always just pretty much been for suckers and continues to be, the open source aspect being pretty well a moot point.

Re:Who's had sucess?

[geminidomino]

What about SAINT? (An honest question, I don't know how well they're doing.)

Re:Who's had sucess?

[tepples]

Commercial SSH has always just pretty much been for suckers

By "always", did you mean even before there was PuTTY or another decent free SSH client for Windows? Or did you mean Windows itself is for suckers?

Re:Who's had sucess?

[TorKlingberg]

NX used to be really great and a number of people I know used it. But I don't know any one that still does.

So what are they using now that is better than NX?

Re:Who's had sucess?

[GameboyRMH]

selling to companies that presume it's more secure because they have to pay for it.

Yep, the "nobody ever got fired for buying IBM" factor is still a big thing. At my company we'd use a lot of common, open, well-proven *nix tools instead of shitty, proprietary commercial crap that sort of does the same thing, if it was sold from a well-known vendor who could be called up and blamed if something went wrong - apparently even Red Hat isn't a good enough brand name for these guys...

Re:Who's had sucess?

[Junta]

I'll take the latter. Honestly though, I forgot about that aspect, but at least *now* I really dislike Tectia SSH Windows client compared to PuTTY even if both were free.

I really was only thinking of the use of it on Unix and Linux systems. I know companies uninstalling OpenSSH from RedHat and doing Tectia, which I cannot begin to fathom.

Re:Who's had sucess?

[Carewolf]

The SSH client for windows has always been free. It was the server you paid for. I still use the "commercial ssh" client at home, because it is free, and it works. Putty is only for netcafes, where the fact that it is a single executable that requires no installation is really convinient.

Re:Who's had sucess?

[Jumpy]

I just started using nx technology (freenx, nx, and the No Machine client) about a year ago and I absolutely depend on it. I find it better to remote into a Linux box to work rather than use rdesktop to a windows session. And then have to use xwin32 + putty with some windows crap which is just a shitty experience... I am using an EVDO modem (like from a cell phone company) and either rdesktop or NX are the only things that are fast enough to be usable for GUI type apps with my network connection from my personal experience. NX seems quite a bit faster then even rdesktop to windows... I am sad they are closing the source. Hopefully the freenx stuff will continue to work for a few years.

And for the people who say "Just use a browser, dumbass!" I'd love to do that or just use CLI type apps via ssh but sometimes it isn't an option. Sometimes the customer wants their backup and you are on the road or home sick and you have to launch some X windows app to do that because the company you work for bought a proprietary app to do that with and you have to use the X Windows GUI... and it has to be launched on some specific fileserver. I admit these kind of things are rare but the NX stuff is just awesome for that kind of situation.

Nobody gives a shit.

[RightSaidFred99]

It seems to me NX came up with some good technology and they think they should therefore be able to profit greatly off of it.

Like many pieces of great technology, however, nobody gives a shit about the problem NX is trying to solve. The market is tiny for the number of people who would buy this kind of technology, so they're not going to be the next VMWare.

Their only hope is to adapt it to something that is useful and sell it to some larger company.

Re:Nobody gives a shit.

[moonbender]

Not sure why you'd think that. There are a lot of people who want to do remote X (more remote than LAN), and NX currently seems to be the best, most convenient tool for that job.

Re:Nobody gives a shit.

[AaronW]

I was just using NX last night to connect to my Linux work machine from home. I've used VNC but my experience is that NX is much faster over my internet connection (20/8) than VNC was over a LAN, and this is running NX on Windows in a VM on my Linux box (because I've had some issues with the VPN in Linux).

NX is a lot more intelligent than VNC in that it caches a lot of stuff on the client side and is X aware. I.e. it keeps track of X bitmaps and will use jpeg compression on them when sending them across, renders fonts locally, etc. It is *MUCH* more responsive than running an X app remotely over ssh.

I've found NX to be quite usable even when the available bandwidth is fairly low whereas VNC would be useless. It actually seemed faster to run my web browser over NX rather than running it locally.

Sometimes I wish it could behave more like running a remote X application without having to bring up the entire desktop, but other than I'm a convert.

I've also seen a lot of VNC servers get borked where a VNC session suddenly starts gobbling up 100% of the CPU. I don't know if that's been fixed yet, but it was a major problem when I used it.

Re:Nobody gives a shit.

[madprof]

Well NX is a totally different thing to VNC anyway isn't it? I mean it's not just sending pixel data, which is all VNC ever does, even if it is compressed pixel data. NX to me is a Unix-flavoured version of RDP where it does clever stuff. Only it is tied to a platform too which is its biggest downfall.

Re:Nobody gives a shit.

[agm]

The performance of VNC over even "normal" broadband connections can be abysmal. NX runs rings around it for performance.

Re:Nobody gives a shit.

[X0563511]

Chicken and egg problem.

I find that from gigabit LAN usage, I would MOST DEFINITELY use the hell out of "X Forwarding" if it wasn't so godawful slow. I'm sure I'm not alone in this opinion.

Conventional systems such as RDP and VNC are tough, because the 'windowed' mode of it wastes a lot of space and makes it hard to integrate what your doing into your local environment. Systems that get around this (like Citrix) really are doing what X Forwarding does anyways.

Re:Nobody gives a shit.

[X0563511]

VNC opens a window that everything else goes in.

Running shit via "ssh -X [-Y]" makes anything I run integrate seamlessly into my environment.

Big difference, that.

Re:Nobody gives a shit.

[GameboyRMH]

Try to be more believable next time, maybe you'll get some lulz instead of just being modded down immediately.

Re:Nobody gives a shit.

[GameboyRMH]

Huh? I've never seen VNC pegging the CPU, and I have VNC sessions running 24/7 on my work servers and my home server. I have had a problem once on a CentOS box where I logged into a VNC session and got a black screen, I had to restart the VNC session, no big deal, it's only there as a convenience for me and as the only way the textophobic Windows admins could bring themselves to work with it.

Re:Nobody gives a shit.

[Jamori]

> Sometimes I wish it could behave more like running a remote X application without having to bring up the entire desktop, but other than I'm a convert.

It can! NX can integrate remote windows into your local environment in the same way that an app launched via 'ssh -X' does, rather than a full remote desktop via (like a VNC session).

Bring up the configuration window. Under the 'Desktop' settings, rather than choosing KDE, GNOME, XDM, etc, click 'Settings' and select the "Floating windows" radio option rather than "New virtual desktop".

Launch a new session (you can't resume your old one and have this setting apply itself), and you have smoothly integrated remote application windows into your local environment.

Re:Nobody gives a shit.

[neiko]

Sometimes I wish it could behave more like running a remote X application without having to bring up the entire desktop, but other than I'm a convert.

It can...that's the only way I use it. Check the options for your client.

Re:Nobody gives a shit.

[bill_mcgonigle]

Bring up the configuration window. Under the 'Desktop' settings, rather than choosing KDE, GNOME, XDM, etc, click 'Settings' and select the "Floating windows" radio option rather than "New virtual desktop".

Thanks!

Re:Nobody gives a shit.

[AaronW]

I have seen this on multiple servers, usually when a person doesn't disconnect nicely. I had one clueless user that would do this frequently with one server I administered and at another company where we had to use VNC I would often find the server load to be quite high with one or more VNC sessions pegging the CPU. It's possible it may have been due to users letting the screen saver run or something but I think the users usually had disconnected.

Part of it was good...

[Junta]

Specifically, the core tech isn't too shabby. Their session management stuff with awkward use of ssh keys and a separate user on top of the user's real account, a real awkward mess. I use a complete separate management of nxagent/nxproxy that doesn't use any extraneous user and it's nice.

The real shame is their strategy seemed to be oriented around selling their crappy session management stuff and giving away the quality stuff.

Er...

[Michael Hunt]

The reason that the "core" bits of NX were always Free is because dxpc (and, thus, mlview-dxpc, from which NX sprang) is only available under the GPL.

If i was involved in dxpc (or mlview-dxpc, really, although I'd imagine most of those changes are owned by the NX folks) development I'd be lawyering up at this point, if only to get some kind of proof that I wasn't being ripped off.

Re:Er...

[tepples]

If i was involved in dxpc (or mlview-dxpc, really, although I'd imagine most of those changes are owned by the NX folks) development I'd be lawyering up at this point, if only to get some kind of proof that I wasn't being ripped off.

NoMachine's defense might be to claim a ground-up rewrite of the parts of the program demonstrably derived from dxpc, just as the LAME team wrote replacements for ISO's MP3 encoder demonstration code piece by piece. How would your lawyer talk a judge into authorizing a fishing expedition to see whether this is the case?

Re:Er...

[MemoryDragon]

Actually I assume they now closed it up because they have removed the last trace of the old GPL core.

Well

[sajuuk]

Time to find an alternative to NoMachine then. I guess I could just go back to using nothing but the shell to remote-manage my web server.

Re:NX significantly better

[nurb432]

But better enough to pay for it? Most will say that its not. Some will.

Re:source fork?

[AaronW]

This seems the obvious solution to me. If their previous releases were under GPL, there's nothing stopping anyone from releasing a fork.

I am an NX user

[Jailbrekr]

There is nothing software based on the market right now which comes close to offering what NX has. Nothing. X, VNC and Spice all have their drawbacks and it boils down to how they behave with limited bandwidth and high latency. With NX, our engineers can actually be productive in an NX session. While I am philosophically opposed to them closing the source for version 4 and not sharing, the pragmatist in me is thinking "hey, built in web server, cool, that'll make it easier for me to allocate VMs to the engineers via a web gui. Screw that NX plugin crap.".

Until someone can come up with an open source product that does what NX does, as well as offer a web client, we're sticking with it. It simply is the best that is out there.

Isn't the cat already out of the bag?

[Thalaric]

I've been using FreeNX since Fedora 8 or so and it's superior in just about every way to VNC. Yeah, it's annoying brittle when the distro's packages are broken or not configured correctly, but that's just an implementation issue. The way it caches X calls makes a remote desktop run at local speeds over most broadband. This is a technology that needs to be brought mainstream.

That said, this is probably not the end of the world. They weren't bragging about any protocol level enhancements, more user level enhancements like a "revamped client GUI" which were never available under an open source license anyway. The core NX library is probably pretty mature by now. Since all future development on their side will be closed, it presents the opportunity for a real project to grow up around a fork. No more being held back by lockstep to a company that never wanted a successful open solution to begin with.

Re:NX significantly better

[aztracker1]

Not for me, as a Win/Mac/Nix user, I want something with clients/servers where i need them... VNC/VNC+SSH works... not as well as NX or RDP, but hey, that's the price I pay...

Wouldn't remote X11 over BZ stream cut it similarly?

You nobrainers and thieves

[BillBrains]

Instead of all just sitting there wondering why they are closing the source, and wondering what you can use in its place, just download and try it. They have obviously closed the source because they got fed up with "the community" taking a technology for free without giving a shit back. They have no doubt got some tricks up their sleeves, and who said no one uses it? It is clear you don't read much, you just sit there and drink your free beer. Who cares about forks, it's nomachine that's holding the knife by its handle.

Re:You nobrainers and thieves

[jimicus]

Whether intentional or not, that was definitely worded in a somewhat trollish way.

In any case, I'll bite.

The simple truth is... sssh, come closer.

Closer.

The "F/OSS community" is by and large abysmal at coming up with great new solutions. It's far better at taking an existing idea and running with it - probably because nobody has ever got a group of people together to say "With our experience and problems that we and so many others face, what fantastic solution can we put together?" without paying them first. OTOH, lots of successful projects have come about as a result of someone - or a small group of people - saying "We've got this product which has the potential to be a fantastic solution to problems X and Y, and maybe Z as well. Right now it's OK with X and Y, if a little rough around the edges. We haven't even started on Z, but we're opening it to the community".

What's really embarrassing is when the F/OSS community not only doesn't solve a problem, but pronounces it unsolveable even in the face of evidence to the contrary in the form of commercial software which is generally considered to solve it quite nicely, TYVM. I can think of quite a few OSS projects where the leaders have announced that they don't plan to even try to solve a particular problem because they have mathematically proven that it is impossible, only to proudly announce about 12-18 months later that their latest release does actually solve it.

Re:You nobrainers and thieves

[BillBrains]

Whether it sounds trollish or not, (I would say that it was intentional), the point is that so many projects give their hard work to the community who take, and then take and continue to take, and give nothing in return other than the malicious and evil comments (not all) that you can find on this page.

All the forks that the readers mention herein exist thanks to the developers behind NoMachine NX. Neatx would not exist without NoMachine, likewise all those other projects that continue to take the NX compression libraries and even complain that certain things don't work, or don't work well. If they don't work, or don't work well, why don't they just fix it themselves? What a shame it is that all these forks do exist now, only to disappear later. And they will because they don't have anyone holding their hand.

Linus himself says Git is not as good as BitKeeper

[melted]

Linus himself says Git is not as good as BitKeeper.

Re:NX significantly better

[MemoryDragon]

Actually when I had a need for Linux remoting which actually was not do slow, I paid for NX already years ago. It simply was so much better.
The glorified plain X remoting is a dog once you get out of the Lan area, and it is one of the reasons why I hate X passionately and wished the protocol finally would be overhauled for modern times.

VNC also almost never is a solution in WAN scenarios due to streaming also being dog slow. RDP and NX simply are the best protocols for such things there is no discussion.
And NX is basically X done right or more along the lines, X fixed up so that the glorified remote protocol actually is usable.

Re:Linus himself says Git is not as good as BitKee

[Dr_Barnowl]

[citation needed]

Linus : "I'm an egotistical bastard"

Imagining Linus actually admitting that something was better than something he wrote is hard. git soundly beats BitKeeper in terms of performance ; BK was managing a kernel patch merge in 6 seconds at a time that git was managing 6 patches per second. Squishy-feeling UI considerations don't sound like Linus' bag, and he wrote git specifically to cater to his needs, so I'm struggling to grasp what he would consider "better" about BitKeeper over git.

He thought BitKeeper was better than all the other VCS systems at the time he started git, that I will concede.

NX may well become obsolete anyway

[DrXym]

If major dists like Ubuntu move to Wayland (and run X11 over the top much the way OS X does when it needs X11), then all the over the wire remoting will have to radically change anyway.

Screenscaping (VNC) will suck as much as it always did and X11 derived solutions (NX) won't work. Now might be the time to consider how over the wire functionality such as 3d, compositing etc. occurs and plumb it in in a much more coherent and integrated way with Wayland. Really there should be no reason for there to be a separate product to do this anyway on the server side. On the client side I imagine there will be lots of competing products both commercial and open source which would consume the content.

How graphics went over the wire really depends on the capabilities of the server and of the client and the datapipe between them. I expect that OpenGL instructions would form the basis of it, but would there be scope to merge, concatenate, cull instructions over the wire? Would there be scope for the server to perform certain transforms to spare the client the effort? Should the server send all textures, shaders, meshes etc. as is or also provide functionality to simplify them? Perhaps the client has a more powerful GPU than the server. Or perhaps it's the other way around. Perhaps the server would have to render most of the screen in some cases or not other. Perhaps the server has too many connections and doesn't have the GPU available to do this. Lots of questions, but it seems like this is all plumbing that should be considered for Wayland.

Obviously things like mouse, clipboard have to be considered too but I expect those issues are largely solved and fairly trivial compared to the graphical bits.

Good riddance

[hackel]

lol, they've just signed their company's death warrant. They couldn't even get people to use NX when it was free! What hope is there now? Oh well... Hopefully Wayland will be able to do this much better anyway.

Re:Good riddance

[GameboyRMH]

You can hope, but Wayland has no desktop forwarding capability and they have no plans whatsoever of adding it.

woah

[fireylord]

and just how much of their work was based on other people's work?

Re:Hmmmm...

[GameboyRMH]

VNC's simplicity and basically universal compatibility has always been its trump card to me...it's a clunky old thing but it works with anything and doesn't give trouble, for basic remote desktop capability it's the obvious choice.

Anyone remember 4Front Technologies?

[Ant P.]

No, and for this reason.

Re: Commercial SSH can be worthwhile

[tangent]

Commercial SSH has always just pretty much been for suckers...

That's probably true if you refer only to command line ssh.

For GUI SSH, however, spending a bit of coin can be very much worth it.

I have to use Windows on my work desktop, from which I help manage hundreds of remote Linux boxen. It would be madness to do that that with PuTTY, with its horrible UI, or with ssh in a Cygwin window, where session management is entirely missing, unless you count ~/.ssh/config. The commercial GUI SSH client we use[*] lets us organize all those sessions, save login info, and sync all that among the handful of computers used by the people who I help with those remote servers. A user name, password, or IP changes about once a week, somewhere. Managing that with just command line tools would be less efficient. Further, we get features like a tabbed UI, integrated file transfer, scripting, etc. We pay back our hundred bucks a seat in efficiency pretty quickly.

[*] no point naming it, I'm not here to sell anyone anything

NX vs. VNC side by side test

[rockytopchip]

I frequently use NX and have just now performed a side by side test which at least shows for my situation that NX is better than VNC. I've got two identical PC's on a local network, each running ubuntu 10.04 updated. Local machine has two flat screen displays, I've got an NX full screen connection on the right display, and a VNC full screen connection on the left. The remote machine is used to access some graphical waveform control screens. Output of top shows VNC (vinagre) is using 8% of CPU compared to 4% of CPU for NX (nxssh). Both are using same amount of memory at 0.6%. Just moving the mouse around in the window shows the difference between the two, the mouse pointer movement is smooth on the NX display and jumpy on VNC. Windows open immediately on the NX display and are slow to display on VNC. I tried VNC with and without JPEG compression option and it made no apparent difference.

Nobody thinks VNC is fast

[OrangeTide]

I have no doubt that VNC is inferior in performance to NX. But a much harder comparison to perform is to compare NX to RDesktop. The two protocols are more like each other than like VNC, and I'd expect their performance to be alike as well. NX is more like LBX (low bandwidth X) in terms of performance, although significantly better in my opinion, and more portable (LBX never really caught on).

VNC is popular not because it is fast, but because it is so very easy. It's way easier to setup VNC than NX. There are lots and lots of clients and servers available for VNC. And having written my own VNC server from scratch I can tell you one reason for this is because it is easy to develop VNC software. In addition there are plenty of open source VNC library forks to begin development on a VNC based project.

so what

[pak9rabid]

NX sucks anyways. Yeah the compression is good and once setup it works alright, but getting it setup is a bitchy process. This is where x2go and (from what I've heard) Xrdp shine. Good luck with your shitty proprietary technology that nobody will use in the future.

What's the impact?

[DaVince21]

It doesn't seem of any importance whatsoever to me, considering I had never heard of NX before opening this page.