XLiveCD: Cygwin and X For Windows On A Live CD

mallumax writes "OSnews is running a story on XliveCD which runs an X server (from X.org) from the CD using Cygwin. Also included are awk, sed, perl, vim, bash, grep, other text utilities, and most importantly an OpenSSH client. XliveCD is being developed by University Technology Services of Indiana University. Now you can carry Cygwin with you! I have been looking for something like this for a long time. Torrent link."

Comments

The point?

[Random Data]

I was going to ask what the point was, given the number of Live CDs such as Knoppix, etc. Then I actually RTFA and they suggest it's for use in public access Windows boxes, where a reboot may not be available but running stuff from the CD is.

I still suspect VNC on a USB key or CD might be easier, and the difference between forwarding X and using VNC isn't that much in my experience.

Re:The point?

[fearlezz]

they suggest it's for use in public access Windows boxes

Great idea. There's only one problem with public access Windows boxes: they cannot be trusted. Public access machines may have keyloggers, backdoors and stuff. I don't want anyone to see my password because I login from a windows box. Don't trust internet cafe's... So, personally, I was thinking more of using this when I'm visiting family and friends.

Re:The point?

[Minna Kirai]

Are you kidding? I used to run xterm and Emacs over 2400 baud and it was tolerable,

Your example is inconclusive. Network speed has two independent components: bandwidth and latency. Just because your modem had poor bandwidth doesn't mean the latency was also bad.

And indeed, many aspects of the X11 protocol involve almost gratuitous round-trip queries that can make high latency a killer. Often it's aspects of the GUI toolkits that create this problem- a pretty effect that seems cool & fast on a localbox can be sluggish on the network.

Specific real-life example: in Evolution, you move an email from one folder to another, and the application draws a little translucent icon flying from origin to destination as a feedback indicator. It covers about 200 pixels distance, and for each step, the applcation downloads the remote image of the workspace under that position, alpha-blends a pixmap ontop of it, and sends the pixmap back to the viewer.

On a long-haul link, this can take MINUTES, during which you can't interact with any X11 programs. If you were running the program under VNC, however, the whole animation would be over before a screen update is even transmitted. While the user has missed-out on some eye-candy, this is far better than waiting through all the bidirectional traffic.

On the other hand, TightVNC is not usable for serious work even over a cable modem.

For truely serious work, it's not usable even on a loopback interface to localhost.

Even if everything else was the same, VNC has to refresh the whole screen,

False. VNC is an extensible protocol, so it can support arbitrarily intelligent update mechanisms. But even the original generation of VNC clients were smart enough only to update the screen regions that were actually changing.

(If your VNC experience has primarily been with the Windows server, you might not have noticed this, but that's because it was difficult for them to interface with the server without full screen-scraping and mouse-yanking, as Windows wasn't designed to allow concurrent users)

Re:The point?

[Minna Kirai]

So, configure ssh to use S/Key, generate some one time pass phrases, and carry this live CD with you. Login remotely to your system, be careful not to do anything which is security sensitive, and you are set.

No you're not. If the client machine is compromised, one-time logins cannot protect you. The local SSH client could do evil things in a hidden side-channel to your actual work.

I've seen proof-of-concept modifed SSH clients which secretly download files from the remote homedir whenever anyone connects to anyplace (and that's in addition to logging all the activities of the session, of course).

With work, the operating system could be modified to recognize known popular SSH clients (such as putty.exe on your USB drive, or this X LiveCD thing), and secretly replace it with a compromised version when you attempt to run from your supposedly-trusted removable media.

to do anything which is security sensitive,

Um... if the activity was genuinely insensitive to security, you could run naked telnet. It's true that attacks like I described are probably rare enough that many people would be willing to run the risk, but they should still be aware of the threats and make that choice on their own. Elaborate multi-part attacks will only become more common as time goes by.

Re:I don't get it.

[BrynM]

What is the point?

Consider these ideas:

• Take a CD to work and use X on any machine with a CD drive
• Use this as a framework to add more applications like dev tools, auditing tools or desktop apps (probably forking the project, but interesting)
• Use it to prove to the ill tempered that *Nix is not all bad and quite usable
• Find a way to port it to Flash drives and such
• Another valid use of BitTorrent :D (the download of this iso that is)

Remember that this is an early version. The best uses are to come probably.

Re:Wait...

[ahg]

I'm sorry if I appear to be to be knit-picking but...

You have the client-server relationship wrong, a common mistake when talking about X windows.

The application running on Unix is the client to the X-server running, in this case, off a Windows machines.

Client examples: xterm, konqueror, Kmail, Evolution running on *nix

Server: X.org running on _Windows_ via Cygwin

Re:I don't get it.

[UnderScan]

> What is the point?
The point is easy interoperability between *nix and Windows OSs. See my post at OSNews.com http://www.osnews.com/comment.php?news_id=9163#313 438 of it's ease of use.

You can also find more info from from a paper published for the SIGUCCS of the ACM titled: Easy access to remote graphical UNIX applications for windows users Listed below is the publicly available abstract:
ABSTRACT
A barrier deters Windows users from evaluating graphical scientific software that runs only on remote UNIX systems. Graphical UNIX applications are based on X Windows. To make use of X applications, Windows users must install an X server, install communications software for connecting to remote UNIX systems, and configure their systems to display graphics from remote systems. This barrier can be removed by making use of an X server and communications software that run live from CD-ROM. This poster presents such a CD-ROM known as XLiveCD.

XLiveCD appears to users as an application that provides a command-prompt that allows them to log in to remote computers. Windows XP/NT/ 2000 users insert the CD into a drive and click twice in response to a wizard. A terminal window appears on the screen and provides a command prompt. From the command prompt users run the secure shell (ssh) to connect to a remote computer and launch applications. X graphics windows are forwarded automatically.

XLiveCD is based entirely on open source software and is available free for download. It is a Cygwin environment (from Red Hat, Inc.), including the X.org X server and openssh installed and modified to be run from CD-ROM. The home page is http://xlivecd.indiana.edu/.

Re:Wait...

[steveha]

I really, really wish the X guys hadn't used this terminology. But they did, so we need to keep it straight.

Just think of it this way:

A file server provides files to its clients. A print server provides printing to its clients. An X Window server provides graphical windows to its clients.

Thus, when you run any X application, it is a client to the X Window server. It asks the server for a window to display stuff in.

So, if you buy an expensive rack-mount server machine, and you hook up a thin client that lets you use a GUI, that thin client has an X server on it, and the X server talks to X clients that run on the server.

The neat thing is that in the other universe (the one where Spock has a beard), they call "clients" "servers" and "servers" "clients", but the X guys still did it backwards there so this confusion still applies.

Hope this helps.

steveha

Re:I don't get it.

[dario_moreno]

I see the point for portables computers with WiFi or GPRS Internet access, which can be a pain (or simply impossible) to setup under Linux. Imagine that you can borrow a laptop from time to time at work to go on the road, and cannot afford to take hours to install cygwin, but with this solution instead you can in seconds connect to your remote Unix server.

Re:I don't get it.

[MoneyMan]

The point is that you can sit at any windows machine, which may not necessarily be your own, and have a decent set of utilities to use.

I personaly work on many machines on any given day. The majority of which I do not own. I'm not "allowed" by my customer to go and just start throwing applications onto their system willy-nilly.

With this, I can work on any machine, using a shell I know, (bash), have a functional Xserver available, and access to a bajillion other GNU utilities without ever installing a single app.

Ever needed to tail a file in windows? It's there. Yes, there is a tail app for windows, and it's free. The point here is that this doesn't need to be installed. Grep? same thing.

Just boot to a LiveCD distro, you say? But I need to see what's happening on this Virii / Spyware ridden hunk o' junk while it's running windows.

Could I build my own suite? Yeah... but why would I? This has what I need.

Kudos and my thanks to the Cygwin team.

Re:I don't get it.

[BrynM]

It's a crufty layer on top of Windows

Exactly! You don't have to RTFA to know that this means you don't have to even re-boot your machine to use it. This is an advantake to those of us on W32 boxen at work that are refused a *nix dev machine. If you think it sucks, don't use it. I happen to see some usefulness for me. To each their own.

Re:Wait...

[steveha]

The terminology makes sense, but is it sensible?

When you run the server on your thin client, and the clients all run on your rackmount server, and the newbies are all confused, and we need to write posts explaining why the seemingly backwards terminology is in fact correct... in some sense, it's sensible, but if you take a poll of a bunch of newbies, the consensus would be that it's confusing. (I'll stop now before someone beats me senseless.)

If some terminology makes arguable sense but confuses everyone but hard-core computer geeks, is it really the best terminology? I say no.

steveha