Slashdot Mirror

I do something pretty similar except I stay in the *nix environment unless on my laptop. Sure, vnc can be used like a "pcanywhere" tool, but it is much more.

I never have an X desktop on a server since I rarely have a monitor and mouse attached. But, I do have a video card in the server, usually a cheap S3. I configure XFree86 for the video card and use vncserver to create virtual X screens. Often, I will have two different vncservers running, one in kde, another in windowmaker. I start and stop these as needed via ssh. The virtual X sessions are not displayed anywhere on the server. Each instance of vncviewer, on whatever platform, can display a virtual screen. According to the docs, you can even have multiple vncviewers displaying on the same screen. However, this I have never had a need to try. That might just push me over the edge.

Give vnc a try. Its not the fastest or most secure thing, but it has served me very well. You assign an X screen number when starting the server and connect with :. When you disconnect, the session stays alive until the vncserver is shut down. You can connect to it from anywhere you have vncviewer.

I have been trialing GCC 3.1 and the C++ compilation is a major improvement over past version of GCC. Being C based the GNOME 1.4 and GNOME 2 libraries and most applications compiles and runs without too much hassle. However KDE 3.0.1 is somewhat more problematic, even when neither debugging support nor strict syntax checking is enabled..

This is not the fault of either the KDE or GCC developers. KDE was coded to support the "older" C++ style of pre GCC 2.9x and Microsoft's compilers and the GCC Team is following the new C99 & ISO 14882 C++ standards.

After kludgeing around the defects in the older GCC C++ template and library implementations, GCC 3.1 C++ is real joy to use. It makes it possible to program C++ in a completely new styles, that IMO can be far more productive.

It is difficult layering one type of programing style over another, the older C++ style libraries certainly make Windows programing a pain.

Would it not be better to wait for the KDE team to port KDE to a pure GCC 3.1/ISO 14882 style?

At the very least the debugging support is required for GCC's Profile Driven Optimizations which can greatly improve application performance.

GNOME 2.0 is due for release soon enough, at the very least the GNOME libraries and core should also be included at a United Linux "main component".

But why do they need to go to college to become programmers, especially if they are promising? What they need is a good base in logic, algorithms, craftsmanship, a language, and some hard earned experience. I have no degree and no high-school diploma, but I own my own personal library devoted to programming , read and code regularly, and can still (yes, even after the .bombs) write my own ticket for jobs. Most people in the programming industries these days want to know what you can do and how you can do it - not where you when to school or what your GPA is, because quite honestly there is a huge influx of really poor but really heavily degreed and certified programmers and administrators. One of the hazards of being at the top of the "most desired profession" lists for too long I think....

A ``typical C++ programmer'' writes ``int* p;'' and explains it ``p is a pointer to an int'' emphasizing type. Indeed the type of p is int*. I clearly prefer that emphasis and see it as important for using the more advanced parts of C++ well.

I tend not to use multiple declarations per line; the problems outweigh the positives (With multiple declarations: You cannot comment variable usage later, it's (slightly) more typing to refactor code by removing or changing variables, and there's the problem you bring up).

Once you've adopted that rule, it's a lot clearer (and closer to the truth) to use "type*", since the actual type of "var1" in your example is "pointer to type", not "type".

It's also the style Bjarne uses in his code, if that makes any difference to you.

Well, according to Stroustrop, overloading whitespace and using single-character unicode variable names can help increase clarity. :)

Apple's list of venders promising support for the new server (without any actual product commitments) is at : http://www.apple.com/pr/library/2002/may/14quotes. html

Without a doubt, the BSD-base is the best thing this server has going for it. Without knowing more, I'd say that VNC is going to be a big deal for people wanting to use this thing without necessarily giving a Mac to their network admins. (Speaking of admins, has Apple figured out how to sell the major consulting companies on this thing?)

In at least one incident, a nuclear weapon did come very close to accidental detonation. In 1961, a B-52 with two large warheads crashed near Goldsboro, North Carolina; the impact set off the conventional explosives in one of the bombs, and triggered all but one of the safety mechanisms in the other.

In at least one incident, a nuclear weapon did come very close to accidental detonation. In 1961, a B-52 with two large warheads crashed near Goldsboro, North Carolina; the impact set off the conventional explosives in one of the bombs, and triggered all but one of the safety mechanisms in the other.

On a related note, I've concentrated quite a bit over the past year in generating charts and graphs to multiple output formats simultaneously. Of interest here are some new versions of old standard apps, Gnuplot and Graphviz. Both applications contain SVG output capabilities in their latest builds. I've been using both to generate both PNG, but also PS and SVG. I will generally convert the PS to PDF via Ghostscript. To me this represents an incredible time savings by allowing me to generate PNG's to act as thumbnails for the SVG and PDF versions of the same graph. Consider also the ability to nest SVG objects within a larger SVG picture (or animation). To aid in technical illustration you could actually embed SVG chart animations with the other custom 2D animations that you seek to create to further clarify the idea you are trying to present.

Raster, vector and publication quality visualizations in one fell swoop without spending a dime. Schweet!

AT&T are working on systems that track everything in an office so you desktop can follow you around, phone calls can be routed to the nearest phone to you, etc:

AT&T - Location Systems

They have some videos to watch on the topic... a few years old, but still interesting:

AT&T - The Video Collection

They have a nice away around objections to the system knowing where you are... if someone looks you up, you get their name and details...

It's is a pretty specialised system, though - I can't see it exactly being any use in your average family household...

I mean, who wants that kind of thing in their home?

AT&T are working on systems that track everything in an office so you desktop can follow you around, phone calls can be routed to the nearest phone to you, etc:

AT&T - Location Systems

They have some videos to watch on the topic... a few years old, but still interesting:

AT&T - The Video Collection

They have a nice away around objections to the system knowing where you are... if someone looks you up, you get their name and details...

It's is a pretty specialised system, though - I can't see it exactly being any use in your average family household...

I mean, who wants that kind of thing in their home?

I should have done some more research before posting; seems that VNC will do the modem thing for me: rtfm. So all that's left is to get Linux to cope with my modem. Groovy.

In 2001, Bjarne Stroustrup started a dialogue about the future of C++. As primary inventor of C++, Bjarne is giving interviews, visiting user's groups, and posting in forums, all with the intent of stimulating discussion about where C++ should "go." It's an important topic for software engineers, and everyone has a laundry list of features they'd like to see added in the next revision of C++.

I'll buck the laundry-list trend and suggest some things I don't want to put in the next C++.

In my experience, C++ iterators, algorithms, and containers are inefficient and unnecessarily complex. The actual source code doesn't look terribly confusing — it's the underlying mechanisms that obscure function with too much form. We heap template upon template, giving the compiler nightmares while obscuring what is really happening "under the hood."

Is container abstraction akin to the hiding a car's pistons from its driver? No, because I'm not driving the car, I'm building it. And as any good engineer can tell you, hidden complexity and obfuscated parts have been the bane of many software (and hardware) projects. I have no problem with containers being part of the language — what bothers me is that the current set of containers is complicated and inconsistent. We need to refine the current standard before we begin adding new material; otherwise, we build new code on uncertain foundations.

An official template library also leads to another question: Just what is a "standard" container? Some people argue that, for the sake of completeness, we should add hash-based containers to the standard library. But "completeness" means different things to different people; someone might want balanced binary tree containers, while others would prefer B-Tree or r-tree implementations. And then we get into the whole issue of graphical development — and you end up with Java, that tries to be everything to everyone but does few things particularly well.

The current template library is much too heavy, prone to the "feature creep" inherent in a committee-based standards process. And when the standard includes an inconsistency, (list<>.sort() comes to mind), we're stuck with it. Should a list be sorted via its member method or the sort algorithm? And what constitutes a "required" container feature? I use about 20% of the vector<> template 80% of the time; it seems to me that C++ needs a functional hierarchy that stems from a set of concise "base" containers.

We also have the entire realm of garbage collection and "smart" pointers, which is a nasty tangle of divergent opinions. The auto_ptr<> type has numerous logical and practical problems, as does the Boost smart_ptr<>. I don't believe one type of smart pointer makes sense for all applications — and C++'s experience with auto_ptr<> should teach us to avoid providing specific solutions to general problems. I'm still not convinced that automatic garbage collection is a good idea in most applications; it tends to make programmers lazy about controlling their resources.

I've always preached that code should be no more complicated than necessary — and that includes the code I obtain from language libraries. The C++ container types are heavy and detailed, when what we need is a simple set of light, fast containers, with hooks for adding algorithms that fit individual application needs.

Anything else is trying to be Java. ;)

yes, PHP support is still lacking a bit from our beloved OS X...

here are some links on the subject that may be of some interest:

http://developer.apple.com/internet/macosx/php.htm l - Apple's PHP page where they "admit" that...

Like all Unix variants, Mac OS X has it quirks. It is also a young OS, and not all Open Source or Unix software projects have had time to test their programs on the Mac and make the necessary alterations

Linked from that last page is http://www.entropy.ch/software/macosx/php/ - a site concerning a prebuilt PHP Apache module for OS X... they are working on the 4.2.0 version and I don't have a doubt that these folks will deliver...!

Installing pre-compiled MySQL, PHP and Graphviz packages on a OS X-running powerbook was a great (and quick!...) experience. May everybody taste the sweet pleasure of writing PHP code on their balcony! :-)

Greg

I remember them demonstrating some of their technology at a careers fair at Cambridge University. It was pretty spectacular. For example, take a look at some of the applications of their Sentient Computing Project.

Now for the good news - VNC lives on!

First, the current version of VNC will continue to be available at the original web address, which will soon be re-hosted at Cambridge University where AT&T continues to sponsor research.

Second, the creators of VNC are planning a venture to independently support and develop VNC as an ongoing open source project. You haven't heard much from us recently because we've been busy with other projects such as the Broadband Phone, but now that we have the opportunity :) we're back on the case.

Watch the VNC website, the mailing list, or slashdot for an announcement "real soon"

The VNC Team

A large number of people use the weather station on their roof - I use a little app written by a friend that displays the outside temperature next to my CPU monitor on my taskbar. They have fairly comprehensive archives for the last few years, so it would be a shame if this service stopped as well.

I was looking ahead to the Broadband Phone too! They even had a wireless version running. Sigh.

It's a limitation in most PC bioses, that you must have a video card to be able to boot. Any old card will do. My suggestion -- buy some old S3 Virge cards, for like $15 a piece.

Also, if you still want to use the Windows GUI, try VNC.

From prior Bell Labs ...

• www.research.avaya.com

• www.research.att.com AT&T

• www.bell-labs.com Lucent Technologies

And another unrelated Bell Labs with different mission;
at times major news agencies have mistakenly used their logo :-)

Incidentally hopper is a pretty interesting character too. Having worked on the Cambridge ring which was for a while superior to ethernet, he then became involved with the Acorn computers that every Brit of my generation knew and loved at school. He established the then olivetti lab in the mid 80's and is involved with 2 of the three big startups in Cambridge, ARM and Virata. Oh, he also flies planes, is worth a packet and lectures in scruffy jeans.

Incidentally hopper is a pretty interesting character too. Having worked on the Cambridge ring which was for a while superior to ethernet, he then became involved with the Acorn computers that every Brit of my generation knew and loved at school. He established the then olivetti lab in the mid 80's and is involved with 2 of the three big startups in Cambridge, ARM and Virata. Oh, he also flies planes, is worth a packet and lectures in scruffy jeans.

Incidentally hopper is a pretty interesting character too. Having worked on the Cambridge ring which was for a while superior to ethernet, he then became involved with the Acorn computers that every Brit of my generation knew and loved at school. He established the then olivetti lab in the mid 80's and is involved with 2 of the three big startups in Cambridge, ARM and Virata. Oh, he also flies planes, is worth a packet and lectures in scruffy jeans.

Incidentally hopper is a pretty interesting character too. Having worked on the Cambridge ring which was for a while superior to ethernet, he then became involved with the Acorn computers that every Brit of my generation knew and loved at school. He established the then olivetti lab in the mid 80's and is involved with 2 of the three big startups in Cambridge, ARM and Virata. Oh, he also flies planes, is worth a packet and lectures in scruffy jeans.

I was curious what the lab was working on and found a variety of multimedia materials on their anonymous FTP site here:

ftp://ftp.uk.research.att.com/pub/videos/qsif-200/

What I've seen so far is interesting though not earth-shattering. Take a look.

-David

Things that have come out of AT&T Labs Cambridge recently:

The Active Bat system, which can locate in 3D better than any other deployed system. They are using Bat transmitters as mice in the air, on 50 inch plasma screens. Now that's a cool interface.

A broadband phone, rolled out across the entire staff, which lets then see train timetables, share a doodling screen during phone calls, have active directories so that they can call the nearest phone to someone (c.f. Bat above)

At least visit their website before you start trolling. You might even learn something.

You may want to try AT&T's sfio, coauthored by David Korn of the shell by the same name fame.

Well, I wouldn't be all that interested by this story except for that mention of time scaling.

I have been using my mp3 players of late (I have a d-link 32 meg thing and a Rio Volt) to listen to Project Gutenberg etexts converted to mp3 with TextAloud MP3 using AT&T Natural Voices speech engine. (You can hear a sample here.

Anyway, with time scaling, I'm guessing I'll be able to choose how fast to listen to my texts. That's exciting because your ability to listen to these things at high speed increases the more you listen. (As blind people can listen to audio books at surprisingly high speed, you will find your ability to listen increases as your practice increases.)

So, this seemingly minor feature could actually revolutionize the way you take in media.

Well, I wouldn't be all that interested by this story except for that mention of time scaling.

I have been using my mp3 players of late (I have a d-link 32 meg thing and a Rio Volt) to listen to Project Gutenberg etexts converted to mp3 with TextAloud MP3 using AT&T Natural Voices speech engine. (You can hear a sample here.

Anyway, with time scaling, I'm guessing I'll be able to choose how fast to listen to my texts. That's exciting because your ability to listen to these things at high speed increases the more you listen. (As blind people can listen to audio books at surprisingly high speed, you will find your ability to listen increases as your practice increases.)

So, this seemingly minor feature could actually revolutionize the way you take in media.

• basically they've extended RDP (which was available in W2K Server) onto the desktop

Yeah, and they failed the laugh test on that one too. Did you see the open ports on the IIS machine that Microsoft belatedy moved their We Have The Way Out *nix bashing site to? Port 5900 was open and exposed to the outside world. That's vnc, a freeware open source remote desktop viewing system that does everything that RDA does... only without the encryption. It's recommended to run it over ssl or similar, but apparently that's a bit complicated for the bulging brains at Microsoft.

That shows that Microsoft don't use their own products, that they happy to use "insecure" open source products when it suits them, and (once again) that they neither know nor care about security. But at least they set a good precedent: the first thing I did to my WinXP box was to make sure that RDA was off, and to download and compile VNC (checking that there's no back doors) - but only after I'd set up zebedee, an ssl tunnel, for it to run on. Sigh.

5900 (VNC!!!: no Terminal Services eh Microsoft?)

Not only that, but VNC is that dreaded, evil, viral, open-source type software. Under the GPL no less (http://www.uk.research.att.com/vnc/gpl.html for the link leary).

Gupta seems to have missed PEN (formerly PicoNet), a low power network developed by AT&T Labs Cambridge (formerly Olivetti and Oracle Research Labs). They built some prototype hardware and designed power-saving radio protocols for it. Further work with PEN has been done at the Laboratory for Communications Engineering at Cambridge University.

Gupta seems to have missed PEN (formerly PicoNet), a low power network developed by AT&T Labs Cambridge (formerly Olivetti and Oracle Research Labs). They built some prototype hardware and designed power-saving radio protocols for it. Further work with PEN has been done at the Laboratory for Communications Engineering at Cambridge University.

What you need is to combine VNC with a distributed (ie. multiple-computer, at multiple-locations, with multiple-ISPs on multiple-backbones) setup. I'm not aware of any automated products for this, what I've done in the past is to have an main CVS repository on one box, and a cron-job that updates it every night from a different machine. If you lose that main box, you'll have a worst-case day old copy of the repository on the other machine, and can still continue to do your work. If it's a big problem with several days of downtime, you can then set the backup as the main CVS server.

This could work for any number of "backup" computers, but even having two will pretty much eliminate the downtime you've suffered before. And you can use CVS to backup all sorts of stuff (code, IDE config files, even binary data).

In the *nix world, esp. if you're used to a command-prompt instead of a GUI, using VNC is often overkill - I tend to use an OpenSSH java applet (such as AppGate's excellent MindTerm), which can run over slow connections without issue, and can be served from any number of computers.

(However, for a Wintel solution, you'd have to find a configuration-management tool that can automatically update stuff for you - not sure if Source-Safe or StarTeam can do this).

Years ago, at a JavaOne, Sun bragged about their "webtop", which was based upon their iPlanet software. It sounded like a dream come true: all Sun employees had little authentication devices for coming up with their keys; they could fire up any Java enabled browser, bang in their keys, and access all their email, newsgroups, web pages (including publishing) and so forth anywhere that an internet connection was available.

I thought it was terribly exciting. But I haven't much from it since, so I'm guessing it was yet another big hype job from Sun, with little to support it.

I tried out the iPlanet demo a few months later, but it was horribly confusing and complicated to set up, and I never was able to actually get it running. Not exactly ready for prime time. (It was an acquisition, I believe, that was never properly integrated and deployed in a slick package.) Recently, I think Sun has made some rumblings about iPlanet, so maybe it will be resurrected in a more usable fashion.

The idea of divorcing oneself from a specific piece of hardware for their desktop is going to be the future. If you think about it, the popularity of HotMail and other web-based mail services is very much a specific case of this for email. By making the application web-centric, you free yourself from specific hardware.

One step in the right direction for non-web email is using IMAPI instead of POP. I made the switch awhile ago, and am feeling much more "mobile" with the mail stored on a central server, especially given that a large chunk of my work revolves around email. Any machine running Windows or Linux has a browser or mail client that support MAPI, and I can get at all my messages.

The biggest problem with this is that other than email and browsing (which is inherently portable), few applications have been made to work in such a distributed fashion. Also, even when an application has been done in such a manner, the ubiquity of connectivity is an issue. If you're away from the home or office, or on a plane, or elsewhere, live connectivity can be a problem. This issue will be reduced with time, as airlines add internet connectivity, CDPD and other wide area connectivity becomes more prevalent.

But the apps are still an issue.

For those who haven't checked it out, VNC deserves a good look. You can create a virtual desktop on Windows (or more importantly, Linux :-), which can be access from any Java-enabled browser (or more efficiently, using the VNC client application, which is open source, and heavily ported).

In a slightly more archaic fashion, the textual VNC equivalent, "screen" (standard in RedHat and other Linux distributions) is a powerful and oft-overlooked utility, which I consider essential for anyone using a command line. Disappearing windows or disconnected telnet sessions become irrelevant with screen, which also allows shared sessions, history, and other amazing stuff.

Regarding VNC, the performance isn't as good as PC Anywhere, but it's usable with 28.8K+, in my experience. There's nothing like going to your hotel, and firing up an X session that is actually running at your house.

I firmly believe the "webtop" or it's equivalent (please God, not .NET) will be the future of the internet. The Linux/Open Source community needs to be careful to make sure it isn't .NET services that dominant this, and that there are other alternatives that meet the same needs as well, or preferrably better. (Or if .NET does dominate, the government or the business community makes sure that the protocols are open.)

-me

I run a VNC session over a zebedee tunnel to my home RedHat box while at work. Every document I create is thusly already at home. It was a bit awkward at first trying to cope with the lag of the interface, but in time you get used to it. After a while of this, I even threw down on some PERL scripts to nightly ftp my documents from my linux box to my wife's XP machine, because 'hey, you never know. Shit happens'. In this way I always use the same desktop and all my files on hand. Even when I go to the parental units houses, I can just jump right in and grab files (SFTP), or check something for them (What's that address?). At work they've given me a XP laptop to do my stuff, but all I use it for launching my terminal, and playing BB King and Charles Mingus all damn day.

Here's the big list of links of a lot of VNC derivatives in case anyone's wondering where all these contribs suddeny appeared from (like me ;-) )

What this guys seems to be looking for is a really efficient implementation of XWindows.

You mean like this?

:)

The Linux Terminal Server Project is exactly what you're talking about. I've been using it at home here to play around with for a few months now. It's really slick. I have a bunch of my old computers that would otherwise be in the dumpster that are right now serving as terminals. And they're pretty fast, since all the apps run on my big Athlon box.

It works by netbooting from your server. Some kind of bootrom code, either on your network card or on a floppy disk, initalizes the network card. It uses DHCP to find its own IP address, and then it uses TFTP to download a small Linux kernel over the network. This loads up and uses an NFS-mounted root to run an X server on the local computer. The X server connects back to the main server by XDMCP, and you get your XDM/GDM/KDM login window.

The LTSP guys have done a great job packaging this all up. Take a look. And as for your requirement of running it on a Windows box, see Cygwin's XFree86 port to Windows. You can use it to connect with XDMCP. Of course, I don't know why you wouldn't just pop in a bootdisk...

The biggest drawback to this approach is remote access security. Look at that paragraph and how many daemons and services you need to have running. But I imagine that if it was secured well enough, it'd be fine. Actually, there is a way to make this all go over VNC (or VNC with compression). It's not as fast, but at least that's only one TCP port and a lot easier to get by firewalls.

There's a great bunch of guys working on this project. And its nice to be able to connect to #ltsp on irc.openprojects.net and get the lead developers to answer your questions.

Michael F. Robbins

VNC works great with Windows & Linux clients, and Linux servers (Windows servers are limited to a single desktop at this time I believe). You need to install a VNC client, but I consider it the best alternative.

• details here
• screenshots cose we love 'em.

VNC works great with Windows & Linux clients, and Linux servers (Windows servers are limited to a single desktop at this time I believe). You need to install a VNC client, but I consider it the best alternative.

• details here
• screenshots cose we love 'em.

http://www.uk.research.att.com/spirit/

All you need is VNC.

Well, you can run VNC and you'll have access to the same exact desktop no matter which machine. There are native viewers, as well as the Java one. Its small, platform independent, and GPL'd.

VNC
VNC
VNC

Let's say it again, VNC.

Oh, yeah. Here's a link.

This Link shows how a single window can be shared using VNC....You just have 50 Word sessions open, each person can connect to their own single window...(might need some more code tweaking, but may be possible).

So, if someone wanted to write a "Citrix Like" app based on VNC, it is most certainly possible

So, if I'm reading this right, it will run Windows CE, and allow you to do remote desktop stuff with a Windows XP machine via Windows Terminal Server. That sounds really nifty, except I don't want to run Windows XP, ever.
Perhaps it could also net-boot and run Linux Terminal Server, instead. Are there any good X Servers for Windows CE? Any GPL'd? I think there is a GPL'd Java one that might run on Windows CE.
I know that VNC runs on Windows CE. While VNC is very slow when connecting to a Windows machine, it is quite fast connecting to a Linux machine. From what I understand, it isn't as fast as Cygwin/Xfree ,but the install is sooooooooooo much easier.

It doesn't seem to be a new concept at all.