It's my understanding that kernel modules could get around this legal requirement, because kernel modules are not derivatives of the kernel, and are therefore not subject to the GPL. As I'm sure I'm not the first person to think of this, there has to be a reason this doesn't work. Does anybody know why or why not?

Somebody above pointed out the fact that what he said could be grounds for a lawsuit under unfair competition laws. So Microsoft was left between a rock and a hard place. If they didn't say anything, Apple could try to take them to court over copying their specific look and feel ideas (the 1990's case covering this is also mentioned above), but if they did try to officially retract his statement, the Streisand affect comes into play.

They were in a damned-if-they-do-damned-if-they-don't situation, because let's face it: they did copy a lot of ideas from other operating systems. Perhaps they should try to come up with their own ideas instead of playing follow the leader/s (Apple and the KDE team) in UI design.

Gee, now if only somebody thought of a solution that worked for *removable media* like the parent asked in addition to the NFS part, we wouldn't have to read snarky replies on Slashdot that don't really answer the question completely.

Ah, I see your point, although my understanding is that X forwarding is faster than VNC and RDP. There are advantages and disadvantages for everything though, I suppose.

I used X forwarding over SSH from my school's computers to run Firefox to get access to websites behind the school firewall (the VPN software didn't work with my computer - they switched from the Cisco VPN, which had a nice plugin for NetworkManager, to Juniper, which needed a bunch of messing with and still didn't work on my standard Ubuntu desktop - yes). Without the networking functionality of X, I would have had to call up a friend and divulge my password to the school's computers (which would have voided my agreement to use the network to begin with) or drive the 20 minutes to school to use the computers for 5 minutes and then drive back home - roundtrip time: 45 minutes of my life, 40 of which I thankfully spent doing other (admittedly sometimes non-productive) things.

Do not underestimate the utility of networking functionality in today's software environment; Bill Gates wrote an email to his employees in the *90's* about the rise and usefulness of the internet, and when his employees did not share his vision, Google - a company that also recognized the potential of the internet and the browser - managed to become Gates' worst nightmare: a relevant and *widely successful* competitor to the entrenched monopoly that is Microsoft. To successfully compete against a monopoly is difficult, to surpass the monopoly makes the feat even more impressive (even though Google may turn into a bad monopoly in the future, much as Microsoft did).

I am not familiar with the architecture of X, but somebody smarter than I needs to sit down and think about the architecture of X, whether or not it can improve, and whether or not it needs to be rewritten from scratch. FOSS should not fall into the trap of rewriting everything as pointed out by somebody smarter than I (I think it was Jamie Zawinski, who has criticized the X project for quite awhile, perhaps not the best reference).

The more people you have, the more hardware manufacturers and software companies take notice and actually care about Linux. When hardware companies release open source drivers (and even proprietary drivers - look at nVidia) and when software companies release software for Linux (Wolfram's Mathematica, MathWorks' Matlab, National Instruments' LabVIEW), Linux users do benefit. As more applications become more available on Linux, it becomes a more mainstream desktop OS and the interest in Linux grows and quality of software for Linux improves. Once Linux is mainstream, more and more hardware will "just work" because the Linux kernel includes more drivers. The more the hardware "just works", the easier it is to configure for everybody using clickable GUI tools instead of looking up commands and command options (the command line, while sometimes easier to use than GUIS, is not necessarily hard to use for everything, but it is hard to initially learn). The easier it is to configure, the more "Regular Joe" users will be able to use Linux, not only making Linux a more visible solution to Windows, but making all of FOSS a more visible solution to proprietary software. Finally, as the number of Linux users increases, presumably the number non-duplicate bugs are reported (as some users are technically inclined enough to submit a bug, but not enough to fix it - something even Linus, if I remember correctly, has noted as an advantage of FOSS), making architecture errors easier more visible, giving developers to improve the architecture, which hopefully leads to happier developers and users. All of these changes, while insignificant when standing alone, lead towards a better world of software for everybody, including you. So whether or not Linux developers are aiming for world domination or not, the "Regular Joe" users of the Linux desktop are as important as the developers themselves.

Like Agilent scopes better. At my school, the physics department has two different models of Tektronix scopes, the only difference being that one model can save results to a usb drive. The only problem I've had with them is that they sometimes format the damn usb drive, so you sometimes have to continually swap it out and download the data to a computer, which kind of defeats the purpose (a bit). So, to minimize the problem of people's usb drives and data getting eaten, the lab profs decided to get their own usb drives, which promptly resulted in a few of the usb drives rapidly evolving legs and running out of the room (they were stolen for those of you with no sense of humor). Both models of Tektronix scopes will lie to you if you hit the autoset button.

The engineering department at my school has Agilent scopes, and while they are larger (though they aren't stationary), they have a slightly more intuitive interface. They also don't format usb drives willy nilly without warning you and making sure you know whats going on. They autocalibrate their own probes, and they have storage space for the manual and probes in a compartment on top. Their autoset buttons are better than the Tektronix scopes, but still ended up giving me garbage settings on a few signals. I do vaguely remember somebody having to learn to use the interface and having a little trouble with it, but in a class of 18, that was probably not statistically significant.

My impression was that the Agilent folks had to use their own instruments, which gave them the incentive to go that extra tweak in UI improvement. They also had a spinning wheel for when the drive was being accessed, similar to the Apple beachball. Which looked cool even if it didn't serve a useful purpose.

Disclaimer: I have a few friends that work at Agilent, and love it. I had tried to get a job at Agilent for the summer, but was turned down. I even baked them a cake.