The thing is, we'd have to assume then that the equipment you just stated was not damaged in the EMP. If that's the case, then yeah, they could just fly a freaking helicopter or drive a van to the fault. Also, I dislike people who just randomly quote binary search algorithm and toss nothing with it. If you thought someone would just have a single person walk the entire line, I have a bridge to sell to you. I apologize for assuming that both of those points would be obvious.

Oh well the military already has tried pacemakers versus EMPs. And typically during planned extended outages, like hurricanes, life support patients are typically taken elsewhere. You'll be surprised what you can find within the public domain from the US military.

No, not really. If you had an EMP that could cook via electromagnetic radiation the electrical grid of the US, you'd have worst problems than EMP caused black outs to deal with.

The whole idea of an effective EMP is to fry as much cooper/aluminum wire as one could. Think of really effective EMPs being more like lighting and less like nuclear detonations, since using a nuclear detonation is like trying to cut off the kitchen lights using a bulldozer and thirty tons of sand. If we're strictly talking EMP, lighting and all the static discharge family is your better bet.

So with that said, a lot of electrical companies are prepped for pretty bad EMPs with response teams, though I'm pretty sure most of America would find that hard to believe when their power does actually go out. One thing that is actually going to be slightly more difficult to deal with is first finding the point of failure, bring it down, and then put up the replacement. With a bomb, it's pretty easy to figure out where the failure is and the upshot is that the part to be replaced is already on the ground or missing completely so you can skip that whole removal step. Yeah, we might be talking several kilometers needing to be replaced with a nuclear device, but you know from the first second you need a couple of hundred kilometers of wire. A good EMP keeps you guessing and has a dozen or so employees walking hundreds of kilometers of wire trying to find the failure. Better sections of the US grid have more fine grain reporting points so maybe on a dozen or so kilometers need to be checked. However, the actual transmission lines are the key to a good EMP. That said, you don't need a big "bomb" to be effective, you just need coordinated attacks on major transmission lines. However, doing that alone is just more of a major disruption, rather than a major blow to the nation.

Additionally, power generating plants usually have a lot of counter measures for EMPs. So you really aren't going to take out the generators. It's silly to think that someone could without a massively coordinated attack. Especially if we're talking strictly EMP here. If we're talking a nuclear device, again, you've got bigger problems especially if you had enough to take out all the major plants in the US.

The real danger here, I guess, is consumers. Some EMPs can fry pacemakers and pretty much anyone on life support is dead in a massive EMP. However, it's not the end of society and more so, hardly the end of the US. You can take any example of when some large section of the US had power knocked out for several weeks. A massive EMP would be roughly equal to a hurricane without any advance notice that hit a large section of the US. If you were lucky enough to hit the entire US then multiply the figures in your head by that amount. However, the end of days for the US, hardily. EMP weapons in real life would cause some death but for the most part the US would be tired of martial law long before the US fell under the pressure of an EMP weapon.

The real tactical value of EMPs is not as some silver bullet, but as a disruptive force that is soon followed by other forms of attack. Additionally, you'd want your EMP to be as quiet as possible and look as much like a lighting strike as possible on the grid. Anything else and there would be way too much attention drawn that would get people ready for the obvious next strike. Thinking that an EMP would be a good primary strike is silly. Additionally, some have thought about EMPs in asymmetrical warfare contexts and while they would play a good role in the demoralizing aspect of that, it's just simpler to buy a ton of fertilizer and diesel fuel as opposed to trying to construct something massive enough to disrupt more than just a few dozen people. In other words, the reason low tech seems to win in asymmetrical warfare and terrorist operations is that you get more bang for buck so to say.

I think when you consider it carefully, those who would toss EMPs around like we are under some imminent threat of them, are doing nothing more but trying to push some agenda. EMPs just really aren't that great of a weapon and the technical curve to building ones that would be worth the time and money is just too high for your casual mayhem makers.

I'm a little curious, why you bring up the link to systemd? Is it because it prevents the stack from running on BSD?

If we honestly wanted to follow the Unix philosophy, we should add X11 to that list as well. There's nothing about X that follows the Unix philosophy any more.

I know that some here on Slashdot will be at a loss for sympathy for the project being in such dire circumstance. However, the key thing that some should remember is that a lot of what the GNOME hackers do, goes into the base for many other projects as well. Much of Linux Mint is an eclectic mix of Ubuntu and GNOME. Likewise for Elementary OS.

So while we might be able to argue if this project has finally run its course, which I do want to add that the foundation running out of reserves hardly equates to the death knell for GNOME. One of the things we shouldn't do, or at least it would be in a very short sighted, is think that the actual GNOME Desktop and how ... "not so great," they've ran that ship plays into all of this. Agreed, the people in the project have become quite hard headed, but honestly which OSS project hasn't by now? However, there are a lot of people (Canonical *cough, cough*) who find their software very useful and hardly give anything back, at least to the foundation.

PS: Being using beta now for a month plus some. I honestly think it is getting better but it does need quite a bit more work. I guess I just wanted to add that after seeing all the f*** beta sigs.

Totally hit nail on head there. It's always been a gamble which of the different frameworks will mesh out in the petty internal battle. It's hard to bank on any MS tech because you never know when it might just up and vanish. I've seen some say, oh then I guess we should just stick to COM then. No, but changing the game as often as I change my phone isn't going to help you win converts.

The whole conversation is less about SSH and more about X11. I don't think anyone is saying SSH is slow.

I think, and I dislike this disjoined thread so maybe we can move it all back into one thread, if you ask the Wayland team, they'd be the first to tell you that X11 isn't going anywhere. There is a difference between wishful thinking and the reality. The Wayland people are anxious to have a more testers for their display system. Distros are ultimately the ones who decided to "cram" it down your throat. The Wayland people have good reason to see the end of X, because many of the X developers are on the Wayland team. They see X as taking time away from Wayland, however, they acknowledge that X is a pretty important piece of software and that supporting it for the near future is a pretty big item on the list. Additionally, yeah they are a bit rude, but they have non-stop email after email on their list yelling about how wrong they are. So yeah, maybe they've gotten a bit jaded. Do you blame them?

Feverish much?

There isn't a point because you'll just sit here and come up with half brained reasons as to why it is wrong. So why even go down the road to begin with? It'll just lead to a conversation that neither of us really want to have, wouldn't you agree?

X11's tunneling isn't worth it to me and that's the way I feel about the matter, case closed. It obviously is something you are worth foaming at the mouth about, so if that's what tickles your fancy so be it. We just don't agree on how we'd go about implementing like solutions, like I said, we'll just have to agree to disagree. If you find that hard to swallow, that isn't exactly my problem. However that's what we have in front of us. Neither one of us agrees on the solution, presenting our so called facts will just stroke more flames, and I'm pretty sure both of us have better things we could be doing with our time than to talk about pieces of software.

Do you have anything else you'd like to add? I'll be more than happy to enter a reasonable debate about the matter, but I doubt from your tone that we'd have a reasonable debate. Don't you think that this thread might be a bit too hot to really have any useful discussion?

What do you think happened to XOrg? XOrg has suffered greatly from a lot of in fighting that sounds a lot like what you and I are having at the moment. Why do you think that has happened? You say the split resulted in a much better display system, why do you think that would not happen with Wayland? It is one thing to have outside people say X11 sucks, but it is an altogether different thing to have X11 developers say X11 sucks. Why do you think they say that?

I'm more than happy to entertain the thought that X11 has some saving grace, but I and a lot of X11 developers are drawing blanks when posited the same question. The world has changed and X11 is too large to move as agile as the world would want it. The exact same thing could be said about IPv6, or the new firewall code in the Linux kernel, or btrfs, and so on. There are a ton of pieces of software where better, more agile things have come along and people just aren't ready to give up the old to make way for the new. Eventually those new things will become old and we'll all have the same arguments all over again. All that being said, maybe you can understand why I am so tired of these rehashed arguments. I've had twenty years of having these types of (quote fingers)discussions(quote fingers), I'm pretty much done with them. Maybe, you aren't that way and perhaps that's a fault of me, but I just can't stomach this kind of back and forth. Hopefully you can understand that.

So like I said, maybe we can *both* cool off a bit and have a reasonable discussion some point down the road. However, I just don't think we're going to have that at this moment.

If you honestly want my opinion on remote options I'll relent, but only if you wish to push the matter more. Whatever it is that gives you some piece of mind.

There obviously isn't any point. I don't think I'll ever be able to convince you personally that the feverish hang ups that you have with X are the exact same things that has played out on X11 mailing lists for the last decade and a half that has gotten X11 nowhere. The whole XFree/X.org breakup was all politically motivated. Same thing here. There is just too much drag to worry about trying to fix X11, it just isn't worth the headache to fix. Fixing X just causes more problems and more headaches. Might as well start clean.

There are things that are worth the argue, but I doubt that we'll ever be able to agree on anything with this topic. For me, X11 just isn't worth continuing to prop up. For you, you see it as some vital thing. We'll just have to agree to disagree.

If you think X11 is going away anytime soon, then you've been horribly misled. Yes the Wayland folks want it to go away but we're in too deep for X11 to disappear anytime this decade. That's just a show of how dug in some are with X. I'm not sure what IPC issue you speak of but I do know that X11 takes in 1270ish points in the API for IPC alone. Wayland's stands at 135. The API is greatly simplified so if you have a better scheme in mind, you have a lot of liberty to, 'go it alone'. The last bit isn't a head in the sand. It's a not our domain to implement thing. Which I think it's a good thing that they focus on the display stack and not network bits.

Also, I like how you pointed out X tunneling over ssh. Which has been shown time and time again as the slowest method for remote connections into a system hands down. Yes, it's nice in that it is easy on admin costs and takes very little to get up and running, however, that comes at the cost of it being slow. Just to compare, Doing Mathlab via X11 over ssh versus (randomly grabbing a tool out of thin air that I know is a really bad choice) VNC. X11 over ssh is close to about 35 seconds to finally see the window appear on the remote client. VNC is roughly three tenths a second. That's just doing wall clock figures so don't take that as a scientific thing, but you can check out Google and see all the "Why is X11 tunneling so slow" hits.

I'm just saying, if X11 tunneling is what you are hanging onto, there are tons of other ways to do the exact same thing with less of a bandwidth cost and less suck. Again, I get that X11 tunneling is really easy to setup and other solutions can be a pain to get setup up. But you pay the setup cost once, you pay the bandwidth fee per usage.

As you know there are lot of protocol-oriented application implementations out there are insanely useful and their effectiveness is hardly ever challenged. For example, TCP as a generic and NTP as a more application specific example.

Being protocol based isn't instantly a mark of slowness, just like using a hash map isn't a sign of slowness versus a binary tree. It is how it is used and X has had so much tacked on and so many extensions that need to be negotiated, it's just turned into a really bad use of a pretty good protocol.

I think I had covered that quite well. Most DEs are working well with the Wayland team of patches upstream to ensure that if there is a specific thing they need, they get an out for it. If it is a generic fix, everyone wins as the patch is taken upstream. The problem with MIR is that if the MIR team changes on the dime the underlying API, your client library is going to be needing a slight overhaul. Libraries are a great way to solve what we're talking about so long as the libraries can be kept fairly stable. That's not really a sure thing, or at least I don't think anyone from the MIR team has said, "We are for real, for sure that this part is never going to change except in major revisions." Last I checked, pretty much the whole collection of things in MIR were all works in progress.

Again, nothing bad about MIR except that with it being mostly developed by a single team, it just isn't ready for other to try and base compositors off of it, just yet. However, just because something is implemented is a certain way, doesn't mean, that is gets tossed into the crap realm. Even binary trees are a bad idea for some applications. Go grab a book by Bjarne Stroustrup, he makes a way better case for fighting this kind of thinking than I ever could. That's not a vote of confidence for C++, but that the guy makes some really good generic programming arguments for bad thinking in the Computer Science arena.

Which is why X cannot be fixed the way it should be fixed. There is always going to be some group of people that uses some random function of X that is horribly ineffective. I can pretty much bet you a lunch that one can point to any random function of X and find some group of people that rely on that function. Most of these functions just get in the way of the majority of users that just want to use desktop applications. I mean come on, who really uses XPrint? Or needs a non-rectangular window? I guess if that's what you need, go for it, but Wayland offers a better method for implementing your own method for doing those things rather than a bunch of hacks in X11.

It is also a horribly ineffective way of doing it, but no one is going to stop you from using one of the worst methods of doing that. If that's how you roll, then that's how you roll. But mind you, it is the reason why X cannot be fixed. The network transport is feature of X is horrible because most WMs draw to a pixbuf and then send the pixbuf to the remote client. The only thing is, there are a ton of other clients out there that will draw you a picture and send it across the wire a whole lot faster than X can do it. With built in security too to boot. However, I'm not here to convince you of anything except this. You cannot fix X, it is broken and simply cannot be fixed because, among other reasons, there are too many diehards that will cling to every little feature of X11. These types of discussions where people argue old crusty features are the reason why X developers started working on Wayland. They needed a clean break, they needed a new project where people wouldn't be sitting there yelling about dropping feature A, B, or C. If Wayland doesn't fit for you, don't move to it, that easy. But X11 is old, slow, and bloated and if that's what you need, then go for it.

That's correct, you can as in technically possible. However you cannot because it would cause some breakage with legacy applications that the foaming mob of X11 zealots would stone the developer. The X11 fan base is so feverish they'll scream at any and all changes, I mean look at them when you threaten their 'network transport'. So while it is technically possible, it is impossible given the current inertia that the X11 fans have for change.